MyArxiv
Robotics 36
☆ Bioinspired Soft Quadrotors Jointly Unlock Agility, Squeezability, and Collision Resilience
Natural flyers use soft wings to seamlessly enable a wide range of flight behaviours, including agile manoeuvres, squeezing through narrow passageways, and withstanding collisions. In contrast, conventional quadrotor designs rely on rigid frames that support agile flight but inherently limit collision resilience and squeezability, thereby constraining flight capabilities in cluttered environments. Inspired by the anisotropic stiffness and distributed mass-energy structures observed in biological organisms, we introduce FlexiQuad, a soft-frame quadrotor design approach that limits this trade-off. We demonstrate a 405-gram FlexiQuad prototype, three orders of magnitude more compliant than conventional quadrotors, yet capable of acrobatic manoeuvres with peak speeds above 80 km/h and linear and angular accelerations exceeding 3 g and 300 rad/s$^2$, respectively. Analysis demonstrates it can replicate accelerations of rigid counterparts up to a thrust-to-weight ratio of 8. Simultaneously, FlexiQuad exhibits fourfold higher collision resilience, surviving frontal impacts at 5 m/s without damage and reducing destabilising forces in glancing collisions by a factor of 39. Its frame can fully compress, enabling flight through gaps as narrow as 70% of its nominal width. Our analysis identifies an optimal structural softness range, from 0.006 to 0.77 N/mm, comparable to that of natural flyers' wings, whereby agility, squeezability, and collision resilience are jointly achieved for FlexiQuad models from 20 to 3000 grams. FlexiQuad expands hovering drone capabilities in complex environments, enabling robust physical interactions without compromising flight performance.
comment: 26 pages, 12 figures, 2 tables, 9 videos (not yet disclosed, awaiting peer review)
☆ Stable and Robust SLIP Model Control via Energy Conservation-Based Feedback Cancellation for Quadrupedal Applications
In this paper, we present an energy-conservation based control architecture for stable dynamic motion in quadruped robots. We model the robot as a Spring-loaded Inverted Pendulum (SLIP), a model well-suited to represent the bouncing motion characteristic of running gaits observed in various biological quadrupeds and bio-inspired robotic systems. The model permits leg-orientation control during flight and leg-length control during stance, a design choice inspired by natural quadruped behaviors and prevalent in robotic quadruped systems. Our control algorithm uses the reduced-order SLIP dynamics of the quadruped to track a stable parabolic spline during stance, which is calculated using the principle of energy conservation. Through simulations based on the design specifications of an actual quadruped robot, Ghost Robotics Minitaur, we demonstrate that our control algorithm generates stable bouncing gaits. Additionally, we illustrate the robustness of our controller by showcasing its ability to maintain stable bouncing even when faced with up to a 10% error in sensor measurements.
☆ EveryDayVLA: A Vision-Language-Action Model for Affordable Robotic Manipulation ICRA 2026
While Vision-Language-Action (VLA) models map visual inputs and language instructions directly to robot actions, they often rely on costly hardware and struggle in novel or cluttered scenes. We introduce EverydayVLA, a 6-DOF manipulator that can be assembled for under $300, capable of modest payloads and workspace. A single unified model jointly outputs discrete and continuous actions, and our adaptive-horizon ensemble monitors motion uncertainty to trigger on-the-fly re-planning for safe, reliable operation. On LIBERO, EverydayVLA matches state-of-the-art success rates, and in real-world tests it outperforms prior methods by 49% in-distribution and 34.9% out-of-distribution. By combining a state-of-the-art VLA with cost-effective hardware, EverydayVLA democratizes access to a robotic foundation model and paves the way for economical use in homes and research labs alike. Experiment videos and details: https://everydayvla.github.io/
comment: Submitted to ICRA 2026
☆ Sample Complexity of Distributionally Robust Off-Dynamics Reinforcement Learning with Online Interaction ICML 2025
Off-dynamics reinforcement learning (RL), where training and deployment transition dynamics are different, can be formulated as learning in a robust Markov decision process (RMDP) where uncertainties in transition dynamics are imposed. Existing literature mostly assumes access to generative models allowing arbitrary state-action queries or pre-collected datasets with a good state coverage of the deployment environment, bypassing the challenge of exploration. In this work, we study a more realistic and challenging setting where the agent is limited to online interaction with the training environment. To capture the intrinsic difficulty of exploration in online RMDPs, we introduce the supremal visitation ratio, a novel quantity that measures the mismatch between the training dynamics and the deployment dynamics. We show that if this ratio is unbounded, online learning becomes exponentially hard. We propose the first computationally efficient algorithm that achieves sublinear regret in online RMDPs with $f$-divergence based transition uncertainties. We also establish matching regret lower bounds, demonstrating that our algorithm achieves optimal dependence on both the supremal visitation ratio and the number of interaction episodes. Finally, we validate our theoretical results through comprehensive numerical experiments.
comment: 53 pages, 6 figures, 3 tables. Published in Proceedings of the 42nd International Conference on Machine Learning (ICML 2025)
☆ ETHOS: A Robotic Encountered-Type Haptic Display for Social Interaction in Virtual Reality
We present ETHOS (Encountered-Type Haptics for On-demand Social Interaction), a dynamic encountered-type haptic display (ETHD) that enables natural physical contact in virtual reality (VR) during social interactions such as handovers, fist bumps, and high-fives. The system integrates a torque-controlled robotic manipulator with interchangeable passive props (silicone hand replicas and a baton), marker-based physical-virtual registration via a ChArUco board, and a safety monitor that gates motion based on the user's head and hand pose. We introduce two control strategies: (i) a static mode that presents a stationary prop aligned with its virtual counterpart, consistent with prior ETHD baselines, and (ii) a dynamic mode that continuously updates prop position by exponentially blending an initial mid-point trajectory with real-time hand tracking, generating a unique contact point for each interaction. Bench tests show static colocation accuracy of 5.09 +/- 0.94 mm, while user interactions achieved temporal alignment with an average contact latency of 28.53 +/- 31.21 ms across all interaction and control conditions. These results demonstrate the feasibility of recreating socially meaningful haptics in VR. By incorporating essential safety and control mechanisms, ETHOS establishes a practical foundation for high-fidelity, dynamic interpersonal interactions in virtual environments.
comment: 8 pages
☆ SAD-Flower: Flow Matching for Safe, Admissible, and Dynamically Consistent Planning
Flow matching (FM) has shown promising results in data-driven planning. However, it inherently lacks formal guarantees for ensuring state and action constraints, whose satisfaction is a fundamental and crucial requirement for the safety and admissibility of planned trajectories on various systems. Moreover, existing FM planners do not ensure the dynamical consistency, which potentially renders trajectories inexecutable. We address these shortcomings by proposing SAD-Flower, a novel framework for generating Safe, Admissible, and Dynamically consistent trajectories. Our approach relies on an augmentation of the flow with a virtual control input. Thereby, principled guidance can be derived using techniques from nonlinear control theory, providing formal guarantees for state constraints, action constraints, and dynamic consistency. Crucially, SAD-Flower operates without retraining, enabling test-time satisfaction of unseen constraints. Through extensive experiments across several tasks, we demonstrate that SAD-Flower outperforms various generative-model-based baselines in ensuring constraint satisfaction.
☆ Cleaning Maintenance Logs with LLM Agents for Improved Predictive Maintenance
Economic constraints, limited availability of datasets for reproducibility and shortages of specialized expertise have long been recognized as key challenges to the adoption and advancement of predictive maintenance (PdM) in the automotive sector. Recent progress in large language models (LLMs) presents an opportunity to overcome these barriers and speed up the transition of PdM from research to industrial practice. Under these conditions, we explore the potential of LLM-based agents to support PdM cleaning pipelines. Specifically, we focus on maintenance logs, a critical data source for training well-performing machine learning (ML) models, but one often affected by errors such as typos, missing fields, near-duplicate entries, and incorrect dates. We evaluate LLM agents on cleaning tasks involving six distinct types of noise. Our findings show that LLMs are effective at handling generic cleaning tasks and offer a promising foundation for future industrial applications. While domain-specific errors remain challenging, these results highlight the potential for further improvements through specialized training and enhanced agentic capabilities.
☆ Force-Safe Environment Maps and Real-Time Detection for Soft Robot Manipulators
Soft robot manipulators have the potential for deployment in delicate environments to perform complex manipulation tasks. However, existing obstacle detection and avoidance methods do not consider limits on the forces that manipulators may exert upon contact with delicate obstacles. This work introduces a framework that maps force safety criteria from task space (i.e. positions along the robot's body) to configuration space (i.e. the robot's joint angles) and enables real-time force safety detection. We incorporate limits on allowable environmental contact forces for given task-space obstacles, and map them into configuration space (C-space) through the manipulator's forward kinematics. This formulation ensures that configurations classified as safe are provably below the maximum force thresholds, thereby allowing us to determine force-safe configurations of the soft robot manipulator in real-time. We validate our approach in simulation and hardware experiments on a two-segment pneumatic soft robot manipulator. Results demonstrate that the proposed method accurately detects force safety during interactions with deformable obstacles, thereby laying the foundation for real-time safe planning of soft manipulators in delicate, cluttered environments.
☆ TwinVLA: Data-Efficient Bimanual Manipulation with Twin Single-Arm Vision-Language-Action Models
Vision-language-action models (VLAs) trained on large-scale robotic datasets have demonstrated strong performance on manipulation tasks, including bimanual tasks. However, because most public datasets focus on single-arm demonstrations, adapting VLAs for bimanual tasks typically requires substantial additional bimanual data and fine-tuning. To address this challenge, we introduce TwinVLA, a modular framework that composes two copies of a pretrained single-arm VLA into a coordinated bimanual VLA. Unlike monolithic cross-embodiment models trained on mixtures of single-arm and bimanual data, TwinVLA improves both data efficiency and performance by composing pretrained single-arm policies. Across diverse bimanual tasks in real-world and simulation settings, TwinVLA outperforms a comparably-sized monolithic RDT-1B model without requiring any bimanual pretraining. Furthermore, it narrows the gap to state-of-the-art model, $\pi_0$ which rely on extensive proprietary bimanual data and compute cost. These results establish our modular composition approach as a data-efficient and scalable path toward high-performance bimanual manipulation, leveraging public single-arm data.
comment: Project webpage : https://jellyho.github.io/TwinVLA/
☆ Context-aware Learned Mesh-based Simulation via Trajectory-Level Meta-Learning
Simulating object deformations is a critical challenge across many scientific domains, including robotics, manufacturing, and structural mechanics. Learned Graph Network Simulators (GNSs) offer a promising alternative to traditional mesh-based physics simulators. Their speed and inherent differentiability make them particularly well suited for applications that require fast and accurate simulations, such as robotic manipulation or manufacturing optimization. However, existing learned simulators typically rely on single-step observations, which limits their ability to exploit temporal context. Without this information, these models fail to infer, e.g., material properties. Further, they rely on auto-regressive rollouts, which quickly accumulate error for long trajectories. We instead frame mesh-based simulation as a trajectory-level meta-learning problem. Using Conditional Neural Processes, our method enables rapid adaptation to new simulation scenarios from limited initial data while capturing their latent simulation properties. We utilize movement primitives to directly predict fast, stable and accurate simulations from a single model call. The resulting approach, Movement-primitive Meta-MeshGraphNet (M3GN), provides higher simulation accuracy at a fraction of the runtime cost compared to state-of-the-art GNSs across several tasks.
comment: 35 pages. Submitted to Transactions on Machine Learning Research (TMLR)
☆ Beyond Master and Apprentice: Grounding Foundation Models for Symbiotic Interactive Learning in a Shared Latent Space
Today's autonomous agents can understand free-form natural language instructions and execute long-horizon tasks in a manner akin to human-level reasoning. These capabilities are mostly driven by large-scale pre-trained foundation models (FMs). However, the approaches with which these models are grounded for human-robot interaction (HRI) perpetuate a master-apprentice model, where the apprentice (embodied agent) passively receives and executes the master's (human's) commands without reciprocal learning. This reactive interaction approach does not capture the co-adaptive dynamics inherent in everyday multi-turn human-human interactions. To address this, we propose a Symbiotic Interactive Learning (SIL) approach that enables both the master and the apprentice to co-adapt through mutual, bidirectional interactions. We formalised SIL as a co-adaptation process within a shared latent task space, where the agent and human maintain joint belief states that evolve based on interaction history. This enables the agent to move beyond reactive execution to proactive clarification, adaptive suggestions, and shared plan refinement. To realise these novel behaviours, we leveraged pre-trained FMs for spatial perception and reasoning, alongside a lightweight latent encoder that grounds the models' outputs into task-specific representations. Furthermore, to ensure stability as the tasks evolve, we augment SIL with a memory architecture that prevents the forgetting of learned task-space representations. We validate SIL on both simulated and real-world embodied tasks, including instruction following, information retrieval, query-oriented reasoning, and interactive dialogues. Demos and resources are public at:~\href{https://linusnep.github.io/SIL/}{https://linusnep.github.io/SIL/}.
☆ Let Me Show You: Learning by Retrieving from Egocentric Video for Robotic Manipulation IROS 2025
Robots operating in complex and uncertain environments face considerable challenges. Advanced robotic systems often rely on extensive datasets to learn manipulation tasks. In contrast, when humans are faced with unfamiliar tasks, such as assembling a chair, a common approach is to learn by watching video demonstrations. In this paper, we propose a novel method for learning robot policies by Retrieving-from-Video (RfV), using analogies from human demonstrations to address manipulation tasks. Our system constructs a video bank comprising recordings of humans performing diverse daily tasks. To enrich the knowledge from these videos, we extract mid-level information, such as object affordance masks and hand motion trajectories, which serve as additional inputs to enhance the robot model's learning and generalization capabilities. We further feature a dual-component system: a video retriever that taps into an external video bank to fetch task-relevant video based on task specification, and a policy generator that integrates this retrieved knowledge into the learning cycle. This approach enables robots to craft adaptive responses to various scenarios and generalize to tasks beyond those in the training data. Through rigorous testing in multiple simulated and real-world settings, our system demonstrates a marked improvement in performance over conventional robotic systems, showcasing a significant breakthrough in the field of robotics.
comment: Accepted by IROS 2025
☆ Procedimiento de auditoría de ciberseguridad para sistemas autónomos: metodología, amenazas y mitigaciones SC
The deployment of autonomous systems has experienced remarkable growth in recent years, driven by their integration into sectors such as industry, medicine, logistics, and domestic environments. This expansion is accompanied by a series of security issues that entail significant risks due to the critical nature of autonomous systems, especially those operating in human-interaction environments. Furthermore, technological advancement and the high operational and architectural complexity of autonomous systems have resulted in an increased attack surface. This article presents a specific security auditing procedure for autonomous systems, based on a layer-structured methodology, a threat taxonomy adapted to the robotic context, and a set of concrete mitigation measures. The validity of the proposed approach is demonstrated through four practical case studies applied to representative robotic platforms: the Vision 60 military quadruped from Ghost Robotics, the A1 robot from Unitree Robotics, the UR3 collaborative arm from Universal Robots, and the Pepper social robot from Aldebaran Robotics.
comment: 32 pages, in Spanish language, 7 tables, 12 Figures. White paper under the TESCAC project
☆ Follow-Me in Micro-Mobility with End-to-End Imitation Learning
Autonomous micro-mobility platforms face challenges from the perspective of the typical deployment environment: large indoor spaces or urban areas that are potentially crowded and highly dynamic. While social navigation algorithms have progressed significantly, optimizing user comfort and overall user experience over other typical metrics in robotics (e.g., time or distance traveled) is understudied. Specifically, these metrics are critical in commercial applications. In this paper, we show how imitation learning delivers smoother and overall better controllers, versus previously used manually-tuned controllers. We demonstrate how DAAV's autonomous wheelchair achieves state-of-the-art comfort in follow-me mode, in which it follows a human operator assisting persons with reduced mobility (PRM). This paper analyzes different neural network architectures for end-to-end control and demonstrates their usability in real-world production-level deployments.
☆ Decomposed Object Manipulation via Dual-Actor Policy
Object manipulation, which focuses on learning to perform tasks on similar parts across different types of objects, can be divided into an approaching stage and a manipulation stage. However, previous works often ignore this characteristic of the task and rely on a single policy to directly learn the whole process of object manipulation. To address this problem, we propose a novel Dual-Actor Policy, termed DAP, which explicitly considers different stages and leverages heterogeneous visual priors to enhance each stage. Specifically, we introduce an affordance-based actor to locate the functional part in the manipulation task, thereby improving the approaching process. Following this, we propose a motion flow-based actor to capture the movement of the component, facilitating the manipulation process. Finally, we introduce a decision maker to determine the current stage of DAP and select the corresponding actor. Moreover, existing object manipulation datasets contain few objects and lack the visual priors needed to support training. To address this, we construct a simulated dataset, the Dual-Prior Object Manipulation Dataset, which combines the two visual priors and includes seven tasks, including two challenging long-term, multi-stage tasks. Experimental results on our dataset, the RoboTwin benchmark and real-world scenarios illustrate that our method consistently outperforms the SOTA method by 5.55%, 14.7% and 10.4% on average respectively.
comment: 9 pages, 7 figures, 5 tables
☆ TAPOM: Task-Space Topology-Guided Motion Planning for Manipulating Elongated Object in Cluttered Environments
Robotic manipulation in complex, constrained spaces is vital for widespread applications but challenging, particularly when navigating narrow passages with elongated objects. Existing planning methods often fail in these low-clearance scenarios due to the sampling difficulties or the local minima. This work proposes Topology-Aware Planning for Object Manipulation (TAPOM), which explicitly incorporates task-space topological analysis to enable efficient planning. TAPOM uses a high-level analysis to identify critical pathways and generate guiding keyframes, which are utilized in a low-level planner to find feasible configuration space trajectories. Experimental validation demonstrates significantly high success rates and improved efficiency over state-of-the-art methods on low-clearance manipulation tasks. This approach offers broad implications for enhancing manipulation capabilities of robots in complex real-world environments.
☆ Epically Powerful: An open-source software and mechatronics infrastructure for wearable robotic systems
Epically Powerful is an open-source robotics infrastructure that streamlines the underlying framework of wearable robotic systems - managing communication protocols, clocking, actuator commands, visualization, sensor data acquisition, data logging, and more - while also providing comprehensive guides for hardware selection, system assembly, and controller implementation. Epically Powerful contains a code base enabling simplified user implementation via Python that seamlessly interfaces with various commercial state-of-the-art quasi-direct drive (QDD) actuators, single-board computers, and common sensors, provides example controllers, and enables real-time visualization. To further support device development, the package also includes a recommended parts list and compatibility guide and detailed documentation on hardware and software implementation. The goal of Epically Powerful is to lower the barrier to developing and deploying custom wearable robotic systems without a pre-specified form factor, enabling researchers to go from raw hardware to modular, robust devices quickly and effectively. Though originally designed with wearable robotics in mind, Epically Powerful is broadly applicable to other robotic domains that utilize QDD actuators, single-board computers, and sensors for closed-loop control.
comment: 11 pages, 5 figures. This work has been submitted to the IEEE for possible publication
☆ Tunable Passivity Control for Centralized Multiport Networked Systems
Centralized Multiport Networked Dynamic (CMND) systems have emerged as a key architecture with applications in several complex network systems, such as multilateral telerobotics and multi-agent control. These systems consist of a hub node/subsystem connecting with multiple remote nodes/subsystems via a networked architecture. One challenge for this system is stability, which can be affected by non-ideal network artifacts. Conventional passivity-based approaches can stabilize the system under specialized applications like small-scale networked systems. However, those conventional passive stabilizers have several restrictions, such as distributing compensation across subsystems in a decentralized manner, limiting flexibility, and, at the same time, relying on the restrictive assumptions of node passivity. This paper synthesizes a centralized optimal passivity-based stabilization framework for CMND systems. It consists of a centralized passivity observer monitoring overall energy flow and an optimal passivity controller that distributes the just-needed dissipation among various nodes, guaranteeing strict passivity and, thus, L2 stability. The proposed data-driven model-free approach, i.e., Tunable Centralized Optimal Passivity Control (TCoPC), optimizes total performance based on the prescribed dissipation distribution strategy while ensuring stability. The controller can put high dissipation loads on some sub-networks while relaxing the dissipation on other nodes. Simulation results demonstrate the proposed frameworks performance in a complex task under different time-varying delay scenarios while relaxing the remote nodes minimum phase and passivity assumption, enhancing the scalability and generalizability.
☆ MoE-DP: An MoE-Enhanced Diffusion Policy for Robust Long-Horizon Robotic Manipulation with Skill Decomposition and Failure Recovery
Diffusion policies have emerged as a powerful framework for robotic visuomotor control, yet they often lack the robustness to recover from subtask failures in long-horizon, multi-stage tasks and their learned representations of observations are often difficult to interpret. In this work, we propose the Mixture of Experts-Enhanced Diffusion Policy (MoE-DP), where the core idea is to insert a Mixture of Experts (MoE) layer between the visual encoder and the diffusion model. This layer decomposes the policy's knowledge into a set of specialized experts, which are dynamically activated to handle different phases of a task. We demonstrate through extensive experiments that MoE-DP exhibits a strong capability to recover from disturbances, significantly outperforming standard baselines in robustness. On a suite of 6 long-horizon simulation tasks, this leads to a 36% average relative improvement in success rate under disturbed conditions. This enhanced robustness is further validated in the real world, where MoE-DP also shows significant performance gains. We further show that MoE-DP learns an interpretable skill decomposition, where distinct experts correspond to semantic task primitives (e.g., approaching, grasping). This learned structure can be leveraged for inference-time control, allowing for the rearrangement of subtasks without any re-training.Our video and code are available at the https://moe-dp-website.github.io/MoE-DP-Website/.
☆ Multi-agent Coordination via Flow Matching
This work presents MAC-Flow, a simple yet expressive framework for multi-agent coordination. We argue that requirements of effective coordination are twofold: (i) a rich representation of the diverse joint behaviors present in offline data and (ii) the ability to act efficiently in real time. However, prior approaches often sacrifice one for the other, i.e., denoising diffusion-based solutions capture complex coordination but are computationally slow, while Gaussian policy-based solutions are fast but brittle in handling multi-agent interaction. MAC-Flow addresses this trade-off by first learning a flow-based representation of joint behaviors, and then distilling it into decentralized one-step policies that preserve coordination while enabling fast execution. Across four different benchmarks, including $12$ environments and $34$ datasets, MAC-Flow alleviates the trade-off between performance and computational cost, specifically achieving about $\boldsymbol{\times14.5}$ faster inference compared to diffusion-based MARL methods, while maintaining good performance. At the same time, its inference speed is similar to that of prior Gaussian policy-based offline multi-agent reinforcement learning (MARL) methods.
☆ Encoding Biomechanical Energy Margin into Passivity-based Synchronization for Networked Telerobotic Systems
Maintaining system stability and accurate position tracking is imperative in networked robotic systems, particularly for haptics-enabled human-robot interaction. Recent literature has integrated human biomechanics into the stabilizers implemented for teleoperation, enhancing force preservation while guaranteeing convergence and safety. However, position desynchronization due to imperfect communication and non-passive behaviors remains a challenge. This paper proposes a two-port biomechanics-aware passivity-based synchronizer and stabilizer, referred to as TBPS2. This stabilizer optimizes position synchronization by leveraging human biomechanics while reducing the stabilizer's conservatism in its activation. We provide the mathematical design synthesis of the stabilizer and the proof of stability. We also conducted a series of grid simulations and systematic experiments, comparing their performance with that of state-of-the-art solutions under varying time delays and environmental conditions.
☆ A semi-analytical approach for computing the largest singularity-free spheres of a class of 6-6 Stewart-Gough platforms for specified orientation workspaces
This article presents a method for computing the largest singularity-free sphere (SFS) of a 6-6 Stewart-Gough platform manipulator (SGPM) over a specified orientation workspace. For a fixed orientation of the moving platform, the SFS is computed analytically. This process is repeated over a set of samples generated within the orientation workspace, and the smallest among them is designated as the desired SFS for the given orientation workspace. Numerical experiments are performed on four distinct architectures of the SGPM to understand their relative performances w.r.t. SFS volumes over the same orientation workspace. This study demonstrates the potential utility of the proposed computational method both in analysis and design of SGPMs.
☆ iFlyBot-VLM Technical Report
We introduce iFlyBot-VLM, a general-purpose Vision-Language Model (VLM) used to improve the domain of Embodied Intelligence. The central objective of iFlyBot-VLM is to bridge the cross-modal semantic gap between high-dimensional environmental perception and low-level robotic motion control. To this end, the model abstracts complex visual and spatial information into a body-agnostic and transferable Operational Language, thereby enabling seamless perception-action closed-loop coordination across diverse robotic platforms. The architecture of iFlyBot-VLM is systematically designed to realize four key functional capabilities essential for embodied intelligence: 1) Spatial Understanding and Metric Reasoning; 2) Interactive Target Grounding; 3) Action Abstraction and Control Parameter Generation; 4) Task Planning and Skill Sequencing. We envision iFlyBot-VLM as a scalable and generalizable foundation model for embodied AI, facilitating the progression from specialized task-oriented systems toward generalist, cognitively capable agents. We conducted evaluations on 10 current mainstream embodied intelligence-related VLM benchmark datasets, such as Blink and Where2Place, and achieved optimal performance while preserving the model's general capabilities. We will publicly release both the training data and model weights to foster further research and development in the field of Embodied Intelligence.
♻ ☆ Periodic Skill Discovery NeurIPS 2025
Unsupervised skill discovery in reinforcement learning (RL) aims to learn diverse behaviors without relying on external rewards. However, current methods often overlook the periodic nature of learned skills, focusing instead on increasing the mutual dependence between states and skills or maximizing the distance traveled in latent space. Considering that many robotic tasks - particularly those involving locomotion - require periodic behaviors across varying timescales, the ability to discover diverse periodic skills is essential. Motivated by this, we propose Periodic Skill Discovery (PSD), a framework that discovers periodic behaviors in an unsupervised manner. The key idea of PSD is to train an encoder that maps states to a circular latent space, thereby naturally encoding periodicity in the latent representation. By capturing temporal distance, PSD can effectively learn skills with diverse periods in complex robotic tasks, even with pixel-based observations. We further show that these learned skills achieve high performance on downstream tasks such as hurdling. Moreover, integrating PSD with an existing skill discovery method offers more diverse behaviors, thus broadening the agent's repertoire. Our code and demos are available at https://jonghaepark.github.io/psd/
comment: NeurIPS 2025
♻ ☆ Tactical Decision Making for Autonomous Trucks by Deep Reinforcement Learning with Total Cost of Operation Based Reward
We develop a deep reinforcement learning framework for tactical decision making in an autonomous truck, specifically for Adaptive Cruise Control (ACC) and lane change maneuvers in a highway scenario. Our results demonstrate that it is beneficial to separate high-level decision-making processes and low-level control actions between the reinforcement learning agent and the low-level controllers based on physical models. In the following, we study optimizing the performance with a realistic and multi-objective reward function based on Total Cost of Operation (TCOP) of the truck using different approaches; by adding weights to reward components, by normalizing the reward components and by using curriculum learning techniques.
comment: Paper is accepted for publication in Artificial Intelligence Review
♻ ☆ Ethics-Aware Safe Reinforcement Learning for Rare-Event Risk Control in Interactive Urban Driving
Autonomous vehicles hold great promise for reducing traffic fatalities and improving transportation efficiency, yet their widespread adoption hinges on embedding credible and transparent ethical reasoning into routine and emergency maneuvers, particularly to protect vulnerable road users (VRUs) such as pedestrians and cyclists. Here, we present a hierarchical Safe Reinforcement Learning (Safe RL) framework that augments standard driving objectives with ethics-aware cost signals. At the decision level, a Safe RL agent is trained using a composite ethical risk cost, combining collision probability and harm severity, to generate high-level motion targets. A dynamic, risk-sensitive Prioritized Experience Replay mechanism amplifies learning from rare but critical, high-risk events. At the execution level, polynomial path planning coupled with Proportional-Integral-Derivative (PID) and Stanley controllers translates these targets into smooth, feasible trajectories, ensuring both accuracy and comfort. We train and validate our approach on closed-loop simulation environments derived from large-scale, real-world traffic datasets encompassing diverse vehicles, cyclists, and pedestrians, and demonstrate that it outperforms baseline methods in reducing risk to others while maintaining ego performance and comfort. This work provides a reproducible benchmark for Safe RL with explicitly ethics-aware objectives in human-mixed traffic scenarios. Our results highlight the potential of combining formal control theory and data-driven learning to advance ethically accountable autonomy that explicitly protects those most at risk in urban traffic environments. Across two interactive benchmarks and five random seeds, our policy decreases conflict frequency by 25-45% compared to matched task successes while maintaining comfort metrics within 5%.
♻ ☆ Holistic Evaluation of Multimodal LLMs on Spatial Intelligence
Multimodal models have achieved remarkable progress in recent years. Nevertheless, they continue to exhibit notable limitations in spatial understanding and reasoning, the very capability that anchors artificial general intelligence in the physical world. With the recent release of GPT-5, allegedly the most powerful AI model to date, it is timely to examine where the leading models (GPT, Gemini, Grok, Seed, Qwen, and Intern) stand on the path toward spatial intelligence. We thus propose EASI for holistic Evaluation of multimodAl LLMs on Spatial Intelligence. EASI conceptualizes a comprehensive taxonomy of spatial tasks that unifies existing benchmarks and a standardized protocol for the fair evaluation of state-of-the-art proprietary and open-source models. In this report, we conduct the study across eight key benchmarks, at a cost exceeding ten billion total tokens. Our empirical study then reveals that (1) GPT-5 demonstrates unprecedented strength in spatial intelligence (SI), yet (2) still falls short of human performance significantly across a broad spectrum of SI-tasks. Moreover, we (3) show that SI-tasks expose greater model capability deficiency than non-SI tasks, to the extent that (4) proprietary models do not exhibit a decisive advantage when facing the most difficult ones. In addition, we conduct a qualitative evaluation across a diverse set of scenarios that are intuitive for humans, yet fail even the most advanced multimodal models.
comment: Codebase: https://github.com/EvolvingLMMs-Lab/EASI/
♻ ☆ Pogobot: an Open-Source, Low-Cost Robot for Swarm Robotics and Programmable Active Matter
This paper describes the Pogobot, an open-source platform specifically designed for research at the interface of swarm robotics and active matter. Pogobot features vibration-based or wheel-based locomotion, fast infrared communication, and an array of sensors in a cost-effective package (approx. 250euros/unit). The platform's modular design, comprehensive API, and extensible architecture facilitate the implementation of swarm intelligence algorithms and collective motion. Pogobots offer an accessible alternative to existing platforms while providing advanced capabilities including directional communication between units and fast locomotion, all with a compact form factor. More than 200 Pogobots are already being used on a daily basis in several Universities to study self-organizing systems, programmable active matter, discrete reaction-diffusion-advection systems and computational models of social learning and evolution. This paper details the hardware and software architecture, communication protocols, locomotion mechanisms, and the infrastructure built around the Pogobots.
♻ ☆ Mean-Shift Theory and Its Applications in Swarm Robotics: A New Way to Enhance the Efficiency of Multi-Robot Collaboration
Swarms evolving from collective behaviors among multiple individuals are commonly seen in nature, which enables biological systems to exhibit more efficient and robust collaboration. Creating similar swarm intelligence in engineered robots poses challenges to the design of collaborative algorithms that can be programmed at large scales. The assignment-based method has played an eminent role for a very long time in solving collaboration problems of robot swarms. However, it faces fundamental limitations in terms of efficiency and robustness due to its unscalability to swarm variants. This article presents a tutorial review on recent advances in assignment-free collaboration of robot swarms, focusing on the problem of shape formation. A key theoretical component is the recently developed \emph{mean-shift exploration} strategy, which improves the collaboration efficiency of large-scale swarms by dozens of times. Further, the efficiency improvement is more significant as the swarm scale increases. Finally, this article discusses three important applications of the mean-shift exploration strategy, including precise shape formation, area coverage formation, and maneuvering formation, as well as their corresponding industrial scenarios in smart warehousing, area exploration, and cargo transportation.
♻ ☆ Affordance-based Robot Manipulation with Flow Matching
We present a framework for assistive robot manipulation, which focuses on two fundamental challenges: first, efficiently adapting large-scale models to downstream scene affordance understanding tasks, especially in daily living scenarios where gathering multi-task data involving humans requires strenuous effort; second, effectively learning robot action trajectories by grounding the visual affordance model. We tackle the first challenge by employing a parameter-efficient prompt tuning method that prepends learnable text prompts to the frozen vision model to predict manipulation affordances in multi-task scenarios. Then we propose to learn robot action trajectories guided by affordances in a supervised flow matching method. Flow matching represents a robot visuomotor policy as a conditional process of flowing random waypoints to desired robot action trajectories. Finally, we introduce a real-world dataset with 10 tasks across Activities of Daily Living to test our framework. Our extensive evaluation highlights that the proposed prompt tuning method for learning manipulation affordance achieves competitive performance and even outperforms some other finetuning protocols across data scales, while satisfying parameter efficiency. Learning multi-task robot action trajectories with flow matching leads to consistently favorable results in several robot manipulation benchmarks than some alternative behavior cloning methods. This includes more stable training and evaluation, and noticeably faster inference, while maintaining comparable generalization performance to diffusion policy, where flow matching performs marginally better in most cases. Our framework seamlessly unifies affordance learning and action generation with flow matching for robot manipulation.
♻ ☆ Learning to Navigate Socially Through Proactive Risk Perception
In this report, we describe the technical details of our submission to the IROS 2025 RoboSense Challenge Social Navigation Track. This track focuses on developing RGBD-based perception and navigation systems that enable autonomous agents to navigate safely, efficiently, and socially compliantly in dynamic human-populated indoor environments. The challenge requires agents to operate from an egocentric perspective using only onboard sensors including RGB-D observations and odometry, without access to global maps or privileged information, while maintaining social norm compliance such as safe distances and collision avoidance. Building upon the Falcon model, we introduce a Proactive Risk Perception Module to enhance social navigation performance. Our approach augments Falcon with collision risk understanding that learns to predict distance-based collision risk scores for surrounding humans, which enables the agent to develop more robust spatial awareness and proactive collision avoidance behaviors. The evaluation on the Social-HM3D benchmark demonstrates that our method improves the agent's ability to maintain personal space compliance while navigating toward goals in crowded indoor scenes with dynamic human agents, achieving 2nd place among 16 participating teams in the challenge.
♻ ☆ GeoAware-VLA: Implicit Geometry Aware Vision-Language-Action Model
Vision-Language-Action (VLA) models often fail to generalize to novel camera viewpoints, a limitation stemming from their difficulty in inferring robust 3D geometry from 2D images. We introduce GeoAware-VLA, a simple yet effective approach that enhances viewpoint invariance by integrating strong geometric priors into the vision backbone. Instead of training a visual encoder or relying on explicit 3D data, we leverage a frozen, pretrained geometric vision model as a feature extractor. A trainable projection layer then adapts these geometrically-rich features for the policy decoder, relieving it of the burden of learning 3D consistency from scratch. Through extensive evaluations on LIBERO benchmark subsets, we show GeoAware-VLA achieves substantial improvements in zero-shot generalization to novel camera poses, boosting success rates by over 2x in simulation. Crucially, these benefits translate to the physical world; our model shows a significant performance gain on a real robot, especially when evaluated from unseen camera angles. Our approach proves effective across both continuous and discrete action spaces, highlighting that robust geometric grounding is a key component for creating more generalizable robotic agents.
comment: Under Review, Project Page https://alisharey.github.io/GeoAware-VLA/
♻ ☆ Search-TTA: A Multimodal Test-Time Adaptation Framework for Visual Search in the Wild
To perform outdoor visual navigation and search, a robot may leverage satellite imagery to generate visual priors. This can help inform high-level search strategies, even when such images lack sufficient resolution for target recognition. However, many existing informative path planning or search-based approaches either assume no prior information, or use priors without accounting for how they were obtained. Recent work instead utilizes large Vision Language Models (VLMs) for generalizable priors, but their outputs can be inaccurate due to hallucination, leading to inefficient search. To address these challenges, we introduce Search-TTA, a multimodal test-time adaptation framework with a flexible plug-and-play interface compatible with various input modalities (e.g., image, text, sound) and planning methods (e.g., RL-based). First, we pretrain a satellite image encoder to align with CLIP's visual encoder to output probability distributions of target presence used for visual search. Second, our TTA framework dynamically refines CLIP's predictions during search using uncertainty-weighted gradient updates inspired by Spatial Poisson Point Processes. To train and evaluate Search-TTA, we curate AVS-Bench, a visual search dataset based on internet-scale ecological data containing 380k images and taxonomy data. We find that Search-TTA improves planner performance by up to 30.0%, particularly in cases with poor initial CLIP predictions due to domain mismatch and limited training data. It also performs comparably with significantly larger VLMs, and achieves zero-shot generalization via emergent alignment to unseen modalities. Finally, we deploy Search-TTA on a real UAV via hardware-in-the-loop testing, by simulating its operation within a large-scale simulation that provides onboard sensing.
comment: Accepted for presentation at CORL 2025. Code, models, and data are available at https://search-tta.github.io/
♻ ☆ Octopus-like Reaching Motion: A Perspective Inspired by Whipping
The stereotypical reaching motion of the octopus arm has drawn growing attention for its efficient control of a highly deformable body. Previous studies suggest that its characteristic bend propagation may share underlying principles with the dynamics of a whip. This work investigates whether whip-like passive dynamics in water can reproduce the kinematic features observed in biological reaching and their similarities and differences. Platform-based whipping tests were performed in water and air while systematically varying material stiffness and driving speed. Image-based quantification revealed that the Ecoflex Gel 2 arm driven at 150 rpm (motor speed) reproduced curvature propagation similar to that observed in octopus reaching. However, its bend-point velocity decreased monotonically rather than exhibiting the biological bell-shaped profile, confirming that the octopus reaching movement is not merely a passive whipping behavior. The absence of propagation in air further highlights the critical role of the surrounding medium in forming octopus-like reaching motion. This study provides a new perspective for understand biological reaching movement, and offers a potential platform for future hydrodynamic research.
comment: The first two listed authors contributed equally. Yiyuan Zhang is the corresponding author
♻ ☆ ReNiL: Event-Driven Pedestrian Bayesian Localization Using IMU for Real-World Applications
Pedestrian inertial localization is key for mobile and IoT services because it provides infrastructure-free positioning. Yet most learning-based methods depend on fixed sliding-window integration, struggle to adapt to diverse motion scales and cadences, and yield inconsistent uncertainty, limiting real-world use. We present ReNiL, a Bayesian deep-learning framework for accurate, efficient, and uncertainty-aware pedestrian localization. ReNiL introduces Inertial Positioning Demand Points (IPDPs) to estimate motion at contextually meaningful waypoints instead of dense tracking, and supports inference on IMU sequences at any scale so cadence can match application needs. It couples a motion-aware orientation filter with an Any-Scale Laplace Estimator (ASLE), a dual-task network that blends patch-based self-supervision with Bayesian regression. By modeling displacements with a Laplace distribution, ReNiL provides homogeneous Euclidean uncertainty that integrates cleanly with other sensors. A Bayesian inference chain links successive IPDPs into consistent trajectories. On RoNIN-ds and a new WUDataset covering indoor and outdoor motion from 28 participants, ReNiL achieves state-of-the-art displacement accuracy and uncertainty consistency, outperforming TLIO, CTIN, iMoT, and RoNIN variants while reducing computation. Application studies further show robustness and practicality for mobile and IoT localization, making ReNiL a scalable, uncertainty-aware foundation for next-generation positioning.
comment: This work has been submitted to the ACM for possible publication
♻ ☆ Generalizing Robot Trajectories from Single-Context Human Demonstrations: A Probabilistic Approach
Generalizing robot trajectories from human demonstrations to new contexts remains a key challenge in Learning from Demonstration (LfD), particularly when only single-context demonstrations are available. We present a novel Gaussian Mixture Model (GMM)-based approach that enables systematic generalization from single-context demonstrations to a wide range of unseen start and goal configurations. Our method performs component-level reparameterization of the GMM, adapting both mean vectors and covariance matrices, followed by Gaussian Mixture Regression (GMR) to generate smooth trajectories. We evaluate the approach on a dual-arm pick-and-place task with varying box placements, comparing against several baselines. Results show that our method significantly outperforms baselines in trajectory success and fidelity, maintaining accuracy even under combined translational and rotational variations of task configurations. These results demonstrate that our method generalizes effectively while ensuring boundary convergence and preserving the intrinsic structure of demonstrated motions.
Robotics 45
☆ GentleHumanoid: Learning Upper-body Compliance for Contact-rich Human and Object Interaction
Humanoid robots are expected to operate in human-centered environments where safe and natural physical interaction is essential. However, most recent reinforcement learning (RL) policies emphasize rigid tracking and suppress external forces. Existing impedance-augmented approaches are typically restricted to base or end-effector control and focus on resisting extreme forces rather than enabling compliance. We introduce GentleHumanoid, a framework that integrates impedance control into a whole-body motion tracking policy to achieve upper-body compliance. At its core is a unified spring-based formulation that models both resistive contacts (restoring forces when pressing against surfaces) and guiding contacts (pushes or pulls sampled from human motion data). This formulation ensures kinematically consistent forces across the shoulder, elbow, and wrist, while exposing the policy to diverse interaction scenarios. Safety is further supported through task-adjustable force thresholds. We evaluate our approach in both simulation and on the Unitree G1 humanoid across tasks requiring different levels of compliance, including gentle hugging, sit-to-stand assistance, and safe object manipulation. Compared to baselines, our policy consistently reduces peak contact forces while maintaining task success, resulting in smoother and more natural interactions. These results highlight a step toward humanoid robots that can safely and effectively collaborate with humans and handle objects in real-world environments.
comment: Home page: https://gentle-humanoid.axell.top
☆ X-Diffusion: Training Diffusion Policies on Cross-Embodiment Human Demonstrations
Human videos can be recorded quickly and at scale, making them an appealing source of training data for robot learning. However, humans and robots differ fundamentally in embodiment, resulting in mismatched action execution. Direct kinematic retargeting of human hand motion can therefore produce actions that are physically infeasible for robots. Despite these low-level differences, human demonstrations provide valuable motion cues about how to manipulate and interact with objects. Our key idea is to exploit the forward diffusion process: as noise is added to actions, low-level execution differences fade while high-level task guidance is preserved. We present X-Diffusion, a principled framework for training diffusion policies that maximally leverages human data without learning dynamically infeasible motions. X-Diffusion first trains a classifier to predict whether a noisy action is executed by a human or robot. Then, a human action is incorporated into policy training only after adding sufficient noise such that the classifier cannot discern its embodiment. Actions consistent with robot execution supervise fine-grained denoising at low noise levels, while mismatched human actions provide only coarse guidance at higher noise levels. Our experiments show that naive co-training under execution mismatches degrades policy performance, while X-Diffusion consistently improves it. Across five manipulation tasks, X-Diffusion achieves a 16% higher average success rate than the best baseline. The project website is available at https://portal-cornell.github.io/X-Diffusion/.
☆ Real-to-Sim Robot Policy Evaluation with Gaussian Splatting Simulation of Soft-Body Interactions
Robotic manipulation policies are advancing rapidly, but their direct evaluation in the real world remains costly, time-consuming, and difficult to reproduce, particularly for tasks involving deformable objects. Simulation provides a scalable and systematic alternative, yet existing simulators often fail to capture the coupled visual and physical complexity of soft-body interactions. We present a real-to-sim policy evaluation framework that constructs soft-body digital twins from real-world videos and renders robots, objects, and environments with photorealistic fidelity using 3D Gaussian Splatting. We validate our approach on representative deformable manipulation tasks, including plush toy packing, rope routing, and T-block pushing, demonstrating that simulated rollouts correlate strongly with real-world execution performance and reveal key behavioral patterns of learned policies. Our results suggest that combining physics-informed reconstruction with high-quality rendering enables reproducible, scalable, and accurate evaluation of robotic manipulation policies. Website: https://real2sim-eval.github.io/
comment: Website: https://real2sim-eval.github.io/
☆ SAFe-Copilot: Unified Shared Autonomy Framework
Autonomous driving systems remain brittle in rare, ambiguous, and out-of-distribution scenarios, where human driver succeed through contextual reasoning. Shared autonomy has emerged as a promising approach to mitigate such failures by incorporating human input when autonomy is uncertain. However, most existing methods restrict arbitration to low-level trajectories, which represent only geometric paths and therefore fail to preserve the underlying driving intent. We propose a unified shared autonomy framework that integrates human input and autonomous planners at a higher level of abstraction. Our method leverages Vision Language Models (VLMs) to infer driver intent from multi-modal cues -- such as driver actions and environmental context -- and to synthesize coherent strategies that mediate between human and autonomous control. We first study the framework in a mock-human setting, where it achieves perfect recall alongside high accuracy and precision. A human-subject survey further shows strong alignment, with participants agreeing with arbitration outcomes in 92% of cases. Finally, evaluation on the Bench2Drive benchmark demonstrates a substantial reduction in collision rate and improvement in overall performance compared to pure autonomy. Arbitration at the level of semantic, language-based representations emerges as a design principle for shared autonomy, enabling systems to exercise common-sense reasoning and maintain continuity with human intent.
☆ Evo-1: Lightweight Vision-Language-Action Model with Preserved Semantic Alignment
Vision-Language-Action (VLA) models have emerged as a powerful framework that unifies perception, language, and control, enabling robots to perform diverse tasks through multimodal understanding. However, current VLA models typically contain massive parameters and rely heavily on large-scale robot data pretraining, leading to high computational costs during training, as well as limited deployability for real-time inference. Moreover, most training paradigms often degrade the perceptual representations of the vision-language backbone, resulting in overfitting and poor generalization to downstream tasks. In this work, we present Evo-1, a lightweight VLA model that reduces computation and improves deployment efficiency, while maintaining strong performance without pretraining on robot data. Evo-1 builds on a native multimodal Vision-Language model (VLM), incorporating a novel cross-modulated diffusion transformer along with an optimized integration module, together forming an effective architecture. We further introduce a two-stage training paradigm that progressively aligns action with perception, preserving the representations of the VLM. Notably, with only 0.77 billion parameters, Evo-1 achieves state-of-the-art results on the Meta-World and RoboTwin suite, surpassing the previous best models by 12.4% and 6.9%, respectively, and also attains a competitive result of 94.8% on LIBERO. In real-world evaluations, Evo-1 attains a 78% success rate with high inference frequency and low memory overhead, outperforming all baseline methods. We release code, data, and model weights to facilitate future research on lightweight and efficient VLA models.
comment: Github: https://github.com/MINT-SJTU/Evo-1
☆ Temporal Action Selection for Action Chunking
Action chunking is a widely adopted approach in Learning from Demonstration (LfD). By modeling multi-step action chunks rather than single-step actions, action chunking significantly enhances modeling capabilities for human expert policies. However, the reduced decision frequency restricts the utilization of recent observations, degrading reactivity - particularly evident in the inadequate adaptation to sensor noise and dynamic environmental changes. Existing efforts to address this issue have primarily resorted to trading off reactivity against decision consistency, without achieving both. To address this limitation, we propose a novel algorithm, Temporal Action Selector (TAS), which caches predicted action chunks from multiple timesteps and dynamically selects the optimal action through a lightweight selector network. TAS achieves balanced optimization across three critical dimensions: reactivity, decision consistency, and motion coherence. Experiments across multiple tasks with diverse base policies show that TAS significantly improves success rates - yielding an absolute gain of up to 73.3%. Furthermore, integrating TAS as a base policy with residual reinforcement learning (RL) substantially enhances training efficiency and elevates the performance plateau. Experiments in both simulation and physical robots confirm the method's efficacy.
☆ BoRe-Depth: Self-supervised Monocular Depth Estimation with Boundary Refinement for Embedded Systems IROS 2025
Depth estimation is one of the key technologies for realizing 3D perception in unmanned systems. Monocular depth estimation has been widely researched because of its low-cost advantage, but the existing methods face the challenges of poor depth estimation performance and blurred object boundaries on embedded systems. In this paper, we propose a novel monocular depth estimation model, BoRe-Depth, which contains only 8.7M parameters. It can accurately estimate depth maps on embedded systems and significantly improves boundary quality. Firstly, we design an Enhanced Feature Adaptive Fusion Module (EFAF) which adaptively fuses depth features to enhance boundary detail representation. Secondly, we integrate semantic knowledge into the encoder to improve the object recognition and boundary perception capabilities. Finally, BoRe-Depth is deployed on NVIDIA Jetson Orin, and runs efficiently at 50.7 FPS. We demonstrate that the proposed model significantly outperforms previous lightweight models on multiple challenging datasets, and we provide detailed ablation studies for the proposed methods. The code is available at https://github.com/liangxiansheng093/BoRe-Depth.
comment: 8 pages, 5 figures, published to IROS 2025
☆ ForeRobo: Unlocking Infinite Simulation Data for 3D Goal-driven Robotic Manipulation
Efficiently leveraging simulation to acquire advanced manipulation skills is both challenging and highly significant. We introduce \textit{ForeRobo}, a generative robotic agent that utilizes generative simulations to autonomously acquire manipulation skills driven by envisioned goal states. Instead of directly learning low-level policies, we advocate integrating generative paradigms with classical control. Our approach equips a robotic agent with a self-guided \textit{propose-generate-learn-actuate} cycle. The agent first proposes the skills to be acquired and constructs the corresponding simulation environments; it then configures objects into appropriate arrangements to generate skill-consistent goal states (\textit{ForeGen}). Subsequently, the virtually infinite data produced by ForeGen are used to train the proposed state generation model (\textit{ForeFormer}), which establishes point-wise correspondences by predicting the 3D goal position of every point in the current state, based on the scene state and task instructions. Finally, classical control algorithms are employed to drive the robot in real-world environments to execute actions based on the envisioned goal states. Compared with end-to-end policy learning methods, ForeFormer offers superior interpretability and execution efficiency. We train and benchmark ForeFormer across a variety of rigid-body and articulated-object manipulation tasks, and observe an average improvement of 56.32\% over the state-of-the-art state generation models, demonstrating strong generality across different manipulation patterns. Moreover, in real-world evaluations involving more than 20 robotic tasks, ForeRobo achieves zero-shot sim-to-real transfer and exhibits remarkable generalization capabilities, attaining an average success rate of 79.28\%.
☆ Studying the Effect of Explicit Interaction Representations on Learning Scene-level Distributions of Human Trajectories
Effectively capturing the joint distribution of all agents in a scene is relevant for predicting the true evolution of the scene and in turn providing more accurate information to the decision processes of autonomous vehicles. While new models have been developed for this purpose in recent years, it remains unclear how to best represent the joint distributions particularly from the perspective of the interactions between agents. Thus far there is no clear consensus on how best to represent interactions between agents; whether they should be learned implicitly from data by neural networks, or explicitly modeled using the spatial and temporal relations that are more grounded in human decision-making. This paper aims to study various means of describing interactions within the same network structure and their effect on the final learned joint distributions. Our findings show that more often than not, simply allowing a network to establish interactive connections between agents based on data has a detrimental effect on performance. Instead, having well defined interactions (such as which agent of an agent pair passes first at an intersection) can often bring about a clear boost in performance.
☆ GraSP-VLA: Graph-based Symbolic Action Representation for Long-Horizon Planning with VLA Policies
Deploying autonomous robots that can learn new skills from demonstrations is an important challenge of modern robotics. Existing solutions often apply end-to-end imitation learning with Vision-Language Action (VLA) models or symbolic approaches with Action Model Learning (AML). On the one hand, current VLA models are limited by the lack of high-level symbolic planning, which hinders their abilities in long-horizon tasks. On the other hand, symbolic approaches in AML lack generalization and scalability perspectives. In this paper we present a new neuro-symbolic approach, GraSP-VLA, a framework that uses a Continuous Scene Graph representation to generate a symbolic representation of human demonstrations. This representation is used to generate new planning domains during inference and serves as an orchestrator for low-level VLA policies, scaling up the number of actions that can be reproduced in a row. Our results show that GraSP-VLA is effective for modeling symbolic representations on the task of automatic planning domain generation from observations. In addition, results on real-world experiments show the potential of our Continuous Scene Graph representation to orchestrate low-level VLA policies in long-horizon tasks.
☆ MacroNav: Multi-Task Context Representation Learning Enables Efficient Navigation in Unknown Environments
Autonomous navigation in unknown environments requires compact yet expressive spatial understanding under partial observability to support high-level decision making. Existing approaches struggle to balance rich contextual representation with navigation efficiency. We present MacroNav, a learning-based navigation framework featuring two key components: (1) a lightweight context encoder trained via multi-task self-supervised learning to capture multi-scale, navigation-centric spatial representations; and (2) a reinforcement learning policy that seamlessly integrates these representations with graph-based reasoning for efficient action selection. Extensive experiments demonstrate the context encoder's efficient and robust environmental understanding. Real-world deployments further validate MacroNav's effectiveness, yielding significant gains over state-of-the-art navigation methods in both Success Rate (SR) and Success weighted by Path Length (SPL), while maintaining low computational cost. Code will be released upon acceptance.
☆ Design and Control of a Coaxial Dual-rotor Reconfigurable Tailsitter UAV Based on Swashplateless Mechanism
The tailsitter vertical takeoff and landing (VTOL) UAV is widely used due to its lower dead weight, which eliminates the actuators and mechanisms for tilting. However, the tailsitter UAV is susceptible to wind disturbances in multi-rotor mode, as it exposes a large frontal fuselage area. To address this issue, our tailsitter UAV features a reconfigurable wing design, allowing wings to retract in multi-rotor mode and extend in fixed- wing mode. Considering power efficiency, we design a coaxial heterogeneous dual-rotor configuration, which significantly re- duces the total power consumption. To reduce structural weight and simplify structural complexity, we employ a swashplateless mechanism with an improved design to control pitch and roll in multi-rotor mode. We optimize the structure of the swashplateless mechanism by adding flapping hinges, which reduces vibration during cyclic acceleration and deceleration. Finally, we perform comprehensive transition flight tests to validate stable flight performance across the entire flight envelope of the tailsitter UAV.
comment: 8 pages 12 figures
☆ Can Context Bridge the Reality Gap? Sim-to-Real Transfer of Context-Aware Policies
Sim-to-real transfer remains a major challenge in reinforcement learning (RL) for robotics, as policies trained in simulation often fail to generalize to the real world due to discrepancies in environment dynamics. Domain Randomization (DR) mitigates this issue by exposing the policy to a wide range of randomized dynamics during training, yet leading to a reduction in performance. While standard approaches typically train policies agnostic to these variations, we investigate whether sim-to-real transfer can be improved by conditioning the policy on an estimate of the dynamics parameters -- referred to as context. To this end, we integrate a context estimation module into a DR-based RL framework and systematically compare SOTA supervision strategies. We evaluate the resulting context-aware policies in both a canonical control benchmark and a real-world pushing task using a Franka Emika Panda robot. Results show that context-aware policies outperform the context-agnostic baseline across all settings, although the best supervision strategy depends on the task.
☆ GraspView: Active Perception Scoring and Best-View Optimization for Robotic Grasping in Cluttered Environments
Robotic grasping is a fundamental capability for autonomous manipulation, yet remains highly challenging in cluttered environments where occlusion, poor perception quality, and inconsistent 3D reconstructions often lead to unstable or failed grasps. Conventional pipelines have widely relied on RGB-D cameras to provide geometric information, which fail on transparent or glossy objects and degrade at close range. We present GraspView, an RGB-only robotic grasping pipeline that achieves accurate manipulation in cluttered environments without depth sensors. Our framework integrates three key components: (i) global perception scene reconstruction, which provides locally consistent, up-to-scale geometry from a single RGB view and fuses multi-view projections into a coherent global 3D scene; (ii) a render-and-score active perception strategy, which dynamically selects next-best-views to reveal occluded regions; and (iii) an online metric alignment module that calibrates VGGT predictions against robot kinematics to ensure physical scale consistency. Building on these tailor-designed modules, GraspView performs best-view global grasping, fusing multi-view reconstructions and leveraging GraspNet for robust execution. Experiments on diverse tabletop objects demonstrate that GraspView significantly outperforms both RGB-D and single-view RGB baselines, especially under heavy occlusion, near-field sensing, and with transparent objects. These results highlight GraspView as a practical and versatile alternative to RGB-D pipelines, enabling reliable grasping in unstructured real-world environments.
☆ PUL-SLAM: Path-Uncertainty Co-Optimization with Lightweight Stagnation Detection for Efficient Robotic Exploration
Existing Active SLAM methodologies face issues such as slow exploration speed and suboptimal paths. To address these limitations, we propose a hybrid framework combining a Path-Uncertainty Co-Optimization Deep Reinforcement Learning framework and a Lightweight Stagnation Detection mechanism. The Path-Uncertainty Co-Optimization framework jointly optimizes travel distance and map uncertainty through a dual-objective reward function, balancing exploration and exploitation. The Lightweight Stagnation Detection reduces redundant exploration through Lidar Static Anomaly Detection and Map Update Stagnation Detection, terminating episodes on low expansion rates. Experimental results show that compared with the frontier-based method and RRT method, our approach shortens exploration time by up to 65% and reduces path distance by up to 42%, significantly improving exploration efficiency in complex environments while maintaining reliable map completeness. Ablation studies confirm that the collaborative mechanism accelerates training convergence. Empirical validation on a physical robotic platform demonstrates the algorithm's practical applicability and its successful transferability from simulation to real-world environments.
☆ BFM-Zero: A Promptable Behavioral Foundation Model for Humanoid Control Using Unsupervised Reinforcement Learning
Building Behavioral Foundation Models (BFMs) for humanoid robots has the potential to unify diverse control tasks under a single, promptable generalist policy. However, existing approaches are either exclusively deployed on simulated humanoid characters, or specialized to specific tasks such as tracking. We propose BFM-Zero, a framework that learns an effective shared latent representation that embeds motions, goals, and rewards into a common space, enabling a single policy to be prompted for multiple downstream tasks without retraining. This well-structured latent space in BFM-Zero enables versatile and robust whole-body skills on a Unitree G1 humanoid in the real world, via diverse inference methods, including zero-shot motion tracking, goal reaching, and reward optimization, and few-shot optimization-based adaptation. Unlike prior on-policy reinforcement learning (RL) frameworks, BFM-Zero builds upon recent advancements in unsupervised RL and Forward-Backward (FB) models, which offer an objective-centric, explainable, and smooth latent representation of whole-body motions. We further extend BFM-Zero with critical reward shaping, domain randomization, and history-dependent asymmetric learning to bridge the sim-to-real gap. Those key design choices are quantitatively ablated in simulation. A first-of-its-kind model, BFM-Zero establishes a step toward scalable, promptable behavioral foundation models for whole-body humanoid control.
☆ CBMC-V3: A CNS-inspired Control Framework Towards Manipulation Agility with SNN
As robotic arm applications extend beyond industrial settings into healthcare, service, and daily life, existing control algorithms struggle to achieve the agile manipulation required for complex environments with dynamic trajectories, unpredictable interactions, and diverse objects. This paper presents a biomimetic control framework based on Spiking Neural Networks (SNN), inspired by the human Central Nervous System (CNS), to achieve agile control in such environments. The proposed framework features five control modules (cerebral cortex, cerebellum, thalamus, brainstem, spinal cord), three hierarchical control levels (first-order, second-order, third-order), and two information pathways (ascending, descending). Each module is fully implemented using SNN. The spinal cord module uses spike encoding and Leaky Integrate-and-Fire (LIF) neurons for feedback control. The brainstem module employs a network of LIF and non-spiking LIF neurons to dynamically adjust spinal cord parameters via reinforcement learning. The thalamus module similarly adjusts the cerebellum's torque outputs. The cerebellum module uses a recurrent SNN to learn the robotic arm's dynamics through regression, providing feedforward gravity compensation torques. The framework is validated both in simulation and on real-world robotic arm platform under various loads and trajectories. Results demonstrate that our method outperforms the industrial-grade position control in manipulation agility.
☆ Enhancing Fault-Tolerant Space Computing: Guidance Navigation and Control (GNC) and Landing Vision System (LVS) Implementations on Next-Gen Multi-Core Processors
Future planetary exploration missions demand high-performance, fault-tolerant computing to enable autonomous Guidance, Navigation, and Control (GNC) and Lander Vision System (LVS) operations during Entry, Descent, and Landing (EDL). This paper evaluates the deployment of GNC and LVS algorithms on next-generation multi-core processors--HPSC, Snapdragon VOXL2, and AMD Xilinx Versal--demonstrating up to 15x speedup for LVS image processing and over 250x speedup for Guidance for Fuel-Optimal Large Divert (GFOLD) trajectory optimization compared to legacy spaceflight hardware. To ensure computational reliability, we present ARBITER (Asynchronous Redundant Behavior Inspection for Trusted Execution and Recovery), a Multi-Core Voting (MV) mechanism that performs real-time fault detection and correction across redundant cores. ARBITER is validated in both static optimization tasks (GFOLD) and dynamic closed-loop control (Attitude Control System). A fault injection study further identifies the gradient computation stage in GFOLD as the most sensitive to bit-level errors, motivating selective protection strategies and vector-based output arbitration. This work establishes a scalable and energy-efficient architecture for future missions, including Mars Sample Return, Enceladus Orbilander, and Ceres Sample Return, where onboard autonomy, low latency, and fault resilience are critical.
☆ An LLM-based Framework for Human-Swarm Teaming Cognition in Disaster Search and Rescue
Large-scale disaster Search And Rescue (SAR) operations are persistently challenged by complex terrain and disrupted communications. While Unmanned Aerial Vehicle (UAV) swarms offer a promising solution for tasks like wide-area search and supply delivery, yet their effective coordination places a significant cognitive burden on human operators. The core human-machine collaboration bottleneck lies in the ``intention-to-action gap'', which is an error-prone process of translating a high-level rescue objective into a low-level swarm command under high intensity and pressure. To bridge this gap, this study proposes a novel LLM-CRF system that leverages Large Language Models (LLMs) to model and augment human-swarm teaming cognition. The proposed framework initially captures the operator's intention through natural and multi-modal interactions with the device via voice or graphical annotations. It then employs the LLM as a cognitive engine to perform intention comprehension, hierarchical task decomposition, and mission planning for the UAV swarm. This closed-loop framework enables the swarm to act as a proactive partner, providing active feedback in real-time while reducing the need for manual monitoring and control, which considerably advances the efficacy of the SAR task. We evaluate the proposed framework in a simulated SAR scenario. Experimental results demonstrate that, compared to traditional order and command-based interfaces, the proposed LLM-driven approach reduced task completion time by approximately $64.2\%$ and improved task success rate by $7\%$. It also leads to a considerable reduction in subjective cognitive workload, with NASA-TLX scores dropping by $42.9\%$. This work establishes the potential of LLMs to create more intuitive and effective human-swarm collaborations in high-stakes scenarios.
☆ Integrating Ergonomics and Manipulability for Upper Limb Postural Optimization in Bimanual Human-Robot Collaboration IROS 2025
This paper introduces an upper limb postural optimization method for enhancing physical ergonomics and force manipulability during bimanual human-robot co-carrying tasks. Existing research typically emphasizes human safety or manipulative efficiency, whereas our proposed method uniquely integrates both aspects to strengthen collaboration across diverse conditions (e.g., different grasping postures of humans, and different shapes of objects). Specifically, the joint angles of a simplified human skeleton model are optimized by minimizing the cost function to prioritize safety and manipulative capability. To guide humans towards the optimized posture, the reference end-effector poses of the robot are generated through a transformation module. A bimanual model predictive impedance controller (MPIC) is proposed for our human-like robot, CURI, to recalibrate the end effector poses through planned trajectories. The proposed method has been validated through various subjects and objects during human-human collaboration (HHC) and human-robot collaboration (HRC). The experimental results demonstrate significant improvement in muscle conditions by comparing the activation of target muscles before and after optimization.
comment: 7 pages, 7 figures, IROS 2025 accepted
☆ Learning Vision-Driven Reactive Soccer Skills for Humanoid Robots
Humanoid soccer poses a representative challenge for embodied intelligence, requiring robots to operate within a tightly coupled perception-action loop. However, existing systems typically rely on decoupled modules, resulting in delayed responses and incoherent behaviors in dynamic environments, while real-world perceptual limitations further exacerbate these issues. In this work, we present a unified reinforcement learning-based controller that enables humanoid robots to acquire reactive soccer skills through the direct integration of visual perception and motion control. Our approach extends Adversarial Motion Priors to perceptual settings in real-world dynamic environments, bridging motion imitation and visually grounded dynamic control. We introduce an encoder-decoder architecture combined with a virtual perception system that models real-world visual characteristics, allowing the policy to recover privileged states from imperfect observations and establish active coordination between perception and action. The resulting controller demonstrates strong reactivity, consistently executing coherent and robust soccer behaviors across various scenarios, including real RoboCup matches.
comment: Project page: https://humanoid-kick.github.io
☆ Dynamic Shape Control of Soft Robots Enabled by Data-Driven Model Reduction
Soft robots have shown immense promise in settings where they can leverage dynamic control of their entire bodies. However, effective dynamic shape control requires a controller that accounts for the robot's high-dimensional dynamics--a challenge exacerbated by a lack of general-purpose tools for modeling soft robots amenably for control. In this work, we conduct a comparative study of data-driven model reduction techniques for generating linear models amendable to dynamic shape control. We focus on three methods--the eigensystem realization algorithm, dynamic mode decomposition with control, and the Lagrangian operator inference (LOpInf) method. Using each class of model, we explored their efficacy in model predictive control policies for the dynamic shape control of a simulated eel-inspired soft robot in three experiments: 1) tracking simulated reference trajectories guaranteed to be feasible, 2) tracking reference trajectories generated from a biological model of eel kinematics, and 3) tracking reference trajectories generated by a reduced-scale physical analog. In all experiments, the LOpInf-based policies generated lower tracking errors than policies based on other models.
comment: 20 Pages, 8 Figures
☆ Design Exploration for Protection and Cleaning of Solar Panels with Case Studies for Space Missions
Solar energy is used for many mission-critical applications including space exploration, sensor systems to monitor wildfires, etc. Their operation can be limited or even terminated if solar panels are covered with dust or hit by space debris. To address this issue, we designed panel cleaning mechanisms and tested protective materials. For cleaning mechanisms, we designed and compared a wiper system and a rail system. For protective materials, we found through collision tests that polycarbonate was very promising, though the most important factor was layering a soft material between the panel's surface and a hard material. In the cleaning system comparisons, the wiper-based system was more efficient than the rail-based system in terms of cost, cleaning speed, and total power consumption.
comment: 4 pages, 3 figures (5 assets)
☆ Conformalized Non-uniform Sampling Strategies for Accelerated Sampling-based Motion Planning
Sampling-based motion planners (SBMPs) are widely used to compute dynamically feasible robot paths. However, their reliance on uniform sampling often leads to poor efficiency and slow planning in complex environments. We introduce a novel non-uniform sampling strategy that integrates into existing SBMPs by biasing sampling toward `certified' regions. These regions are constructed by (i) generating an initial, possibly infeasible, path using any heuristic path predictor (e.g., A* or vision-language models) and (ii) applying conformal prediction to quantify the predictor's uncertainty. This process yields prediction sets around the initial-guess path that are guaranteed, with user-specified probability, to contain the optimal solution. To our knowledge, this is the first non-uniform sampling approach for SBMPs that provides such probabilistically correct guarantees on the sampling regions. Extensive evaluations demonstrate that our method consistently finds feasible paths faster and generalizes better to unseen environments than existing baselines.
☆ Isaac Lab: A GPU-Accelerated Simulation Framework for Multi-Modal Robot Learning
We present Isaac Lab, the natural successor to Isaac Gym, which extends the paradigm of GPU-native robotics simulation into the era of large-scale multi-modal learning. Isaac Lab combines high-fidelity GPU parallel physics, photorealistic rendering, and a modular, composable architecture for designing environments and training robot policies. Beyond physics and rendering, the framework integrates actuator models, multi-frequency sensor simulation, data collection pipelines, and domain randomization tools, unifying best practices for reinforcement and imitation learning at scale within a single extensible platform. We highlight its application to a diverse set of challenges, including whole-body control, cross-embodiment mobility, contact-rich and dexterous manipulation, and the integration of human demonstrations for skill acquisition. Finally, we discuss upcoming integration with the differentiable, GPU-accelerated Newton physics engine, which promises new opportunities for scalable, data-efficient, and gradient-based approaches to robot learning. We believe Isaac Lab's combination of advanced simulation capabilities, rich sensing, and data-center scale execution will help unlock the next generation of breakthroughs in robotics research.
comment: Code and documentation are available here: https://github.com/isaac-sim/IsaacLab
☆ Pixi: Unified Software Development and Distribution for Robotics and AI
The reproducibility crisis in scientific computing constrains robotics research. Existing studies reveal that up to 70% of robotics algorithms cannot be reproduced by independent teams, while many others fail to reach deployment because creating shareable software environments remains prohibitively complex. These challenges stem from fragmented, multi-language, and hardware-software toolchains that lead to dependency hell. We present Pixi, a unified package-management framework that addresses these issues by capturing exact dependency states in project-level lockfiles, ensuring bit-for-bit reproducibility across platforms. Its high-performance SAT solver achieves up to 10x faster dependency resolution than comparable tools, while integration of the conda-forge and PyPI ecosystems removes the need for multiple managers. Adopted in over 5,300 projects since 2023, Pixi reduces setup times from hours to minutes and lowers technical barriers for researchers worldwide. By enabling scalable, reproducible, collaborative research infrastructure, Pixi accelerates progress in robotics and AI.
comment: 20 pages, 3 figures, 11 code snippets
☆ Unified Multimodal Diffusion Forcing for Forceful Manipulation
Given a dataset of expert trajectories, standard imitation learning approaches typically learn a direct mapping from observations (e.g., RGB images) to actions. However, such methods often overlook the rich interplay between different modalities, i.e., sensory inputs, actions, and rewards, which is crucial for modeling robot behavior and understanding task outcomes. In this work, we propose Multimodal Diffusion Forcing, a unified framework for learning from multimodal robot trajectories that extends beyond action generation. Rather than modeling a fixed distribution, MDF applies random partial masking and trains a diffusion model to reconstruct the trajectory. This training objective encourages the model to learn temporal and cross-modal dependencies, such as predicting the effects of actions on force signals or inferring states from partial observations. We evaluate MDF on contact-rich, forceful manipulation tasks in simulated and real-world environments. Our results show that MDF not only delivers versatile functionalities, but also achieves strong performance, and robustness under noisy observations. More visualizations can be found on our website https://unified-df.github.io
comment: Project website: https://unified-df.github.io
☆ ReGen: Generative Robot Simulation via Inverse Design
Simulation plays a key role in scaling robot learning and validating policies, but constructing simulations remains a labor-intensive process. This paper introduces ReGen, a generative simulation framework that automates simulation design via inverse design. Given a robot's behavior -- such as a motion trajectory or an objective function -- and its textual description, ReGen infers plausible scenarios and environments that could have caused the behavior. ReGen leverages large language models to synthesize scenarios by expanding a directed graph that encodes cause-and-effect relationships, relevant entities, and their properties. This structured graph is then translated into a symbolic program, which configures and executes a robot simulation environment. Our framework supports (i) augmenting simulations based on ego-agent behaviors, (ii) controllable, counterfactual scenario generation, (iii) reasoning about agent cognition and mental states, and (iv) reasoning with distinct sensing modalities, such as braking due to faulty GPS signals. We demonstrate ReGen in autonomous driving and robot manipulation tasks, generating more diverse, complex simulated environments compared to existing simulations with high success rates, and enabling controllable generation for corner cases. This approach enhances the validation of robot policies and supports data or simulation augmentation, advancing scalable robot learning for improved generalization and robustness. We provide code and example videos at: https://regen-sim.github.io/
☆ ScheduleStream: Temporal Planning with Samplers for GPU-Accelerated Multi-Arm Task and Motion Planning & Scheduling
Bimanual and humanoid robots are appealing because of their human-like ability to leverage multiple arms to efficiently complete tasks. However, controlling multiple arms at once is computationally challenging due to the growth in the hybrid discrete-continuous action space. Task and Motion Planning (TAMP) algorithms can efficiently plan in hybrid spaces but generally produce plans, where only one arm is moving at a time, rather than schedules that allow for parallel arm motion. In order to extend TAMP to produce schedules, we present ScheduleStream, the first general-purpose framework for planning & scheduling with sampling operations. ScheduleStream models temporal dynamics using hybrid durative actions, which can be started asynchronously and persist for a duration that's a function of their parameters. We propose domain-independent algorithms that solve ScheduleStream problems without any application-specific mechanisms. We apply ScheduleStream to Task and Motion Planning & Scheduling (TAMPAS), where we use GPU acceleration within samplers to expedite planning. We compare ScheduleStream algorithms to several ablations in simulation and find that they produce more efficient solutions. We demonstrate ScheduleStream on several real-world bimanual robot tasks at https://schedulestream.github.io.
comment: Project website: https://schedulestream.github.io
♻ ☆ Particle-Grid Neural Dynamics for Learning Deformable Object Models from RGB-D Videos
Modeling the dynamics of deformable objects is challenging due to their diverse physical properties and the difficulty of estimating states from limited visual information. We address these challenges with a neural dynamics framework that combines object particles and spatial grids in a hybrid representation. Our particle-grid model captures global shape and motion information while predicting dense particle movements, enabling the modeling of objects with varied shapes and materials. Particles represent object shapes, while the spatial grid discretizes the 3D space to ensure spatial continuity and enhance learning efficiency. Coupled with Gaussian Splattings for visual rendering, our framework achieves a fully learning-based digital twin of deformable objects and generates 3D action-conditioned videos. Through experiments, we demonstrate that our model learns the dynamics of diverse objects -- such as ropes, cloths, stuffed animals, and paper bags -- from sparse-view RGB-D recordings of robot-object interactions, while also generalizing at the category level to unseen instances. Our approach outperforms state-of-the-art learning-based and physics-based simulators, particularly in scenarios with limited camera views. Furthermore, we showcase the utility of our learned models in model-based planning, enabling goal-conditioned object manipulation across a range of tasks. The project page is available at https://kywind.github.io/pgnd .
comment: Project page: https://kywind.github.io/pgnd
♻ ☆ The Mini Wheelbot: A Testbed for Learning-based Balancing, Flips, and Articulated Driving
The Mini Wheelbot is a balancing, reaction wheel unicycle robot designed as a testbed for learning-based control. It is an unstable system with highly nonlinear yaw dynamics, non-holonomic driving, and discrete contact switches in a small, powerful, and rugged form factor. The Mini Wheelbot can use its wheels to stand up from any initial orientation - enabling automatic environment resets in repetitive experiments and even challenging half flips. We illustrate the effectiveness of the Mini Wheelbot as a testbed by implementing two popular learning-based control algorithms. First, we showcase Bayesian optimization for tuning the balancing controller. Second, we use imitation learning from an expert nonlinear MPC that uses gyroscopic effects to reorient the robot and can track higher-level velocity and orientation commands. The latter allows the robot to drive around based on user commands - for the first time in this class of robots. The Mini Wheelbot is not only compelling for testing learning-based control algorithms, but it is also just fun to work with, as demonstrated in the video of our experiments.
♻ ☆ Action Deviation-Aware Inference for Low-Latency Wireless Robots
To support latency-sensitive AI applications ranging from autonomous driving to industrial robot manipulation, 6G envisions distributed ML with computational resources in mobile, edge, and cloud connected over hyper-reliable low-latency communication (HRLLC). In this setting, speculative decoding can facilitate collaborative inference of models distributively deployed: a lightweight on-device model locally generates drafts while a more capable remote target model on a server verifies and corrects them in parallel with speculative sampling, thus resulting in lower latency without compromising accuracy. However, unlike autoregressive text generation, behavior cloning policies, typically used for embodied AI applications, cannot parallelize verification and correction for multiple drafts as each generated action depends on observation updated by a previous action. To this end, we propose Action Deviation-Aware Hybrid Inference (ADAHI), wherein drafts are selectively transmitted and verified based on action deviation, which has a strong correlation with action's rejection probability by the target model. By invoking server operation only when necessary, communication and computational overhead can be reduced while accuracy gain from speculative sampling is preserved. Experiments on our testbed show that ADAHI reduces transmission and server operations by approximately 40%, lowers end-to-end latency by 39.2%, and attains up to 97.2% of the task-success rate of baseline that invokes speculative sampling for every draft embedding vector.
♻ ☆ When Semantics Connect the Swarm: LLM-Driven Fuzzy Control for Cooperative Multi-Robot Underwater Coverage
Underwater multi-robot cooperative coverage remains challenging due to partial observability, limited communication, environmental uncertainty, and the lack of access to global localization. To address these issues, this paper presents a semantics-guided fuzzy control framework that couples Large Language Models (LLMs) with interpretable control and lightweight coordination. Raw multimodal observations are compressed by the LLM into compact, human-interpretable semantic tokens that summarize obstacles, unexplored regions, and Objects Of Interest (OOIs) under uncertain perception. A fuzzy inference system with pre-defined membership functions then maps these tokens into smooth and stable steering and gait commands, enabling reliable navigation without relying on global positioning. Then, we further coordinate multiple robots by introducing semantic communication that shares intent and local context in linguistic form, enabling agreement on who explores where while avoiding redundant revisits. Extensive simulations in unknown reef-like environments show that, under limited sensing and communication, the proposed framework achieves robust OOI-oriented navigation and cooperative coverage with improved efficiency and adaptability, narrowing the gap between semantic cognition and distributed underwater control in GPS-denied, map-free conditions.
comment: This paper has been submitted to IEEE Transactions on Mobile Computing. Jingzehua Xu, Weihang Zhang, and Yangyang Li contributed equally to this work and are recognized as the co-first authors of the paper
♻ ☆ SLAM&Render: A Benchmark for the Intersection Between Neural Rendering, Gaussian Splatting and SLAM
Models and methods originally developed for Novel View Synthesis and Scene Rendering, such as Neural Radiance Fields (NeRF) and Gaussian Splatting, are increasingly being adopted as representations in Simultaneous Localization and Mapping (SLAM). However, existing datasets fail to include the specific challenges of both fields, such as sequential operations and, in many settings, multi-modality in SLAM or generalization across viewpoints and illumination conditions in neural rendering. Additionally, the data are often collected using sensors which are handheld or mounted on drones or mobile robots, which complicates the accurate reproduction of sensor motions. To bridge these gaps, we introduce SLAM&Render, a novel dataset designed to benchmark methods in the intersection between SLAM, Novel View Rendering and Gaussian Splatting. Recorded with a robot manipulator, it uniquely includes 40 sequences with time-synchronized RGB-D images, IMU readings, robot kinematic data, and ground-truth pose streams. By releasing robot kinematic data, the dataset also enables the assessment of recent integrations of SLAM paradigms within robotic applications. The dataset features five setups with consumer and industrial objects under four controlled lighting conditions, each with separate training and test trajectories. All sequences are static with different levels of object rearrangements and occlusions. Our experimental results, obtained with several baselines from the literature, validate SLAM&Render as a relevant benchmark for this emerging research area.
comment: 9 pages, 8 figures, submitted to The International Journal of Robotics Research (IJRR)
♻ ☆ Application Management in C-ITS: Orchestrating Demand-Driven Deployments and Reconfigurations SC 2025
Vehicles are becoming increasingly automated and interconnected, enabling the formation of cooperative intelligent transport systems (C-ITS) and the use of offboard services. As a result, cloud-native techniques, such as microservices and container orchestration, play an increasingly important role in their operation. However, orchestrating applications in a large-scale C-ITS poses unique challenges due to the dynamic nature of the environment and the need for efficient resource utilization. In this paper, we present a demand-driven application management approach that leverages cloud-native techniques - specifically Kubernetes - to address these challenges. Taking into account the demands originating from different entities within the C-ITS, the approach enables the automation of processes, such as deployment, reconfiguration, update, upgrade, and scaling of microservices. Executing these processes on demand can, for example, reduce computing resource consumption and network traffic. A demand may include a request for provisioning an external supporting service, such as a collective environment model. The approach handles changing and new demands by dynamically reconciling them through our proposed application management framework built on Kubernetes and the Robot Operating System (ROS 2). We demonstrate the operation of our framework in the C-ITS use case of collective environment perception and make the source code of the prototypical framework publicly available at https://github.com/ika-rwth-aachen/application_manager.
comment: 7 pages, 2 figures, 2 tables; Accepted to be published as part of the 2025 IEEE International Conference on Intelligent Transportation Systems (ITSC 2025), Gold Coast, Australia, November 18-21, 2025
♻ ☆ SafeVLA: Towards Safety Alignment of Vision-Language-Action Model via Constrained Learning NeurIPS 2025
Vision-language-action models (VLAs) show potential as generalist robot policies. However, these models pose extreme safety challenges during real-world deployment, including the risk of harm to the environment, the robot itself, and humans. How can safety constraints be explicitly integrated into VLAs? We address this by exploring an integrated safety approach (ISA), systematically modeling safety requirements, then actively eliciting diverse unsafe behaviors, effectively constraining VLA policies via safe reinforcement learning, and rigorously assuring their safety through targeted evaluations. Leveraging the constrained Markov decision process (CMDP) paradigm, ISA optimizes VLAs from a min-max perspective against elicited safety risks. Thus, policies aligned through this comprehensive approach achieve the following key features: (I) effective safety-performance trade-offs, reducing the cumulative cost of safety violations by 83.58% compared to the state-of-the-art method, while also maintaining task success rate (+3.85%). (II) strong safety assurance, with the ability to mitigate long-tail risks and handle extreme failure scenarios. (III) robust generalization of learned safety behaviors to various out-of-distribution perturbations. The effectiveness is evaluated on long-horizon mobile manipulation tasks. Our data, models and newly proposed benchmark environment are available at https://pku-safevla.github.io.
comment: Accepted by NeurIPS 2025 Spotlight Presentation
♻ ☆ HiMaCon: Discovering Hierarchical Manipulation Concepts from Unlabeled Multi-Modal Data NeurIPS 2025
Effective generalization in robotic manipulation requires representations that capture invariant patterns of interaction across environments and tasks. We present a self-supervised framework for learning hierarchical manipulation concepts that encode these invariant patterns through cross-modal sensory correlations and multi-level temporal abstractions without requiring human annotation. Our approach combines a cross-modal correlation network that identifies persistent patterns across sensory modalities with a multi-horizon predictor that organizes representations hierarchically across temporal scales. Manipulation concepts learned through this dual structure enable policies to focus on transferable relational patterns while maintaining awareness of both immediate actions and longer-term goals. Empirical evaluation across simulated benchmarks and real-world deployments demonstrates significant performance improvements with our concept-enhanced policies. Analysis reveals that the learned concepts resemble human-interpretable manipulation primitives despite receiving no semantic supervision. This work advances both the understanding of representation learning for manipulation and provides a practical approach to enhancing robotic performance in complex scenarios.
comment: Accepted at 39th Conference on Neural Information Processing Systems (NeurIPS 2025)
♻ ☆ Learning to Navigate Socially Through Proactive Risk Perception
In this report, we describe the technical details of our submission to the IROS 2025 RoboSense Challenge Social Navigation Track. This track focuses on developing RGBD-based perception and navigation systems that enable autonomous agents to navigate safely, efficiently, and socially compliantly in dynamic human-populated indoor environments. The challenge requires agents to operate from an egocentric perspective using only onboard sensors including RGB-D observations and odometry, without access to global maps or privileged information, while maintaining social norm compliance such as safe distances and collision avoidance. Building upon the Falcon model, we introduce a Proactive Risk Perception Module to enhance social navigation performance. Our approach augments Falcon with collision risk understanding that learns to predict distance-based collision risk scores for surrounding humans, which enables the agent to develop more robust spatial awareness and proactive collision avoidance behaviors. The evaluation on the Social-HM3D benchmark demonstrates that our method improves the agent's ability to maintain personal space compliance while navigating toward goals in crowded indoor scenes with dynamic human agents, achieving 2nd place among 16 participating teams in the challenge.
♻ ☆ MLP-SLAM: Multilayer Perceptron-Based Simultaneous Localization and Mapping
The Visual Simultaneous Localization and Mapping (V-SLAM) system has seen significant development in recent years, demonstrating high precision in environments with limited dynamic objects. However, their performance significantly deteriorates when deployed in settings with a higher presence of movable objects, such as environments with pedestrians, cars, and buses, which are common in outdoor scenes. To address this issue, we propose a Multilayer Perceptron (MLP)-based real-time stereo SLAM system that leverages complete geometry information to avoid information loss. Moreover, there is currently no publicly available dataset for directly evaluating the effectiveness of dynamic and static feature classification methods, and to bridge this gap, we have created a publicly available dataset containing over 50,000 feature points. Experimental results demonstrate that our MLP-based dynamic and static feature point discriminator has achieved superior performance compared to other methods on this dataset. Furthermore, the MLP-based real-time stereo SLAM system has shown the highest average precision and fastest speed on the outdoor KITTI tracking datasets compared to other dynamic SLAM systems.The open-source code and datasets are available at https://github.com/TaozheLi/MLP-SLAM.
comment: Dynamic SLAM
♻ ☆ Poutine: Vision-Language-Trajectory Pre-Training and Reinforcement Learning Post-Training Enable Robust End-to-End Autonomous Driving
Maintaining good driving behavior in out-of-distribution scenarios remains a critical challenge in autonomous driving. A promising direction is to leverage the generalist knowledge and reasoning capabilities of large-language models by treating unusual driving scenarios as a logical reasoning task. In this work, we present Poutine, a method that uses an off-the-shelf 3B-parameter vision-language model (VLM) - without any additional components - to achieve robust end-to-end autonomous driving via a simple and scalable training recipe. To learn strong base driving capabilities, we first train Poutine-Base using self-supervised next-token prediction over vision, language, and trajectory (VLT) tokens, leveraging both nominal and long-tail driving data. In the second stage, we fine-tune Poutine-Base using Group Relative Policy Optimization (GRPO) with a small set of human preference-labeled examples. We evaluated our approach on the Waymo end-to-end driving benchmark curated for long-tail scenarios. The final Poutine model achieves an RFS of 7.99 on the test set, placing 1st in the 2025 Waymo Vision-Based End-to-End Driving Challenge by a significant margin. Our results suggest that handcrafted tokenizers or custom architectural components added to base VLMs in prior work are not necessary to achieve strong driving performance. Instead, this work highlights the potential of scalable VLT pretraining combined with lightweight RL fine-tuning to enable robust and generalizable autonomous driving.
♻ ☆ Team Xiaomi EV-AD VLA: Caption-Guided Retrieval System for Cross-Modal Drone Navigation -- Technical Report for IROS 2025 RoboSense Challenge Track 4
Cross-modal drone navigation remains a challenging task in robotics, requiring efficient retrieval of relevant images from large-scale databases based on natural language descriptions. The RoboSense 2025 Track 4 challenge addresses this challenge, focusing on robust, natural language-guided cross-view image retrieval across multiple platforms (drones, satellites, and ground cameras). Current baseline methods, while effective for initial retrieval, often struggle to achieve fine-grained semantic matching between text queries and visual content, especially in complex aerial scenes. To address this challenge, we propose a two-stage retrieval refinement method: Caption-Guided Retrieval System (CGRS) that enhances the baseline coarse ranking through intelligent reranking. Our method first leverages a baseline model to obtain an initial coarse ranking of the top 20 most relevant images for each query. We then use Vision-Language-Model (VLM) to generate detailed captions for these candidate images, capturing rich semantic descriptions of their visual content. These generated captions are then used in a multimodal similarity computation framework to perform fine-grained reranking of the original text query, effectively building a semantic bridge between the visual content and natural language descriptions. Our approach significantly improves upon the baseline, achieving a consistent 5\% improvement across all key metrics (Recall@1, Recall@5, and Recall@10). Our approach win TOP-2 in the challenge, demonstrating the practical value of our semantic refinement strategy in real-world robotic navigation scenarios.
♻ ☆ OmniVLA: Physically-Grounded Multimodal VLA with Unified Multi-Sensor Perception for Robotic Manipulation
Vision-language-action (VLA) models have shown strong generalization for robotic action prediction through large-scale vision-language pretraining. However, most existing models rely solely on RGB cameras, limiting their perception and, consequently, manipulation capabilities. We present OmniVLA, an omni-modality VLA model that integrates novel sensing modalities for physically-grounded spatial intelligence beyond RGB perception. The core of our approach is the sensor-masked image, a unified representation that overlays spatially grounded and physically meaningful masks onto the RGB images, derived from sensors including an infrared camera, a mmWave radar, and a microphone array. This image-native unification keeps sensor input close to RGB statistics to facilitate training, provides a uniform interface across sensor hardware, and enables data-efficient learning with lightweight per-sensor projectors. Built on this, we present a multisensory vision-language-action model architecture and train the model based on an RGB-pretrained VLA backbone. We evaluate OmniVLA on challenging real-world tasks where sensor-modality perception guides the robotic manipulation. OmniVLA achieves an average task success rate of 84%, significantly outperforms both RGB-only and raw-sensor-input baseline models by 59% and 28% respectively, meanwhile showing higher learning efficiency and stronger generalization capability.
♻ ☆ Text to Robotic Assembly of Multi Component Objects using 3D Generative AI and Vision Language Models NeurIPS 2025
Advances in 3D generative AI have enabled the creation of physical objects from text prompts, but challenges remain in creating objects involving multiple component types. We present a pipeline that integrates 3D generative AI with vision-language models (VLMs) to enable the robotic assembly of multi-component objects from natural language. Our method leverages VLMs for zero-shot, multi-modal reasoning about geometry and functionality to decompose AI-generated meshes into multi-component 3D models using predefined structural and panel components. We demonstrate that a VLM is capable of determining which mesh regions need panel components in addition to structural components, based on the object's geometry and functionality. Evaluation across test objects shows that users preferred the VLM-generated assignments 90.6% of the time, compared to 59.4% for rule-based and 2.5% for random assignment. Lastly, the system allows users to refine component assignments through conversational feedback, enabling greater human control and agency in making physical objects with generative AI and robotics.
comment: Accepted to NeurIPS 2025, Conference on Neural Information Processing Systems, Creative AI Track
♻ ☆ Toward Engineering AGI: Benchmarking the Engineering Design Capabilities of LLMs NeurIPS 2025
Modern engineering, spanning electrical, mechanical, aerospace, civil, and computer disciplines, stands as a cornerstone of human civilization and the foundation of our society. However, engineering design poses a fundamentally different challenge for large language models (LLMs) compared with traditional textbook-style problem solving or factual question answering. Although existing benchmarks have driven progress in areas such as language understanding, code synthesis, and scientific problem solving, real-world engineering design demands the synthesis of domain knowledge, navigation of complex trade-offs, and management of the tedious processes that consume much of practicing engineers' time. Despite these shared challenges across engineering disciplines, no benchmark currently captures the unique demands of engineering design work. In this work, we introduce EngDesign, an Engineering Design benchmark that evaluates LLMs' abilities to perform practical design tasks across nine engineering domains. Unlike existing benchmarks that focus on factual recall or question answering, EngDesign uniquely emphasizes LLMs' ability to synthesize domain knowledge, reason under constraints, and generate functional, objective-oriented engineering designs. Each task in EngDesign represents a real-world engineering design problem, accompanied by a detailed task description specifying design goals, constraints, and performance requirements. EngDesign pioneers a simulation-based evaluation paradigm that moves beyond textbook knowledge to assess genuine engineering design capabilities and shifts evaluation from static answer checking to dynamic, simulation-driven functional verification, marking a crucial step toward realizing the vision of engineering Artificial General Intelligence (AGI).
comment: To Appear in NeurIPS 2025 Datasets & Benchmarks Track
♻ ☆ Joint Verification and Refinement of Language Models for Safety-Constrained Planning
Large language models possess impressive capabilities in generating programs (e.g., Python) from natural language descriptions to execute robotic tasks. However, these generated programs often contain errors that violate externally given task specifications. Without an effective method to verify their correctness, the reliable deployment of language models in real-world systems is practically infeasible. We develop a method that converts generated robot programs into an automaton-based representation and verifies them against task-relevant safety specifications. We establish a theorem that any arbitrary combination of the verified programs will also satisfy the safety specifications. Hence, the method eliminates the need to verify complex programs composed of multiple simpler ones, reducing computation complexity. We then introduce an automated fine-tuning procedure that leverages verification outcomes for supervision. By applying the theorem, this procedure only requires training the model to generate safe sub-components, thereby improving training efficiency. Empirical results on robot applications show a 30 percent increase in the probability of generating specification-compliant programs, with training time reduced by half compared to fine-tuning on generating full programs.
Robotics 39
☆ Source-Free Bistable Fluidic Gripper for Size-Selective and Stiffness-Adaptive Grasping
Conventional fluid-driven soft grippers typically depend on external sources, which limit portability and long-term autonomy. This work introduces a self-contained soft gripper with fixed size that operates solely through internal liquid redistribution among three interconnected bistable snap-through chambers. When the top sensing chamber deforms upon contact, the displaced liquid triggers snap-through expansion of the grasping chambers, enabling stable and size-selective grasping without continuous energy input. The internal hydraulic feedback further allows passive adaptation of gripping pressure to object stiffness. This source-free and compact design opens new possibilities for lightweight, stiffness-adaptive fluid-driven manipulation in soft robotics, providing a feasible approach for targeted size-specific sampling and operation in underwater and field environments.
☆ Unconscious and Intentional Human Motion Cues for Expressive Robot-Arm Motion Design
This study investigates how human motion cues can be used to design expressive robot-arm movements. Using the imperfect-information game Geister, we analyzed two types of human piece-moving motions: natural gameplay (unconscious tendencies) and instructed expressions (intentional cues). Based on these findings, we created phase-specific robot motions by varying movement speed and stop duration, and evaluated observer impressions under two presentation modalities: a physical robot and a recorded video. Results indicate that late-phase motion timing, particularly during withdrawal, plays an important role in impression formation and that physical embodiment enhances the interpretability of motion cues. These findings provide insights for designing expressive robot motions based on human timing behavior.
comment: 5 pages, 5 figures, HAI2025 Workshop on Socially Aware and Cooperative Intelligent Systems
☆ Motion Planning Under Temporal Logic Specifications In Semantically Unknown Environments
This paper addresses a motion planning problem to achieve spatio-temporal-logical tasks, expressed by syntactically co-safe linear temporal logic specifications (scLTL\next), in uncertain environments. Here, the uncertainty is modeled as some probabilistic knowledge on the semantic labels of the environment. For example, the task is "first go to region 1, then go to region 2"; however, the exact locations of regions 1 and 2 are not known a priori, instead a probabilistic belief is available. We propose a novel automata-theoretic approach, where a special product automaton is constructed to capture the uncertainty related to semantic labels, and a reward function is designed for each edge of this product automaton. The proposed algorithm utilizes value iteration for online replanning. We show some theoretical results and present some simulations/experiments to demonstrate the efficacy of the proposed approach.
comment: 8 pages, 6 figures
☆ Flying Robotics Art: ROS-based Drone Draws the Record-Breaking Mural
This paper presents the innovative design and successful deployment of a pioneering autonomous unmanned aerial system developed for executing the world's largest mural painted by a drone. Addressing the dual challenges of maintaining artistic precision and operational reliability under adverse outdoor conditions such as wind and direct sunlight, our work introduces a robust system capable of navigating and painting outdoors with unprecedented accuracy. Key to our approach is a novel navigation system that combines an infrared (IR) motion capture camera and LiDAR technology, enabling precise location tracking tailored specifically for largescale artistic applications. We employ a unique control architecture that uses different regulation in tangential and normal directions relative to the planned path, enabling precise trajectory tracking and stable line rendering. We also present algorithms for trajectory planning and path optimization, allowing for complex curve drawing and area filling. The system includes a custom-designed paint spraying mechanism, specifically engineered to function effectively amidst the turbulent airflow generated by the drone's propellers, which also protects the drone's critical components from paint-related damage, ensuring longevity and consistent performance. Experimental results demonstrate the system's robustness and precision in varied conditions, showcasing its potential for autonomous large-scale art creation and expanding the functional applications of robotics in creative fields.
☆ Multi-robot searching with limited sensing range for static and mobile intruders
We consider the problem of searching for an intruder in a geometric domain by utilizing multiple search robots. The domain is a simply connected orthogonal polygon with edges parallel to the cartesian coordinate axes. Each robot has a limited sensing capability. We study the problem for both static and mobile intruders. It turns out that the problem of finding an intruder is NP-hard, even for a stationary intruder. Given this intractability, we turn our attention towards developing efficient and robust algorithms, namely methods based on space-filling curves, random search, and cooperative random search. Moreover, for each proposed algorithm, we evaluate the trade-off between the number of search robots and the time required for the robots to complete the search process while considering the geometric properties of the connected orthogonal search area.
☆ Manifold-constrained Hamilton-Jacobi Reachability Learning for Decentralized Multi-Agent Motion Planning
Safe multi-agent motion planning (MAMP) under task-induced constraints is a critical challenge in robotics. Many real-world scenarios require robots to navigate dynamic environments while adhering to manifold constraints imposed by tasks. For example, service robots must carry cups upright while avoiding collisions with humans or other robots. Despite recent advances in decentralized MAMP for high-dimensional systems, incorporating manifold constraints remains difficult. To address this, we propose a manifold-constrained Hamilton-Jacobi reachability (HJR) learning framework for decentralized MAMP. Our method solves HJR problems under manifold constraints to capture task-aware safety conditions, which are then integrated into a decentralized trajectory optimization planner. This enables robots to generate motion plans that are both safe and task-feasible without requiring assumptions about other agents' policies. Our approach generalizes across diverse manifold-constrained tasks and scales effectively to high-dimensional multi-agent manipulation problems. Experiments show that our method outperforms existing constrained motion planners and operates at speeds suitable for real-world applications. Video demonstrations are available at https://youtu.be/RYcEHMnPTH8 .
☆ Multi-User Personalisation in Human-Robot Interaction: Using Quantitative Bipolar Argumentation Frameworks for Preferences Conflict Resolution
While personalisation in Human-Robot Interaction (HRI) has advanced significantly, most existing approaches focus on single-user adaptation, overlooking scenarios involving multiple stakeholders with potentially conflicting preferences. To address this, we propose the Multi-User Preferences Quantitative Bipolar Argumentation Framework (MUP-QBAF), a novel multi-user personalisation framework based on Quantitative Bipolar Argumentation Frameworks (QBAFs) that explicitly models and resolves multi-user preference conflicts. Unlike prior work in Argumentation Frameworks, which typically assumes static inputs, our approach is tailored to robotics: it incorporates both users' arguments and the robot's dynamic observations of the environment, allowing the system to adapt over time and respond to changing contexts. Preferences, both positive and negative, are represented as arguments whose strength is recalculated iteratively based on new information. The framework's properties and capabilities are presented and validated through a realistic case study, where an assistive robot mediates between the conflicting preferences of a caregiver and a care recipient during a frailty assessment task. This evaluation further includes a sensitivity analysis of argument base scores, demonstrating how preference outcomes can be shaped by user input and contextual observations. By offering a transparent, structured, and context-sensitive approach to resolving competing user preferences, this work advances the field of multi-user HRI. It provides a principled alternative to data-driven methods, enabling robots to navigate conflicts in real-world environments.
comment: Preprint submitted to a journal
☆ OneOcc: Semantic Occupancy Prediction for Legged Robots with a Single Panoramic Camera
Robust 3D semantic occupancy is crucial for legged/humanoid robots, yet most semantic scene completion (SSC) systems target wheeled platforms with forward-facing sensors. We present OneOcc, a vision-only panoramic SSC framework designed for gait-introduced body jitter and 360{\deg} continuity. OneOcc combines: (i) Dual-Projection fusion (DP-ER) to exploit the annular panorama and its equirectangular unfolding, preserving 360{\deg} continuity and grid alignment; (ii) Bi-Grid Voxelization (BGV) to reason in Cartesian and cylindrical-polar spaces, reducing discretization bias and sharpening free/occupied boundaries; (iii) a lightweight decoder with Hierarchical AMoE-3D for dynamic multi-scale fusion and better long-range/occlusion reasoning; and (iv) plug-and-play Gait Displacement Compensation (GDC) learning feature-level motion correction without extra sensors. We also release two panoramic occupancy benchmarks: QuadOcc (real quadruped, first-person 360{\deg}) and Human360Occ (H3O) (CARLA human-ego 360{\deg} with RGB, Depth, semantic occupancy; standardized within-/cross-city splits). OneOcc sets new state-of-the-art (SOTA): on QuadOcc it beats strong vision baselines and popular LiDAR ones; on H3O it gains +3.83 mIoU (within-city) and +8.08 (cross-city). Modules are lightweight, enabling deployable full-surround perception for legged/humanoid robots. Datasets and code will be publicly available at https://github.com/MasterHow/OneOcc.
comment: Datasets and code will be publicly available at https://github.com/MasterHow/OneOcc
☆ Indicating Robot Vision Capabilities with Augmented Reality
Research indicates that humans can mistakenly assume that robots and humans have the same field of view (FoV), possessing an inaccurate mental model of robots. This misperception may lead to failures during human-robot collaboration tasks where robots might be asked to complete impossible tasks about out-of-view objects. The issue is more severe when robots do not have a chance to scan the scene to update their world model while focusing on assigned tasks. To help align humans' mental models of robots' vision capabilities, we propose four FoV indicators in augmented reality (AR) and conducted a user human-subjects experiment (N=41) to evaluate them in terms of accuracy, confidence, task efficiency, and workload. These indicators span a spectrum from egocentric (robot's eye and head space) to allocentric (task space). Results showed that the allocentric blocks at the task space had the highest accuracy with a delay in interpreting the robot's FoV. The egocentric indicator of deeper eye sockets, possible for physical alteration, also increased accuracy. In all indicators, participants' confidence was high while cognitive load remained low. Finally, we contribute six guidelines for practitioners to apply our AR indicators or physical alterations to align humans' mental models with robots' vision capabilities.
☆ ROSBag MCP Server: Analyzing Robot Data with LLMs for Agentic Embodied AI Applications
Agentic AI systems and Physical or Embodied AI systems have been two key research verticals at the forefront of Artificial Intelligence and Robotics, with Model Context Protocol (MCP) increasingly becoming a key component and enabler of agentic applications. However, the literature at the intersection of these verticals, i.e., Agentic Embodied AI, remains scarce. This paper introduces an MCP server for analyzing ROS and ROS 2 bags, allowing for analyzing, visualizing and processing robot data with natural language through LLMs and VLMs. We describe specific tooling built with robotics domain knowledge, with our initial release focused on mobile robotics and supporting natively the analysis of trajectories, laser scan data, transforms, or time series data. This is in addition to providing an interface to standard ROS 2 CLI tools ("ros2 bag list" or "ros2 bag info"), as well as the ability to filter bags with a subset of topics or trimmed in time. Coupled with the MCP server, we provide a lightweight UI that allows the benchmarking of the tooling with different LLMs, both proprietary (Anthropic, OpenAI) and open-source (through Groq). Our experimental results include the analysis of tool calling capabilities of eight different state-of-the-art LLM/VLM models, both proprietary and open-source, large and small. Our experiments indicate that there is a large divide in tool calling capabilities, with Kimi K2 and Claude Sonnet 4 demonstrating clearly superior performance. We also conclude that there are multiple factors affecting the success rates, from the tool description schema to the number of arguments, as well as the number of tools available to the models. The code is available with a permissive license at https://github.com/binabik-ai/mcp-rosbags.
☆ Development of the Bioinspired Tendon-Driven DexHand 021 with Proprioceptive Compliance Control
The human hand plays a vital role in daily life and industrial applications, yet replicating its multifunctional capabilities-including motion, sensing, and coordinated manipulation-with robotic systems remains a formidable challenge. Developing a dexterous robotic hand requires balancing human-like agility with engineering constraints such as complexity, size-to-weight ratio, durability, and force-sensing performance. This letter presents Dex-Hand 021, a high-performance, cable-driven five-finger robotic hand with 12 active and 7 passive degrees of freedom (DoFs), achieving 19 DoFs dexterity in a lightweight 1 kg design. We propose a proprioceptive force-sensing-based admittance control method to enhance manipulation. Experimental results demonstrate its superior performance: a single-finger load capacity exceeding 10 N, fingertip repeatability under 0.001 m, and force estimation errors below 0.2 N. Compared to PID control, joint torques in multi-object grasping are reduced by 31.19%, significantly improves force-sensing capability while preventing overload during collisions. The hand excels in both power and precision grasps, successfully executing 33 GRASP taxonomy motions and complex manipulation tasks. This work advances the design of lightweight, industrial-grade dexterous hands and enhances proprioceptive control, contributing to robotic manipulation and intelligent manufacturing.
comment: 8 pages 18 fogures, IEEE RAL accept
☆ Value Elicitation for a Socially Assistive Robot Addressing Social Anxiety: A Participatory Design Approach ECAI 2025
Social anxiety is a prevalent mental health condition that can significantly impact overall well-being and quality of life. Despite its widespread effects, adequate support or treatment for social anxiety is often insufficient. Advances in technology, particularly in social robotics, offer promising opportunities to complement traditional mental health. As an initial step toward developing effective solutions, it is essential to understand the values that shape what is considered meaningful, acceptable, and helpful. In this study, a participatory design workshop was conducted with mental health academic researchers to elicit the underlying values that should inform the design of socially assistive robots for social anxiety support. Through creative, reflective, and envisioning activities, participants explored scenarios and design possibilities, allowing for systematic elicitation of values, expectations, needs, and preferences related to robot-supported interventions. The findings reveal rich insights into design-relevant values-including adaptivity, acceptance, and efficacy-that are core to support for individuals with social anxiety. This study highlights the significance of a research-led approach to value elicitation, emphasising user-centred and context-aware design considerations in the development of socially assistive robots.
comment: Accepted at Value Engineering in AI (VALE) Workshop (ECAI 2025)
☆ GUIDES: Guidance Using Instructor-Distilled Embeddings for Pre-trained Robot Policy Enhancement IROS 2025
Pre-trained robot policies serve as the foundation of many validated robotic systems, which encapsulate extensive embodied knowledge. However, they often lack the semantic awareness characteristic of foundation models, and replacing them entirely is impractical in many situations due to high costs and the loss of accumulated knowledge. To address this gap, we introduce GUIDES, a lightweight framework that augments pre-trained policies with semantic guidance from foundation models without requiring architectural redesign. GUIDES employs a fine-tuned vision-language model (Instructor) to generate contextual instructions, which are encoded by an auxiliary module into guidance embeddings. These embeddings are injected into the policy's latent space, allowing the legacy model to adapt to this new semantic input through brief, targeted fine-tuning. For inference-time robustness, a large language model-based Reflector monitors the Instructor's confidence and, when confidence is low, initiates a reasoning loop that analyzes execution history, retrieves relevant examples, and augments the VLM's context to refine subsequent actions. Extensive validation in the RoboCasa simulation environment across diverse policy architectures shows consistent and substantial improvements in task success rates. Real-world deployment on a UR5 robot further demonstrates that GUIDES enhances motion precision for critical sub-tasks such as grasping. Overall, GUIDES offers a practical and resource-efficient pathway to upgrade, rather than replace, validated robot policies.
comment: 8 pages, 4 figures, Accepted by IEEE IROS 2025 Workshop WIR-M
☆ Collaborative Assembly Policy Learning of a Sightless Robot
This paper explores a physical human-robot collaboration (pHRC) task involving the joint insertion of a board into a frame by a sightless robot and a human operator. While admittance control is commonly used in pHRC tasks, it can be challenging to measure the force/torque applied by the human for accurate human intent estimation, limiting the robot's ability to assist in the collaborative task. Other methods that attempt to solve pHRC tasks using reinforcement learning (RL) are also unsuitable for the board-insertion task due to its safety constraints and sparse rewards. Therefore, we propose a novel RL approach that utilizes a human-designed admittance controller to facilitate more active robot behavior and reduce human effort. Through simulation and real-world experiments, we demonstrate that our approach outperforms admittance control in terms of success rate and task completion time. Additionally, we observed a significant reduction in measured force/torque when using our proposed approach compared to admittance control. The video of the experiments is available at https://youtu.be/va07Gw6YIog.
comment: Accepted by IEEE ROBIO 2025
☆ Periodic Skill Discovery NeurIPS 2025
Unsupervised skill discovery in reinforcement learning (RL) aims to learn diverse behaviors without relying on external rewards. However, current methods often overlook the periodic nature of learned skills, focusing instead on increasing the mutual dependence between states and skills or maximizing the distance traveled in latent space. Considering that many robotic tasks -- particularly those involving locomotion -- require periodic behaviors across varying timescales, the ability to discover diverse periodic skills is essential. Motivated by this, we propose Periodic Skill Discovery (PSD), a framework that discovers periodic behaviors in an unsupervised manner. The key idea of PSD is to train an encoder that maps states to a circular latent space, thereby naturally encoding periodicity in the latent representation. By capturing temporal distance, PSD can effectively learn skills with diverse periods in complex robotic tasks, even with pixel-based observations. We further show that these learned skills achieve high performance on downstream tasks such as hurdling. Moreover, integrating PSD with an existing skill discovery method offers more diverse behaviors, thus broadening the agent's repertoire. Our code and demos are available at https://jonghaepark.github.io/psd/
comment: NeurIPS 2025
☆ Learning-based Cooperative Robotic Paper Wrapping: A Unified Control Policy with Residual Force Control
Human-robot cooperation is essential in environments such as warehouses and retail stores, where workers frequently handle deformable objects like paper, bags, and fabrics. Coordinating robotic actions with human assistance remains difficult due to the unpredictable dynamics of deformable materials and the need for adaptive force control. To explore this challenge, we focus on the task of gift wrapping, which exemplifies a long-horizon manipulation problem involving precise folding, controlled creasing, and secure fixation of paper. Success is achieved when the robot completes the sequence to produce a neatly wrapped package with clean folds and no tears. We propose a learning-based framework that integrates a high-level task planner powered by a large language model (LLM) with a low-level hybrid imitation learning (IL) and reinforcement learning (RL) policy. At its core is a Sub-task Aware Robotic Transformer (START) that learns a unified policy from human demonstrations. The key novelty lies in capturing long-range temporal dependencies across the full wrapping sequence within a single model. Unlike vanilla Action Chunking with Transformer (ACT), typically applied to short tasks, our method introduces sub-task IDs that provide explicit temporal grounding. This enables robust performance across the entire wrapping process and supports flexible execution, as the policy learns sub-goals rather than merely replicating motion sequences. Our framework achieves a 97% success rate on real-world wrapping tasks. We show that the unified transformer-based policy reduces the need for specialized models, allows controlled human supervision, and effectively bridges high-level intent with the fine-grained force control required for deformable object manipulation.
☆ Optimizing Earth-Moon Transfer and Cislunar Navigation: Integrating Low-Energy Trajectories, AI Techniques and GNSS-R Technologies
The rapid growth of cislunar activities, including lunar landings, the Lunar Gateway, and in-space refueling stations, requires advances in cost-efficient trajectory design and reliable integration of navigation and remote sensing. Traditional Earth-Moon transfers suffer from rigid launch windows and high propellant demands, while Earth-based GNSS systems provide little to no coverage beyond geostationary orbit. This limits autonomy and environmental awareness in cislunar space. This review compares four major transfer strategies by evaluating velocity requirements, flight durations, and fuel efficiency, and by identifying their suitability for both crewed and robotic missions. The emerging role of artificial intelligence and machine learning is highlighted: convolutional neural networks support automated crater recognition and digital terrain model generation, while deep reinforcement learning enables adaptive trajectory refinement during descent and landing to reduce risk and decision latency. The study also examines how GNSS-Reflectometry and advanced Positioning, Navigation, and Timing architectures can extend navigation capabilities beyond current limits. GNSS-R can act as a bistatic radar for mapping lunar ice, soil properties, and surface topography, while PNT systems support autonomous rendezvous, Lagrange point station-keeping, and coordinated satellite swarm operations. Combining these developments establishes a scalable framework for sustainable cislunar exploration and long-term human and robotic presence.
☆ Learning Natural and Robust Hexapod Locomotion over Complex Terrains via Motion Priors based on Deep Reinforcement Learning
Multi-legged robots offer enhanced stability to navigate complex terrains with their multiple legs interacting with the environment. However, how to effectively coordinate the multiple legs in a larger action exploration space to generate natural and robust movements is a key issue. In this paper, we introduce a motion prior-based approach, successfully applying deep reinforcement learning algorithms to a real hexapod robot. We generate a dataset of optimized motion priors, and train an adversarial discriminator based on the priors to guide the hexapod robot to learn natural gaits. The learned policy is then successfully transferred to a real hexapod robot, and demonstrate natural gait patterns and remarkable robustness without visual information in complex terrains. This is the first time that a reinforcement learning controller has been used to achieve complex terrain walking on a real hexapod robot.
☆ SENT Map -- Semantically Enhanced Topological Maps with Foundation Models ICRA 2025
We introduce SENT-Map, a semantically enhanced topological map for representing indoor environments, designed to support autonomous navigation and manipulation by leveraging advancements in foundational models (FMs). Through representing the environment in a JSON text format, we enable semantic information to be added and edited in a format that both humans and FMs understand, while grounding the robot to existing nodes during planning to avoid infeasible states during deployment. Our proposed framework employs a two stage approach, first mapping the environment alongside an operator with a Vision-FM, then using the SENT-Map representation alongside a natural-language query within an FM for planning. Our experimental results show that semantic-enhancement enables even small locally-deployable FMs to successfully plan over indoor environments.
comment: Accepted at ICRA 2025 Workshop on Foundation Models and Neuro-Symbolic AI for Robotics
☆ Investigating Robot Control Policy Learning for Autonomous X-ray-guided Spine Procedures
Imitation learning-based robot control policies are enjoying renewed interest in video-based robotics. However, it remains unclear whether this approach applies to X-ray-guided procedures, such as spine instrumentation. This is because interpretation of multi-view X-rays is complex. We examine opportunities and challenges for imitation policy learning in bi-plane-guided cannula insertion. We develop an in silico sandbox for scalable, automated simulation of X-ray-guided spine procedures with a high degree of realism. We curate a dataset of correct trajectories and corresponding bi-planar X-ray sequences that emulate the stepwise alignment of providers. We then train imitation learning policies for planning and open-loop control that iteratively align a cannula solely based on visual information. This precisely controlled setup offers insights into limitations and capabilities of this method. Our policy succeeded on the first attempt in 68.5% of cases, maintaining safe intra-pedicular trajectories across diverse vertebral levels. The policy generalized to complex anatomy, including fractures, and remained robust to varied initializations. Rollouts on real bi-planar X-rays further suggest that the model can produce plausible trajectories, despite training exclusively in simulation. While these preliminary results are promising, we also identify limitations, especially in entry point precision. Full closed-look control will require additional considerations around how to provide sufficiently frequent feedback. With more robust priors and domain knowledge, such models may provide a foundation for future efforts toward lightweight and CT-free robotic intra-operative spinal navigation.
♻ ☆ RoboRAN: A Unified Robotics Framework for Reinforcement Learning-Based Autonomous Navigation
Autonomous robots must navigate and operate in diverse environments, from terrestrial and aquatic settings to aerial and space domains. While Reinforcement Learning (RL) has shown promise in training policies for specific autonomous robots, existing frameworks and benchmarks are often constrained to unique platforms, limiting generalization and fair comparisons across different mobility systems. In this paper, we present a multi-domain framework for training, evaluating and deploying RL-based navigation policies across diverse robotic platforms and operational environments. Our work presents four key contributions: (1) a scalable and modular framework, facilitating seamless robot-task interchangeability and reproducible training pipelines; (2) sim-to-real transfer demonstrated through real-world experiments with multiple robots, including a satellite robotic simulator, an unmanned surface vessel, and a wheeled ground vehicle; (3) the release of the first open-source API for deploying Isaac Lab-trained policies to real robots, enabling lightweight inference and rapid field validation; and (4) uniform tasks and metrics for cross-medium evaluation, through a unified evaluation testbed to assess performance of navigation tasks in diverse operational conditions (aquatic, terrestrial and space). By ensuring consistency between simulation and real-world deployment, RoboRAN lowers the barrier to developing adaptable RL-based navigation strategies. Its modular design enables straightforward integration of new robots and tasks through predefined templates, fostering reproducibility and extension to diverse domains. To support the community, we release RoboRAN as open-source.
comment: Accepted at Transactions on Machine Learning Research (TMLR)
♻ ☆ Depth Matters: Multimodal RGB-D Perception for Robust Autonomous Agents ICRA 2025
Autonomous agents that rely purely on perception to make real-time control decisions require efficient and robust architectures. In this work, we demonstrate that augmenting RGB input with depth information significantly enhances our agents' ability to predict steering commands compared to using RGB alone. We benchmark lightweight recurrent controllers that leverage the fused RGB-D features for sequential decision-making. To train our models, we collect high-quality data using a small-scale autonomous car controlled by an expert driver via a physical steering wheel, capturing varying levels of steering difficulty. Our models were successfully deployed on real hardware and inherently avoided dynamic and static obstacles, under out-of-distribution conditions. Specifically, our findings reveal that the early fusion of depth data results in a highly robust controller, which remains effective even with frame drops and increased noise levels, without compromising the network's focus on the task.
comment: Submitted to ICRA 2025
♻ ☆ An explicit construction of Kaleidocycles by elliptic theta functions
We consider the configuration space of ordered points on the two-dimensional sphere that satisfy a specific system of quadratic equations. We construct periodic orbits in this configuration space using elliptic theta functions and show that they simultaneously satisfy semi-discrete analogues of mKdV and sine-Gordon equations. The configuration space we investigate corresponds to the state space of a linkage mechanism known as the Kaleidocycle, and the constructed orbits describe the characteristic motion of the Kaleidocycle. A key consequence of our construction is the proof that Kaleidocycles exist for any number of tetrahedra greater than five. Our approach is founded on the relationship between the deformation of spatial curves and integrable systems, offering an intriguing example where an integrable system is explicitly solved to generate an orbit in the space of real solutions to polynomial equations defined by geometric constraints.
♻ ☆ Autonomous Robotic Drilling System for Mice Cranial Window Creation
Robotic assistance for experimental manipulation in the life sciences is expected to enable favorable outcomes, regardless of the skill of the scientist. Experimental specimens in the life sciences are subject to individual variability and hence require intricate algorithms for successful autonomous robotic control. As a use case, we are studying the cranial window creation in mice. This operation requires the removal of an 8-mm circular patch of the skull, which is approximately 300 um thick, but the shape and thickness of the mouse skull significantly varies depending on the strain of the mouse, sex, and age. In this work, we develop an autonomous robotic drilling system with no offline planning, consisting of a trajectory planner with execution-time feedback with drilling completion level recognition based on image and force information. In the experiments, we first evaluate the image-and-force-based drilling completion level recognition by comparing it with other state-of-the-art deep learning image processing methods and conduct an ablation study in eggshell drilling to evaluate the impact of each module on system performance. Finally, the system performance is further evaluated in postmortem mice, achieving a success rate of 70% (14/20 trials) with an average drilling time of 9.3 min.
comment: 14 pages, 11 figures, accepted on T-ASE 2025
♻ ☆ Toward Humanoid Brain-Body Co-design: Joint Optimization of Control and Morphology for Fall Recovery
Humanoid robots represent a central frontier in embodied intelligence, as their anthropomorphic form enables natural deployment in humans' workspace. Brain-body co-design for humanoids presents a promising approach to realizing this potential by jointly optimizing control policies and physical morphology. Within this context, fall recovery emerges as a critical capability. It not only enhances safety and resilience but also integrates naturally with locomotion systems, thereby advancing the autonomy of humanoids. In this paper, we propose RoboCraft, a scalable humanoid co-design framework for fall recovery that iteratively improves performance through the coupled updates of control policy and morphology. A shared policy pretrained across multiple designs is progressively finetuned on high-performing morphologies, enabling efficient adaptation without retraining from scratch. Concurrently, morphology search is guided by human-inspired priors and optimization algorithms, supported by a priority buffer that balances reevaluation of promising candidates with the exploration of novel designs. Experiments show that RoboCraft achieves an average performance gain of 44.55% on seven public humanoid robots, with morphology optimization drives at least 40% of improvements in co-designing four humanoid robots, underscoring the critical role of humanoid co-design.
♻ ☆ Mastering Contact-rich Tasks by Combining Soft and Rigid Robotics with Imitation Learning
Soft robots have the potential to revolutionize the use of robotic systems with their capability of establishing safe, robust, and adaptable interactions with their environment, but their precise control remains challenging. In contrast, traditional rigid robots offer high accuracy and repeatability but lack the flexibility of soft robots. We argue that combining these characteristics in a hybrid robotic platform can significantly enhance overall capabilities. This work presents a novel hybrid robotic platform that integrates a rigid manipulator with a fully developed soft arm. This system is equipped with the intelligence necessary to perform flexible and generalizable tasks through imitation learning autonomously. The physical softness and machine learning enable our platform to achieve highly generalizable skills, while the rigid components ensure precision and repeatability.
comment: Update with additional results and experiments
♻ ☆ Augmented Reality for RObots (ARRO): Pointing Visuomotor Policies Towards Visual Robustness
Visuomotor policies trained on human expert demonstrations have recently shown strong performance across a wide range of robotic manipulation tasks. However, these policies remain highly sensitive to domain shifts stemming from background or robot embodiment changes, which limits their generalization capabilities. In this paper, we present ARRO, a novel visual representation that leverages zero-shot open-vocabulary segmentation and object detection models to efficiently mask out task-irrelevant regions of the scene in real time without requiring additional training, modeling of the setup, or camera calibration. By filtering visual distractors and overlaying virtual guides during both training and inference, ARRO improves robustness to scene variations and reduces the need for additional data collection. We extensively evaluate ARRO with Diffusion Policy on a range of tabletop manipulation tasks in both simulation and real-world environments, and further demonstrate its compatibility and effectiveness with generalist robot policies, such as Octo and OpenVLA. Across all settings in our evaluation, ARRO yields consistent performance gains, allows for selective masking to choose between different objects, and shows robustness even to challenging segmentation conditions. Videos showcasing our results are available at: https://augmented-reality-for-robots.github.io/
♻ ☆ mmE-Loc: Facilitating Accurate Drone Landing with Ultra-High-Frequency Localization
For precise, efficient, and safe drone landings, ground platforms should real-time, accurately locate descending drones and guide them to designated spots. While mmWave sensing combined with cameras improves localization accuracy, lower sampling frequency of traditional frame cameras compared to mmWave radar creates bottlenecks in system throughput. In this work, we upgrade traditional frame camera with event camera, a novel sensor that harmonizes in sampling frequency with mmWave radar within ground platform setup, and introduce mmE-Loc, a high-precision, low-latency ground localization system designed for precise drone landings. To fully exploit the \textit{temporal consistency} and \textit{spatial complementarity} between these two modalities, we propose two innovative modules: \textit{(i)} the Consistency-instructed Collaborative Tracking module, which further leverages the drone's physical knowledge of periodic micro-motions and structure for accurate measurements extraction, and \textit{(ii)} the Graph-informed Adaptive Joint Optimization module, which integrates drone motion information for efficient sensor fusion and drone localization. Real-world experiments conducted in landing scenarios with a drone delivery company demonstrate that mmE-Loc significantly outperforms state-of-the-art methods in both accuracy and latency.
comment: 17 pages, 34 figures. Journal extended version of arXiv:2502.14992
♻ ☆ Decentralized Aerial Manipulation of a Cable-Suspended Load using Multi-Agent Reinforcement Learning
This paper presents the first decentralized method to enable real-world 6-DoF manipulation of a cable-suspended load using a team of Micro-Aerial Vehicles (MAVs). Our method leverages multi-agent reinforcement learning (MARL) to train an outer-loop control policy for each MAV. Unlike state-of-the-art controllers that utilize a centralized scheme, our policy does not require global states, inter-MAV communications, nor neighboring MAV information. Instead, agents communicate implicitly through load pose observations alone, which enables high scalability and flexibility. It also significantly reduces computing costs during inference time, enabling onboard deployment of the policy. In addition, we introduce a new action space design for the MAVs using linear acceleration and body rates. This choice, combined with a robust low-level controller, enables reliable sim-to-real transfer despite significant uncertainties caused by cable tension during dynamic 3D motion. We validate our method in various real-world experiments, including full-pose control under load model uncertainties, showing setpoint tracking performance comparable to the state-of-the-art centralized method. We also demonstrate cooperation amongst agents with heterogeneous control policies, and robustness to the complete in-flight loss of one MAV. Videos of experiments: https://autonomousrobots.nl/paper_websites/aerial-manipulation-marl
♻ ☆ Thor: Towards Human-Level Whole-Body Reactions for Intense Contact-Rich Environments
Humanoids hold great potential for service, industrial, and rescue applications, in which robots must sustain whole-body stability while performing intense, contact-rich interactions with the environment. However, enabling humanoids to generate human-like, adaptive responses under such conditions remains a major challenge. To address this, we propose Thor, a humanoid framework for human-level whole-body reactions in contact-rich environments. Based on the robot's force analysis, we design a force-adaptive torso-tilt (FAT2) reward function to encourage humanoids to exhibit human-like responses during force-interaction tasks. To mitigate the high-dimensional challenges of humanoid control, Thor introduces a reinforcement learning architecture that decouples the upper body, waist, and lower body. Each component shares global observations of the whole body and jointly updates its parameters. Finally, we deploy Thor on the Unitree G1, and it substantially outperforms baselines in force-interaction tasks. Specifically, the robot achieves a peak pulling force of 167.7 N (approximately 48% of the G1's body weight) when moving backward and 145.5 N when moving forward, representing improvements of 68.9% and 74.7%, respectively, compared with the best-performing baseline. Moreover, Thor is capable of pulling a loaded rack (130 N) and opening a fire door with one hand (60 N). These results highlight Thor's effectiveness in enhancing humanoid force-interaction capabilities.
♻ ☆ Multi-Agent Reinforcement Learning for Autonomous Multi-Satellite Earth Observation: A Realistic Case Study
The exponential growth of Low Earth Orbit (LEO) satellites has revolutionised Earth Observation (EO) missions, addressing challenges in climate monitoring, disaster management, and more. However, autonomous coordination in multi-satellite systems remains a fundamental challenge. Traditional optimisation approaches struggle to handle the real-time decision-making demands of dynamic EO missions, necessitating the use of Reinforcement Learning (RL) and Multi-Agent Reinforcement Learning (MARL). In this paper, we investigate RL-based autonomous EO mission planning by modelling single-satellite operations and extending to multi-satellite constellations using MARL frameworks. We address key challenges, including energy and data storage limitations, uncertainties in satellite observations, and the complexities of decentralised coordination under partial observability. By leveraging a near-realistic satellite simulation environment, we evaluate the training stability and performance of state-of-the-art MARL algorithms, including PPO, IPPO, MAPPO, and HAPPO. Our results demonstrate that MARL can effectively balance imaging and resource management while addressing non-stationarity and reward interdependency in multi-satellite coordination. The insights gained from this study provide a foundation for autonomous satellite operations, offering practical guidelines for improving policy learning in decentralised EO missions.
♻ ☆ AURA: Autonomous Upskilling with Retrieval-Augmented Agents
Designing reinforcement learning curricula for agile robots traditionally requires extensive manual tuning of reward functions, environment randomizations, and training configurations. We introduce AURA (Autonomous Upskilling with Retrieval-Augmented Agents), a schema-validated curriculum reinforcement learning (RL) framework that leverages Large Language Models (LLMs) as autonomous designers of multi-stage curricula. AURA transforms user prompts into YAML workflows that encode full reward functions, domain randomization strategies, and training configurations. All files are statically validated before any GPU time is used, ensuring efficient and reliable execution. A retrieval-augmented feedback loop allows specialized LLM agents to design, execute, and refine curriculum stages based on prior training results stored in a vector database, enabling continual improvement over time. Quantitative experiments show that AURA consistently outperforms LLM-guided baselines in generation success rate, humanoid locomotion, and manipulation tasks. Ablation studies highlight the importance of schema validation and retrieval for curriculum quality. AURA successfully trains end-to-end policies directly from user prompts and deploys them zero-shot on a custom humanoid robot in multiple environments - capabilities that did not exist previously with manually designed controllers. By abstracting the complexity of curriculum design, AURA enables scalable and adaptive policy learning pipelines that would be complex to construct by hand. Project page: https://aura-research.org/
♻ ☆ Hybrid Dynamics Modeling and Trajectory Planning for a Cable-Trailer System with a Quadruped Robot
Inspired by sled-pulling dogs in transportation, we present a cable-trailer integrated with a quadruped robot system. The motion planning of this system faces challenges due to the interactions between the cable's state transitions, the trailer's nonholonomic constraints, and the system's underactuation. To address these challenges, we first develop a hybrid dynamics model that captures the cable's taut and slack states. A search algorithm is then introduced to compute a suboptimal trajectory while incorporating mode transitions. Additionally, we propose a novel collision avoidance constraint based on geometric polygons to formulate the trajectory optimization problem for the hybrid system. The proposed method is implemented on a Unitree A1 quadruped robot with a customized cable-trailer and validated through experiments. The real system demonstrates both agile and safe motion with cable mode transitions.
comment: 8 pages, 8 figures, Accept by RA-L 2025
♻ ☆ Modeling Elastic-Body Dynamics of Robotic Fish Using a Variational Framework
Fish-inspired aquatic robots are gaining increasing attention in marine robot communities due to their high swimming speeds and efficient propulsion enabled by flexible bodies that generate undulatory motions. To support the design optimization and control of such systems, accurate, interpretable, and computationally tractable modeling of the underlying swimming dynamics is indispensable. In this letter, we present a full-body dynamics model for motor-actuated robotic fish, rigorously derived from Hamilton's principle. The model captures the continuously distributed elasticity of a deformable fish body undergoing large deformations and incorporates fluid-structure coupling effects, enabling self-propelled motion without prescribing kinematics. Preliminary open-loop simulations examine how actuation frequency and body stiffness influence the swimming speed and energy efficiency of the robotic fish. Closed-loop simulations further assess how stiffness distribution impacts the controller's velocity-tracking performance and energy efficiency. The results demonstrate the model's potential for performance evaluation and control optimization of soft robotic swimmers when stiffness is treated as a design variable.
comment: Under review at IEEE Robotics and Automation Letters (RA-L)
♻ ☆ Scalable Multi-Robot Motion Planning Using Workspace Guidance-Informed Hypergraphs
In this work, we propose a method for multiple mobile robot motion planning that efficiently plans for robot teams up to 128 robots (an order of magnitude larger than existing state-of-the-art methods) in congested settings with narrow passages in the environment. We achieve this improvement in scalability by extending the state-of-the-art Decomposable State Space Hypergraph (DaSH) multi-robot planning framework to support mobile robot motion planning in congested environments. This is a problem that DaSH cannot be directly applied to because it lacks a highly structured, easily discretizable task space and features kinodynamic constraints. We accomplish this by exploiting knowledge about the workspace topology to limit exploration of the planning space and through modifying DaSH's conflict resolution scheme. This guidance captures when coordination between robots is necessary, allowing us to decompose the intractably large multi-robot search space while limiting risk of inter-robot conflicts by composing relevant robot groups together while planning.
comment: This work has been submitted for review
♻ ☆ XRoboToolkit: A Cross-Platform Framework for Robot Teleoperation
The rapid advancement of Vision-Language-Action models has created an urgent need for large-scale, high-quality robot demonstration datasets. Although teleoperation is the predominant method for data collection, current approaches suffer from limited scalability, complex setup procedures, and suboptimal data quality. This paper presents XRoboToolkit, a cross-platform framework for extended reality based robot teleoperation built on the OpenXR standard. The system features low-latency stereoscopic visual feedback, optimization-based inverse kinematics, and support for diverse tracking modalities including head, controller, hand, and auxiliary motion trackers. XRoboToolkit's modular architecture enables seamless integration across robotic platforms and simulation environments, spanning precision manipulators, mobile robots, and dexterous hands. We demonstrate the framework's effectiveness through precision manipulation tasks and validate data quality by training VLA models that exhibit robust autonomous performance.
comment: 6 pages, 6 figures, accepted at The 2026 IEEE/SICE International Symposium on System Integration, project link: http://xr-robotics.github.io/
♻ ☆ Stability analysis through folds: An end-loaded elastica with a lever arm
Many physical systems can be modelled as parameter-dependent variational problems. In numerous cases, multiple equilibria co-exist, requiring the evaluation of their stability, and the monitoring of transitions between them. Generally, the stability characteristics of the equilibria change near folds in the parameter space. The direction of stability changes is embedded in a specific projection of the solutions, known as distinguished bifurcation diagrams. In this article, we identify such projections for variational problems characterized by fixed-free ends - a class of problems frequently encountered in mechanics. Using these diagrams, we study an Elastica subject to an end load applied through a rigid lever arm. Several instances of snap-back instability are reported, along with their dependence on system parameters through numerical examples. These findings have potential applications in the design of soft robot arms and other actuator designs.
comment: 22 pages, 12 figures
♻ ☆ Swarmodroid & AMPy: Reconfigurable Bristle-Bots and Software Package for Robotic Active Matter Studies
Large assemblies of extremely simple robots capable only of basic motion activities (like propelling forward or self-rotating) are often applied to study swarming behavior or implement various phenomena characteristic of active matter composed of non-equilibrium particles that convert their energy to a directed motion. As a result, a great abundance of compact swarm robots have been developed. The simplest are bristle-bots that self-propel via converting their vibration with the help of elastic bristles. However, many platforms are optimized for a certain class of studies, are not always made open-source, or have limited customization potential. To address these issues, we develop the open-source Swarmodroid 1.0 platform based on bristle-bots with reconfigurable 3D printed bodies and simple electronics that possess external control of motion velocity and demonstrate basic capabilities of trajectory adjustment. Then, we perform a detailed analysis of individual Swarmodroids' motion characteristics and their kinematics. In addition, we introduce the AMPy software package in Python that features OpenCV-based extraction of robotic swarm kinematics accompanied by the evaluation of key physical quantities describing the collective dynamics. Finally, we discuss potential applications as well as further directions for fundamental studies and Swarmodroid 1.0 platform development.
comment: 17 pages, 6 figures, 1 table + Supplementary Information. Comments are welcome
♻ ☆ A Framework for Human-Reason-Aligned Trajectory Evaluation in Automated Vehicles
One major challenge for the adoption and acceptance of automated vehicles (AVs) is ensuring that they can make sound decisions in everyday situations that involve ethical tension. Much attention has focused on rare, high-stakes dilemmas such as trolley problems. Yet similar conflicts arise in routine driving when human considerations, such as legality, efficiency, and comfort, come into conflict. Current AV planning systems typically rely on rigid rules, which struggle to balance these competing considerations and often lead to behaviour that misaligns with human expectations. This paper introduces a reasons-based trajectory evaluation framework that operationalises the tracking condition of Meaningful Human Control (MHC). The framework represents human agents reasons (e.g., regulatory compliance) as quantifiable functions and evaluates how well candidate trajectories align with them. It assigns adjustable weights to agent priorities and includes a balance function to discourage excluding any agent. To demonstrate the approach, we use a real-world-inspired overtaking scenario, which highlights tensions between compliance, efficiency, and comfort. Our results show that different trajectories emerge as preferable depending on how agents reasons are weighted, and small shifts in priorities can lead to discrete changes in the selected action. This demonstrates that everyday ethical decisions in AV driving are highly sensitive to the weights assigned to the reasons of different human agents.
comment: This version incorporates revisions based on peer-review feedback from a new submission. The work has been accepted and is being prepared for publication
Robotics 54
☆ TWIST2: Scalable, Portable, and Holistic Humanoid Data Collection System
Large-scale data has driven breakthroughs in robotics, from language models to vision-language-action models in bimanual manipulation. However, humanoid robotics lacks equally effective data collection frameworks. Existing humanoid teleoperation systems either use decoupled control or depend on expensive motion capture setups. We introduce TWIST2, a portable, mocap-free humanoid teleoperation and data collection system that preserves full whole-body control while advancing scalability. Our system leverages PICO4U VR for obtaining real-time whole-body human motions, with a custom 2-DoF robot neck (cost around $250) for egocentric vision, enabling holistic human-to-humanoid control. We demonstrate long-horizon dexterous and mobile humanoid skills and we can collect 100 demonstrations in 15 minutes with an almost 100% success rate. Building on this pipeline, we propose a hierarchical visuomotor policy framework that autonomously controls the full humanoid body based on egocentric vision. Our visuomotor policy successfully demonstrates whole-body dexterous manipulation and dynamic kicking tasks. The entire system is fully reproducible and open-sourced at https://yanjieze.com/TWIST2 . Our collected dataset is also open-sourced at https://twist-data.github.io .
comment: Website: https://yanjieze.com/TWIST2
☆ XR-1: Towards Versatile Vision-Language-Action Models via Learning Unified Vision-Motion Representations
Recent progress in large-scale robotic datasets and vision-language models (VLMs) has advanced research on vision-language-action (VLA) models. However, existing VLA models still face two fundamental challenges: (i) producing precise low-level actions from high-dimensional observations, (ii) bridging domain gaps across heterogeneous data sources, including diverse robot embodiments and human demonstrations. Existing methods often encode latent variables from either visual dynamics or robotic actions to guide policy learning, but they fail to fully exploit the complementary multi-modal knowledge present in large-scale, heterogeneous datasets. In this work, we present X Robotic Model 1 (XR-1), a novel framework for versatile and scalable VLA learning across diverse robots, tasks, and environments. XR-1 introduces the \emph{Unified Vision-Motion Codes (UVMC)}, a discrete latent representation learned via a dual-branch VQ-VAE that jointly encodes visual dynamics and robotic motion. UVMC addresses these challenges by (i) serving as an intermediate representation between the observations and actions, and (ii) aligning multimodal dynamic information from heterogeneous data sources to capture complementary knowledge. To effectively exploit UVMC, we propose a three-stage training paradigm: (i) self-supervised UVMC learning, (ii) UVMC-guided pretraining on large-scale cross-embodiment robotic datasets, and (iii) task-specific post-training. We validate XR-1 through extensive real-world experiments with more than 14,000 rollouts on six different robot embodiments, spanning over 120 diverse manipulation tasks. XR-1 consistently outperforms state-of-the-art baselines such as $\pi_{0.5}$, $\pi_0$, RDT, UniVLA, and GR00T-N1.5 while demonstrating strong generalization to novel objects, background variations, distractors, and illumination changes. Our project is at https://xr-1-vla.github.io/.
☆ Non-Contact Manipulation of Induced Magnetic Dipoles
Extending the field of magnetic manipulation to conductive, non-magnetic objects opens the door for a wide array of applications previously limited to hard or soft magnetic materials. Of particular interest is the recycling of space debris through the use of oscillating magnetic fields, which represent a cache of raw materials in an environment particularly suited to the low forces generated from inductive magnetic manipulation. Building upon previous work that demonstrated 3D open-loop position control by leveraging the opposing dipole moment created from induced eddy currents, this work demonstrates closed-loop position control of a semi-buoyant aluminum sphere in lab tests, and the efficacy of varying methods for force inversion is explored. The closed-loop methods represent a critical first step towards wider applications for 3-DOF position control of induced magnetic dipoles.
☆ Many-vs-Many Missile Guidance via Virtual Targets
This paper presents a novel approach to many-vs-many missile guidance using virtual targets (VTs) generated by a Normalizing Flows-based trajectory predictor. Rather than assigning n interceptors directly to m physical targets through conventional weapon target assignment algorithms, we propose a centralized strategy that constructs n VT trajectories representing probabilistic predictions of maneuvering target behavior. Each interceptor is guided toward its assigned VT using Zero-Effort-Miss guidance during midcourse flight, transitioning to Proportional Navigation guidance for terminal interception. This approach treats many-vs-many engagements as many-vs-distribution scenarios, exploiting numerical superiority (n > m) by distributing interceptors across diverse trajectory hypotheses rather than pursuing identical deterministic predictions. Monte Carlo simulations across various target-interceptor configurations (1-6 targets, 1-8 interceptors) demonstrate that the VT method matches or exceeds baseline straight-line prediction performance by 0-4.1% when n = m, with improvements increasing to 5.8-14.4% when n > m. The results confirm that probabilistic VTs enable effective exploitation of numerical superiority, significantly increasing interception probability in many-vs-many scenarios.
comment: will be submitted to Journal of Guidance, Control, and Dynamics as Technical Note
☆ Keeping it Local, Tiny and Real: Automated Report Generation on Edge Computing Devices for Mechatronic-Based Cognitive Systems
Recent advancements in Deep Learning enable hardware-based cognitive systems, that is, mechatronic systems in general and robotics in particular with integrated Artificial Intelligence, to interact with dynamic and unstructured environments. While the results are impressive, the application of such systems to critical tasks like autonomous driving as well as service and care robotics necessitate the evaluation of large amount of heterogeneous data. Automated report generation for Mobile Robotics can play a crucial role in facilitating the evaluation and acceptance of such systems in various domains. In this paper, we propose a pipeline for generating automated reports in natural language utilizing various multi-modal sensors that solely relies on local models capable of being deployed on edge computing devices, thus preserving the privacy of all actors involved and eliminating the need for external services. In particular, we evaluate our implementation on a diverse dataset spanning multiple domains including indoor, outdoor and urban environments, providing quantitative as well as qualitative evaluation results. Various generated example reports and other supplementary materials are available via a public repository.
comment: 6 pages, 4 figures, 1 table; accepted for MECATRONICS-REM 2025 International Conference, PARIS, FRANCE December 3-5 2025
☆ Dexterous Robotic Piano Playing at Scale
Endowing robot hands with human-level dexterity has been a long-standing goal in robotics. Bimanual robotic piano playing represents a particularly challenging task: it is high-dimensional, contact-rich, and requires fast, precise control. We present OmniPianist, the first agent capable of performing nearly one thousand music pieces via scalable, human-demonstration-free learning. Our approach is built on three core components. First, we introduce an automatic fingering strategy based on Optimal Transport (OT), allowing the agent to autonomously discover efficient piano-playing strategies from scratch without demonstrations. Second, we conduct large-scale Reinforcement Learning (RL) by training more than 2,000 agents, each specialized in distinct music pieces, and aggregate their experience into a dataset named RP1M++, consisting of over one million trajectories for robotic piano playing. Finally, we employ a Flow Matching Transformer to leverage RP1M++ through large-scale imitation learning, resulting in the OmniPianist agent capable of performing a wide range of musical pieces. Extensive experiments and ablation studies highlight the effectiveness and scalability of our approach, advancing dexterous robotic piano playing at scale.
☆ From the Laboratory to Real-World Application: Evaluating Zero-Shot Scene Interpretation on Edge Devices for Mobile Robotics
Video Understanding, Scene Interpretation and Commonsense Reasoning are highly challenging tasks enabling the interpretation of visual information, allowing agents to perceive, interact with and make rational decisions in its environment. Large Language Models (LLMs) and Visual Language Models (VLMs) have shown remarkable advancements in these areas in recent years, enabling domain-specific applications as well as zero-shot open vocabulary tasks, combining multiple domains. However, the required computational complexity poses challenges for their application on edge devices and in the context of Mobile Robotics, especially considering the trade-off between accuracy and inference time. In this paper, we investigate the capabilities of state-of-the-art VLMs for the task of Scene Interpretation and Action Recognition, with special regard to small VLMs capable of being deployed to edge devices in the context of Mobile Robotics. The proposed pipeline is evaluated on a diverse dataset consisting of various real-world cityscape, on-campus and indoor scenarios. The experimental evaluation discusses the potential of these small models on edge devices, with particular emphasis on challenges, weaknesses, inherent model biases and the application of the gained information. Supplementary material is provided via the following repository: https://datahub.rz.rptu.de/hstr-csrl-public/publications/scene-interpretation-on-edge-devices/
comment: 15 pages, 6 figures, 1 table; accepted for AI-2025 Forty-fifth SGAI International Conference on Artificial Intelligence CAMBRIDGE, ENGLAND 16-18 DECEMBER 2025
☆ Synthetic Crop-Weed Image Generation and its Impact on Model Generalization
Precise semantic segmentation of crops and weeds is necessary for agricultural weeding robots. However, training deep learning models requires large annotated datasets, which are costly to obtain in real fields. Synthetic data can reduce this burden, but the gap between simulated and real images remains a challenge. In this paper, we present a pipeline for procedural generation of synthetic crop-weed images using Blender, producing annotated datasets under diverse conditions of plant growth, weed density, lighting, and camera angle. We benchmark several state-of-the-art segmentation models on synthetic and real datasets and analyze their cross-domain generalization. Our results show that training on synthetic images leads to a sim-to-real gap of 10%, surpassing previous state-of-the-art methods. Moreover, synthetic data demonstrates good generalization properties, outperforming real datasets in cross-domain scenarios. These findings highlight the potential of synthetic agricultural datasets and support hybrid strategies for more efficient model training.
☆ Whole-body motion planning and safety-critical control for aerial manipulation
Aerial manipulation combines the maneuverability of multirotors with the dexterity of robotic arms to perform complex tasks in cluttered spaces. Yet planning safe, dynamically feasible trajectories remains difficult due to whole-body collision avoidance and the conservativeness of common geometric abstractions such as bounding boxes or ellipsoids. We present a whole-body motion planning and safety-critical control framework for aerial manipulators built on superquadrics (SQs). Using an SQ-plus-proxy representation, we model both the vehicle and obstacles with differentiable, geometry-accurate surfaces. Leveraging this representation, we introduce a maximum-clearance planner that fuses Voronoi diagrams with an equilibrium-manifold formulation to generate smooth, collision-aware trajectories. We further design a safety-critical controller that jointly enforces thrust limits and collision avoidance via high-order control barrier functions. In simulation, our approach outperforms sampling-based planners in cluttered environments, producing faster, safer, and smoother trajectories and exceeding ellipsoid-based baselines in geometric fidelity. Actual experiments on a physical aerial-manipulation platform confirm feasibility and robustness, demonstrating consistent performance across simulation and hardware settings. The video can be found at https://youtu.be/hQYKwrWf1Ak.
comment: Submitted to 2026 IFAC World Congress with the Journal option (MECHATRONICS)
☆ Cycle-Sync: Robust Global Camera Pose Estimation through Enhanced Cycle-Consistent Synchronization NeurIPS 2025
We introduce Cycle-Sync, a robust and global framework for estimating camera poses (both rotations and locations). Our core innovation is a location solver that adapts message-passing least squares (MPLS) -- originally developed for group synchronization -- to camera location estimation. We modify MPLS to emphasize cycle-consistent information, redefine cycle consistencies using estimated distances from previous iterations, and incorporate a Welsch-type robust loss. We establish the strongest known deterministic exact-recovery guarantee for camera location estimation, showing that cycle consistency alone -- without access to inter-camera distances -- suffices to achieve the lowest sample complexity currently known. To further enhance robustness, we introduce a plug-and-play outlier rejection module inspired by robust subspace recovery, and we fully integrate cycle consistency into MPLS for rotation synchronization. Our global approach avoids the need for bundle adjustment. Experiments on synthetic and real datasets show that Cycle-Sync consistently outperforms leading pose estimators, including full structure-from-motion pipelines with bundle adjustment.
comment: NeurIPS 2025 spotlight paper
☆ ZJUNlict Extended Team Description Paper 2025
This paper presents the ZJUNlict team's work over the past year, covering both hardware and software advancements. In the hardware domain, the integration of an IMU into the v2023 robot was completed to enhance posture accuracy and angular velocity planning. On the software side, key modules were optimized, including the strategy and CUDA modules, with significant improvements in decision making efficiency, ball pursuit prediction, and ball possession prediction to adapt to high-tempo game dynamics.
☆ SuckTac: Camera-based Tactile Sucker for Unstructured Surface Perception and Interaction
Suckers are significant for robots in picking, transferring, manipulation and locomotion on diverse surfaces. However, most of the existing suckers lack high-fidelity perceptual and tactile sensing, which impedes them from resolving the fine-grained geometric features and interaction status of the target surface. This limits their robust performance with irregular objects and in complex, unstructured environments. Inspired by the adaptive structure and high-performance sensory capabilities of cephalopod suckers, in this paper, we propose a novel, intelligent sucker, named SuckTac, that integrates a camera-based tactile sensor directly within its optimized structure to provide high-density perception and robust suction. Specifically, through joint structure design and optimization and based on a multi-material integrated casting technique, a camera and light source are embedded into the sucker, which enables in-situ, high-density perception of fine details like surface shape, texture and roughness. To further enhance robustness and adaptability, the sucker's mechanical design is also optimized by refining its profile, adding a compliant lip, and incorporating surface microstructure. Extensive experiments, including challenging tasks such as robotic cloth manipulation and soft mobile robot inspection, demonstrate the superior performance and broad applicability of the proposed system.
☆ LACY: A Vision-Language Model-based Language-Action Cycle for Self-Improving Robotic Manipulation
Learning generalizable policies for robotic manipulation increasingly relies on large-scale models that map language instructions to actions (L2A). However, this one-way paradigm often produces policies that execute tasks without deeper contextual understanding, limiting their ability to generalize or explain their behavior. We argue that the complementary skill of mapping actions back to language (A2L) is essential for developing more holistic grounding. An agent capable of both acting and explaining its actions can form richer internal representations and unlock new paradigms for self-supervised learning. We introduce LACY (Language-Action Cycle), a unified framework that learns such bidirectional mappings within a single vision-language model. LACY is jointly trained on three synergistic tasks: generating parameterized actions from language (L2A), explaining observed actions in language (A2L), and verifying semantic consistency between two language descriptions (L2C). This enables a self-improving cycle that autonomously generates and filters new training data through an active augmentation strategy targeting low-confidence cases, thereby improving the model without additional human labels. Experiments on pick-and-place tasks in both simulation and the real world show that LACY improves task success rates by 56.46% on average and yields more robust language-action grounding for robotic manipulation. Project page: https://vla2026.github.io/LACY/
comment: Preprint. Project page: https://vla2026.github.io/LACY/
☆ A Quantitative Comparison of Centralised and Distributed Reinforcement Learning-Based Control for Soft Robotic Arms
This paper presents a quantitative comparison between centralised and distributed multi-agent reinforcement learning (MARL) architectures for controlling a soft robotic arm modelled as a Cosserat rod in simulation. Using PyElastica and the OpenAI Gym interface, we train both a global Proximal Policy Optimisation (PPO) controller and a Multi-Agent PPO (MAPPO) under identical budgets. Both approaches are based on the arm having $n$ number of controlled sections. The study systematically varies $n$ and evaluates the performance of the arm to reach a fixed target in three scenarios: default baseline condition, recovery from external disturbance, and adaptation to actuator failure. Quantitative metrics used for the evaluation are mean action magnitude, mean final distance, mean episode length, and success rate. The results show that there are no significant benefits of the distributed policy when the number of controlled sections $n\le4$. In very simple systems, when $n\le2$, the centralised policy outperforms the distributed one. When $n$ increases to $4< n\le 12$, the distributed policy shows a high sample efficiency. In these systems, distributed policy promotes a stronger success rate, resilience, and robustness under local observability and yields faster convergence given the same sample size. However, centralised policies achieve much higher time efficiency during training as it takes much less time to train the same size of samples. These findings highlight the trade-offs between centralised and distributed policy in reinforcement learning-based control for soft robotic systems and provide actionable design guidance for future sim-to-real transfer in soft rod-like manipulators.
comment: 7 pages, 4 figures, 2 tables, submitted to RoboSoft 2026
☆ Kinematic and Ergonomic Design of a Robotic Arm for Precision Laparoscopic Surgery
Robotic assistance in minimally invasive surgery can greatly enhance surgical precision and reduce surgeon fatigue. This paper presents a focused investigation on the kinematic and ergonomic design principles for a laparoscopic surgical robotic arm aimed at high-precision tasks. We propose a 7-degree-of-freedom (7-DOF) robotic arm system that incorporates a remote center of motion (RCM) at the instrument insertion point and ergonomic considerations to improve surgeon interaction. The design is implemented on a general-purpose robotic platform, and a series of simulated surgical tasks were performed to evaluate targeting accuracy, task efficiency, and surgeon comfort compared to conventional manual laparoscopy. Experimental results demonstrate that the optimized robotic design achieves significantly improved targeting accuracy (error reduced by over 50%) and shorter task completion times, while substantially lowering operator muscle strain and discomfort. These findings validate the importance of kinematic optimization (such as added articulations and tremor filtering) and human-centered ergonomic design in enhancing the performance of robot-assisted surgery. The insights from this work can guide the development of next-generation surgical robots that improve surgical outcomes and ergonomics for the operating team.
☆ Text to Robotic Assembly of Multi Component Objects using 3D Generative AI and Vision Language Models NeurIPS 2025
Advances in 3D generative AI have enabled the creation of physical objects from text prompts, but challenges remain in creating objects involving multiple component types. We present a pipeline that integrates 3D generative AI with vision-language models (VLMs) to enable the robotic assembly of multi-component objects from natural language. Our method leverages VLMs for zero-shot, multi-modal reasoning about geometry and functionality to decompose AI-generated meshes into multi-component 3D models using predefined structural and panel components. We demonstrate that a VLM is capable of determining which mesh regions need panel components in addition to structural components, based on object functionality. Evaluation across test objects shows that users preferred the VLM-generated assignments 90.6% of the time, compared to 59.4% for rule-based and 2.5% for random assignment. Lastly, the system allows users to refine component assignments through conversational feedback, enabling greater human control and agency in making physical objects with generative AI and robotics.
comment: Accepted to NeurIPS 2025, Conference on Neural Information Processing Systems, Creative AI Track
☆ Census-Based Population Autonomy For Distributed Robotic Teaming
Collaborating teams of robots show promise due in their ability to complete missions more efficiently and with improved robustness, attributes that are particularly useful for systems operating in marine environments. A key issue is how to model, analyze, and design these multi-robot systems to realize the full benefits of collaboration, a challenging task since the domain of multi-robot autonomy encompasses both collective and individual behaviors. This paper introduces a layered model of multi-robot autonomy that uses the principle of census, or a weighted count of the inputs from neighbors, for collective decision-making about teaming, coupled with multi-objective behavior optimization for individual decision-making about actions. The census component is expressed as a nonlinear opinion dynamics model and the multi-objective behavior optimization is accomplished using interval programming. This model can be reduced to recover foundational algorithms in distributed optimization and control, while the full model enables new types of collective behaviors that are useful in real-world scenarios. To illustrate these points, a new method for distributed optimization of subgroup allocation is introduced where robots use a gradient descent algorithm to minimize portions of the cost functions that are locally known, while being influenced by the opinion states from neighbors to account for the unobserved costs. With this method the group can collectively use the information contained in the Hessian matrix of the total global cost. The utility of this model is experimentally validated in three categorically different experiments with fleets of autonomous surface vehicles: an adaptive sampling scenario, a high value unit protection scenario, and a competitive game of capture the flag.
comment: 16 pages, 17 figures
☆ 3D Cal: An Open-Source Software Library for Calibrating Tactile Sensors
Tactile sensing plays a key role in enabling dexterous and reliable robotic manipulation, but realizing this capability requires substantial calibration to convert raw sensor readings into physically meaningful quantities. Despite its near-universal necessity, the calibration process remains ad hoc and labor-intensive. Here, we introduce \libname{}, an open-source library that transforms a low-cost 3D printer into an automated probing device capable of generating large volumes of labeled training data for tactile sensor calibration. We demonstrate the utility of \libname{} by calibrating two commercially available vision-based tactile sensors, DIGIT and GelSight Mini, to reconstruct high-quality depth maps using the collected data and a custom convolutional neural network. In addition, we perform a data ablation study to determine how much data is needed for accurate calibration, providing practical guidelines for researchers working with these specific sensors, and we benchmark the trained models on previously unseen objects to evaluate calibration accuracy and generalization performance. By automating tactile sensor calibration, \libname{} can accelerate tactile sensing research, simplify sensor deployment, and promote the practical integration of tactile sensing in robotic platforms.
☆ WorldPlanner: Monte Carlo Tree Search and MPC with Action-Conditioned Visual World Models
Robots must understand their environment from raw sensory inputs and reason about the consequences of their actions in it to solve complex tasks. Behavior Cloning (BC) leverages task-specific human demonstrations to learn this knowledge as end-to-end policies. However, these policies are difficult to transfer to new tasks, and generating training data is challenging because it requires careful demonstrations and frequent environment resets. In contrast to such policy-based view, in this paper we take a model-based approach where we collect a few hours of unstructured easy-to-collect play data to learn an action-conditioned visual world model, a diffusion-based action sampler, and optionally a reward model. The world model -- in combination with the action sampler and a reward model -- is then used to optimize long sequences of actions with a Monte Carlo Tree Search (MCTS) planner. The resulting plans are executed on the robot via a zeroth-order Model Predictive Controller (MPC). We show that the action sampler mitigates hallucinations of the world model during planning and validate our approach on 3 real-world robotic tasks with varying levels of planning and modeling complexity. Our experiments support the hypothesis that planning leads to a significant improvement over BC baselines on a standard manipulation test environment.
☆ A Collaborative Reasoning Framework for Anomaly Diagnostics in Underwater Robotics ICRA 2026
The safe deployment of autonomous systems in safety-critical settings requires a paradigm that combines human expertise with AI-driven analysis, especially when anomalies are unforeseen. We introduce AURA (Autonomous Resilience Agent), a collaborative framework for anomaly and fault diagnostics in robotics. AURA integrates large language models (LLMs), a high-fidelity digital twin (DT), and human-in-the-loop interaction to detect and respond to anomalous behavior in real time. The architecture uses two agents with clear roles: (i) a low-level State Anomaly Characterization Agent that monitors telemetry and converts signals into a structured natural-language problem description, and (ii) a high-level Diagnostic Reasoning Agent that conducts a knowledge-grounded dialogue with an operator to identify root causes, drawing on external sources. Human-validated diagnoses are then converted into new training examples that refine the low-level perceptual model. This feedback loop progressively distills expert knowledge into the AI, transforming it from a static tool into an adaptive partner. We describe the framework's operating principles and provide a concrete implementation, establishing a pattern for trustworthy, continually improving human-robot teams.
comment: Paper was submitted for ICRA 2026
☆ Comprehensive Assessment of LiDAR Evaluation Metrics: A Comparative Study Using Simulated and Real Data
For developing safe Autonomous Driving Systems (ADS), rigorous testing is required before they are deemed safe for road deployments. Since comprehensive conventional physical testing is impractical due to cost and safety concerns, Virtual Testing Environments (VTE) can be adopted as an alternative. Comparing VTE-generated sensor outputs against their real-world analogues can be a strong indication that the VTE accurately represents reality. Correspondingly, this work explores a comprehensive experimental approach to finding evaluation metrics suitable for comparing real-world and simulated LiDAR scans. The metrics were tested in terms of sensitivity and accuracy with different noise, density, distortion, sensor orientation, and channel settings. From comparing the metrics, we found that Density Aware Chamfer Distance (DCD) works best across all cases. In the second step of the research, a Virtual Testing Environment was generated using real LiDAR scan data. The data was collected in a controlled environment with only static objects using an instrumented vehicle equipped with LiDAR, IMU and cameras. Simulated LiDAR scans were generated from the VTEs using the same pose as real LiDAR scans. The simulated and LiDAR scans were compared in terms of model perception and geometric similarity. Actual and simulated LiDAR scans have a similar semantic segmentation output with a mIoU of 21\% with corrected intensity and an average density aware chamfer distance (DCD) of 0.63. This indicates a slight difference in the geometric properties of simulated and real LiDAR scans and a significant difference between model outputs. During the comparison, density-aware chamfer distance was found to be the most correlated among the metrics with perception methods.
☆ EvtSlowTV -- A Large and Diverse Dataset for Event-Based Depth Estimation
Event cameras, with their high dynamic range (HDR) and low latency, offer a promising alternative for robust depth estimation in challenging environments. However, many event-based depth estimation approaches are constrained by small-scale annotated datasets, limiting their generalizability to real-world scenarios. To bridge this gap, we introduce EvtSlowTV, a large-scale event camera dataset curated from publicly available YouTube footage, which contains more than 13B events across various environmental conditions and motions, including seasonal hiking, flying, scenic driving, and underwater exploration. EvtSlowTV is an order of magnitude larger than existing event datasets, providing an unconstrained, naturalistic setting for event-based depth learning. This work shows the suitability of EvtSlowTV for a self-supervised learning framework to capitalise on the HDR potential of raw event streams. We further demonstrate that training with EvtSlowTV enhances the model's ability to generalise to complex scenes and motions. Our approach removes the need for frame-based annotations and preserves the asynchronous nature of event data.
☆ Toward an Agricultural Operational Design Domain: A Framework
The agricultural sector increasingly relies on autonomous systems that operate in complex and variable environments. Unlike on-road applications, agricultural automation integrates driving and working processes, each of which imposes distinct operational constraints. Handling this complexity and ensuring consistency throughout the development and validation processes requires a structured, transparent, and verified description of the environment. However, existing Operational Design Domain (ODD) concepts do not yet address the unique challenges of agricultural applications. Therefore, this work introduces the Agricultural ODD (Ag-ODD) Framework, which can be used to describe and verify the operational boundaries of autonomous agricultural systems. The Ag-ODD Framework consists of three core elements. First, the Ag-ODD description concept, which provides a structured method for unambiguously defining environmental and operational parameters using concepts from ASAM Open ODD and CityGML. Second, the 7-Layer Model derived from the PEGASUS 6-Layer Model, has been extended to include a process layer to capture dynamic agricultural operations. Third, the iterative verification process verifies the Ag-ODD against its corresponding logical scenarios, derived from the 7-Layer Model, to ensure the Ag-ODD's completeness and consistency. Together, these elements provide a consistent approach for creating unambiguous and verifiable Ag-ODD. Demonstrative use cases show how the Ag-ODD Framework can support the standardization and scalability of environmental descriptions for autonomous agricultural systems.
comment: 18 pages, 7 figures, 2 tables
♻ ☆ Using Fiber Optic Bundles to Miniaturize Vision-Based Tactile Sensors
Vision-based tactile sensors have recently become popular due to their combination of low cost, very high spatial resolution, and ease of integration using widely available miniature cameras. The associated field of view and focal length, however, are difficult to package in a human-sized finger. In this paper we employ optical fiber bundles to achieve a form factor that, at 15 mm diameter, is smaller than an average human fingertip. The electronics and camera are also located remotely, further reducing package size. The sensor achieves a spatial resolution of 0.22 mm and a minimum force resolution 5 mN for normal and shear contact forces. With these attributes, the DIGIT Pinki sensor is suitable for applications such as robotic and teleoperated digital palpation. We demonstrate its utility for palpation of the prostate gland and show that it can achieve clinically relevant discrimination of prostate stiffness for phantom and ex vivo tissue.
comment: This work has been submitted to the IEEE for possible publication. The CAD design files of DIGIT Pinki are available at https://github.com/facebookresearch/digit-design
♻ ☆ Tactile Displays Driven by Projected Light
Tactile displays that lend tangible form to digital content could transform computing interactions. However, achieving the resolution, speed, and dynamic range needed for perceptual fidelity remains challenging. We present a tactile display that directly converts projected light into visible tactile patterns via a photomechanical surface populated with millimeter-scale optotactile pixels. The pixels transduce incident light into mechanical displacements through photostimulated thermal gas expansion, yielding millimeter scale displacements with response times of 2 to 100 milliseconds. Employing projected light for power transmission and addressing renders these displays highly scalable. We demonstrate optically driven displays with up to 1,511 addressable pixels -- several times more pixels than any prior tactile display attaining comparable performance. Perceptual studies confirm that these displays can reproduce diverse spatiotemporal tactile patterns with high fidelity. This research establishes a foundation for practical, versatile high-resolution tactile displays driven by light.
♻ ☆ Replicating Human Anatomy with Vision Controlled Jetting -- A Pneumatic Musculoskeletal Hand and Forearm
The functional replication and actuation of complex structures inspired by nature is a longstanding goal for humanity. Creating such complex structures combining soft and rigid features and actuating them with artificial muscles would further our understanding of natural kinematic structures. We printed a biomimetic hand in a single print process comprised of a rigid skeleton, soft joint capsules, tendons, and printed touch sensors. We showed it's actuation using electric motors. In this work, we expand on this work by adding a forearm that is also closely modeled after the human anatomy and replacing the hand's motors with 22 independently controlled pneumatic artificial muscles (PAMs). Our thin, high-strain (up to 30.1%) PAMs match the performance of state-of-the-art artificial muscles at a lower cost. The system showcases human-like dexterity with independent finger movements, demonstrating successful grasping of various objects, ranging from a small, lightweight coin to a large can of 272g in weight. The performance evaluation, based on fingertip and grasping forces along with finger joint range of motion, highlights the system's potential.
♻ ☆ Mobile Robotic Multi-View Photometric Stereo SP
Multi-View Photometric Stereo (MVPS) is a popular method for fine-detailed 3D acquisition of an object from images. Despite its outstanding results on diverse material objects, a typical MVPS experimental setup requires a well-calibrated light source and a monocular camera installed on an immovable base. This restricts the use of MVPS on a movable platform, limiting us from taking MVPS benefits in 3D acquisition for mobile robotics applications. To this end, we introduce a new mobile robotic system for MVPS. While the proposed system brings advantages, it introduces additional algorithmic challenges. Addressing them, in this paper, we further propose an incremental approach for mobile robotic MVPS. Our approach leverages a supervised learning setup to predict per-view surface normal, object depth, and per-pixel uncertainty in model-predicted results. A refined depth map per view is obtained by solving an MVPS-driven optimization problem proposed in this paper. Later, we fuse the refined depth map while tracking the camera pose w.r.t the reference frame to recover globally consistent object 3D geometry. Experimental results show the advantages of our robotic system and algorithm, featuring the local high-frequency surface detail recovery with globally consistent object shape. Our work is beyond any MVPS system yet presented, providing encouraging results on objects with unknown reflectance properties using fewer frames without a tiring calibration and installation process, enabling computationally efficient robotic automation approach to photogrammetry. The proposed approach is nearly 100 times computationally faster than the state-of-the-art MVPS methods such as [1, 2] while maintaining the similar results when tested on subjects taken from the benchmark DiLiGenT MV dataset [3].
comment: Acknowledgment Added. Published at International Society Journal of Photogrammetry and Remote Sensing (ISPRS). 32 pages, 14 Figures, 5 Tables
♻ ☆ FRASA: An End-to-End Reinforcement Learning Agent for Fall Recovery and Stand Up of Humanoid Robots
Humanoid robotics faces significant challenges in achieving stable locomotion and recovering from falls in dynamic environments. Traditional methods, such as Model Predictive Control (MPC) and Key Frame Based (KFB) routines, either require extensive fine-tuning or lack real-time adaptability. This paper introduces FRASA, a Deep Reinforcement Learning (DRL) agent that integrates fall recovery and stand up strategies into a unified framework. Leveraging the Cross-Q algorithm, FRASA significantly reduces training time and offers a versatile recovery strategy that adapts to unpredictable disturbances. Comparative tests on Sigmaban humanoid robots demonstrate FRASA superior performance against the KFB method deployed in the RoboCup 2023 by the Rhoban Team, world champion of the KidSize League.
♻ ☆ Extended Friction Models for the Physics Simulation of Servo Actuators
Accurate physical simulation is crucial for the development and validation of control algorithms in robotic systems. Recent works in Reinforcement Learning (RL) take notably advantage of extensive simulations to produce efficient robot control. State-of-the-art servo actuator models generally fail at capturing the complex friction dynamics of these systems. This limits the transferability of simulated behaviors to real-world applications. In this work, we present extended friction models that allow to more accurately simulate servo actuator dynamics. We propose a comprehensive analysis of various friction models, present a method for identifying model parameters using recorded trajectories from a pendulum test bench, and demonstrate how these models can be integrated into physics engines. The proposed friction models are validated on four distinct servo actuators and tested on 2R manipulators, showing significant improvements in accuracy over the standard Coulomb-Viscous model. Our results highlight the importance of considering advanced friction effects in the simulation of servo actuators to enhance the realism and reliability of robotic simulations.
♻ ☆ Dual-Stream Diffusion for World-Model Augmented Vision-Language-Action Model
Recently, augmenting vision-language-action models (VLAs) with world-models has shown promise in robotic policy learning. However, it remains challenging to jointly predict next-state observations and action sequences because of the inherent difference between the two modalities. To address this, we propose DUal-STream diffusion (DUST), a world-model augmented VLA framework that handles the modality conflict and enhances the performance of VLAs across diverse tasks. Specifically, we propose a multimodal diffusion transformer architecture that explicitly maintains separate modality streams while enabling cross-modal knowledge sharing. In addition, we propose training techniques such as independent noise perturbations for each modality and a decoupled flow matching loss, which enables the model to learn the joint distribution in a bidirectional manner while avoiding the need for a unified latent space. Furthermore, based on the decoupled training framework, we introduce a sampling method where we sample action and vision tokens asynchronously at different rates, which shows improvement through inference-time scaling. Through experiments on simulated benchmarks such as RoboCasa and GR-1, DUST achieves up to 6% gains over a standard VLA baseline and implicit world-modeling methods, with our inference-time scaling approach providing an additional 2-5% gain on success rate. On real-world tasks with the Franka Research 3, DUST outperforms baselines in success rate by 13%, confirming its effectiveness beyond simulation. Lastly, we demonstrate the effectiveness of DUST in large-scale pretraining with action-free videos from BridgeV2, where DUST leads to significant gain when transferred to the RoboCasa benchmark.
comment: 20 pages, 10 figures
♻ ☆ Virtual Target Trajectory Prediction for Stochastic Targets
Trajectory prediction of aerial vehicles is a key requirement in applications ranging from missile guidance to UAV collision avoidance. While most prediction methods assume deterministic target motion, real-world targets often exhibit stochastic behaviors such as evasive maneuvers or random gliding patterns. This paper introduces a probabilistic framework based on Conditional Normalizing Flows (CNFs) to model and predict such stochastic dynamics directly from trajectory data. The learned model generates probability distributions of future target positions conditioned on initial states and dynamic parameters, enabling efficient sampling and exact density evaluation. To provide deterministic surrogates compatible with existing guidance and planning algorithms, sampled trajectories are clustered using a time series k-means approach, yielding a set of representative "virtual target" trajectories. The method is target-agnostic, computationally efficient, and requires only trajectory data for training, making it suitable as a drop-in replacement for deterministic predictors. Simulated scenarios with maneuvering and ballistic targets demonstrate that the proposed approach bridges the gap between deterministic assumptions and stochastic reality, advancing guidance and control algorithms for autonomous vehicles.
comment: Manuscript accepted by Journal of Guidance, Control, and Dynamics
♻ ☆ Genie Envisioner: A Unified World Foundation Platform for Robotic Manipulation
We introduce Genie Envisioner (GE), a unified world foundation platform for robotic manipulation that integrates policy learning, evaluation, and simulation within a single video-generative framework. At its core, GE-Base is a large-scale, instruction-conditioned video diffusion model that captures the spatial, temporal, and semantic dynamics of real-world robotic interactions in a structured latent space. Built upon this foundation, GE-Act maps latent representations to executable action trajectories through a lightweight, flow-matching decoder, enabling precise and generalizable policy inference across diverse embodiments with minimal supervision. To support scalable evaluation and training, GE-Sim serves as an action-conditioned neural simulator, producing high-fidelity rollouts for closed-loop policy development. The platform is further equipped with EWMBench, a standardized benchmark suite measuring visual fidelity, physical consistency, and instruction-action alignment. Together, these components establish Genie Envisioner as a scalable and practical foundation for instruction-driven, general-purpose embodied intelligence. All code, models, and benchmarks will be released publicly.
comment: https://genie-envisioner.github.io/
♻ ☆ Radar-Based Odometry for Low-Speed Driving
We address automotive odometry for low-speed driving and parking, where centimeter-level accuracy is required due to tight spaces and nearby obstacles. Traditional methods using inertial-measurement units and wheel encoders require vehicle-specific calibration, making them costly for consumer-grade vehicles. To overcome this, we propose a radar-based simultaneous localization and mapping (SLAM) approach that fuses inertial and 4D radar measurements. Our approach tightly couples feature positions and Doppler velocities for accurate localization and robust data association. Key contributions include a tightly coupled radar-Doppler extended Kalman filter, multi-radar support and an information-based feature-pruning strategy. Experiments using both proprietary and public datasets demonstrate high-accuracy localization during low-speed driving.
comment: This work has been submitted to the IEEE for possible publication
♻ ☆ No Plan but Everything Under Control: Robustly Solving Sequential Tasks with Dynamically Composed Gradient Descent ICRA25
We introduce a novel gradient-based approach for solving sequential tasks by dynamically adjusting the underlying myopic potential field in response to feedback and the world's regularities. This adjustment implicitly considers subgoals encoded in these regularities, enabling the solution of long sequential tasks, as demonstrated by solving the traditional planning domain of Blocks World - without any planning. Unlike conventional planning methods, our feedback-driven approach adapts to uncertain and dynamic environments, as demonstrated by one hundred real-world trials involving drawer manipulation. These experiments highlight the robustness of our method compared to planning and show how interactive perception and error recovery naturally emerge from gradient descent without explicitly implementing them. This offers a computationally efficient alternative to planning for a variety of sequential tasks, while aligning with observations on biological problem-solving strategies.
comment: Accepted at ICRA25; Supplementary Material under https://www.tu.berlin/robotics/papers/noplan ; 7 pages + 6 figures;
♻ ☆ UniCoD: Enhancing Robot Policy via Unified Continuous and Discrete Representation Learning
Building generalist robot policies that can handle diverse tasks in open-ended environments is a central challenge in robotics. To leverage knowledge from large-scale pretraining, prior work (VLA) has typically built generalist policies either on top of vision-language understanding models (VLMs) or generative models. However, both semantic understanding from vision-language pretraining and visual dynamics modeling from visual-generation pretraining are crucial for embodied robots. Recent unified models of generation and understanding have demonstrated strong capabilities in both comprehension and generation through large-scale pretraining. We posit that robotic policy learning can likewise benefit from the combined strengths of understanding, planning, and continuous future representation learning. Building on this insight, we introduce UniCoD, which acquires the ability to dynamically model high-dimensional visual features through pretraining on over 1M internet-scale instructional manipulation videos. Subsequently, UniCoD is fine-tuned on data collected from the robot embodiment, enabling the learning of mappings from predictive representations to action tokens. Extensive experiments show our approach consistently outperforms baseline methods in terms of 9\% and 12\% across simulation environments and real-world out-of-distribution tasks.
♻ ☆ Unseen from Seen: Rewriting Observation-Instruction Using Foundation Models for Augmenting Vision-Language Navigation
Data scarcity is a long-standing challenge in the Vision-Language Navigation (VLN) field, which extremely hinders the generalization of agents to unseen environments. Previous works primarily rely on additional simulator data or web-collected images/videos to improve the generalization. However, the simulator environments still face limited diversity, and the web-collected data often requires extensive labor to remove the noise. In this paper, we propose a Rewriting-driven AugMentation (RAM) paradigm for VLN, which directly creates the unseen observation-instruction pairs via rewriting human-annotated training data. Benefiting from our rewriting mechanism, new observation-instruction pairs can be obtained in both simulator-free and labor-saving manners to promote generalization. Specifically, we first introduce Object-Enriched Observation Rewriting, where we combine Vision-Language Models (VLMs) and Large Language Models (LLMs) to derive rewritten object-enriched scene descriptions, enabling observation synthesis with diverse objects and spatial layouts via Text-to-Image Generation Models (T2IMs). Then, we propose Observation-Contrast Instruction Rewriting, which generates observation-aligned rewritten instructions by requiring LLMs to reason the difference between original and new observations. We further develop a mixing-then-focusing training strategy with a random observation cropping scheme, effectively enhancing data distribution diversity while suppressing augmentation data noise during training. Experiments on both the discrete environments (R2R, REVERIE, and R4R datasets) and continuous environments (R2R-CE dataset) show the superior performance and impressive generalization ability of our method. Code is available at https://github.com/SaDil13/VLN-RAM.
comment: Accepted by IEEE Transactions on Neural Networks and Learning Systems
♻ ☆ Kinematify: Open-Vocabulary Synthesis of High-DoF Articulated Objects
A deep understanding of kinematic structures and movable components is essential for enabling robots to manipulate objects and model their own articulated forms. Such understanding is captured through articulated objects, which are essential for tasks such as physical simulation, motion planning, and policy learning. However, creating these models, particularly for objects with high degrees of freedom (DoF), remains a significant challenge. Existing methods typically rely on motion sequences or strong assumptions from hand-curated datasets, which hinders scalability. In this paper, we introduce Kinematify, an automated framework that synthesizes articulated objects directly from arbitrary RGB images or textual descriptions. Our method addresses two core challenges: (i) inferring kinematic topologies for high-DoF objects and (ii) estimating joint parameters from static geometry. To achieve this, we combine MCTS search for structural inference with geometry-driven optimization for joint reasoning, producing physically consistent and functionally valid descriptions. We evaluate Kinematify on diverse inputs from both synthetic and real-world environments, demonstrating improvements in registration and kinematic topology accuracy over prior work.
comment: project page: https://sites.google.com/deemos.com/kinematify
♻ ☆ Generative World Models of Tasks: LLM-Driven Hierarchical Scaffolding for Embodied Agents NeurIPS 2025
Recent advances in agent development have focused on scaling model size and raw interaction data, mirroring successes in large language models. However, for complex, long-horizon multi-agent tasks such as robotic soccer, this end-to-end approach often fails due to intractable exploration spaces and sparse rewards. We propose that an effective world model for decision-making must model the world's physics and also its task semantics. A systematic review of 2024 research in low-resource multi-agent soccer reveals a clear trend towards integrating symbolic and hierarchical methods, such as Hierarchical Task Networks (HTNs) and Bayesian Strategy Networks (BSNs), with multi-agent reinforcement learning (MARL). These methods decompose complex goals into manageable subgoals, creating an intrinsic curriculum that shapes agent learning. We formalize this trend into a framework for Hierarchical Task Environments (HTEs), which are essential for bridging the gap between simple, reactive behaviors and sophisticated, strategic team play. Our framework incorporates the use of Large Language Models (LLMs) as generative world models of tasks, capable of dynamically generating this scaffolding. We argue that HTEs provide a mechanism to guide exploration, generate meaningful learning signals, and train agents to internalize hierarchical structure, enabling the development of more capable and general-purpose agents with greater sample efficiency than purely end-to-end approaches.
comment: In the 39th Conference on Neural Information Processing Systems (NeurIPS 2025) Workshop: Embodied World Models for Decision Making (EWM)
♻ ☆ Neural Network Aided Kalman Filtering with Model Predictive Control Enables Robot-Assisted Drone Recovery on a Wavy Surface
Recovering a drone on a disturbed water surface remains a significant challenge in maritime robotics. In this paper, we propose a unified framework for robot-assisted drone recovery on a wavy surface that addresses two major tasks: Firstly, accurate prediction of a moving drone's position under wave-induced disturbances using KalmanNet Plus Plus (KalmanNet++), a Neural Network Aided Kalman Filtering we proposed. Secondly, effective motion planning using the desired position we got for a manipulator via Receding Horizon Model Predictive Control (RHMPC). Specifically, we compared multiple prediction methods and proposed KalmanNet Plus Plus to predict the position of the UAV, thereby obtaining the desired position. The KalmanNet++ predicts the drone's future position 0.1\,s ahead, while the manipulator plans a capture trajectory in real time, thus overcoming not only wave-induced base motions but also limited constraints such as torque constraints and joint constraints. For the system design, we provide a collaborative system, comprising a manipulator subsystem and a UAV subsystem, enables drone lifting and drone recovery. Simulation and real-world experiments using wave-disturbed motion data demonstrate that our approach achieves a high success rate - above 95\% and outperforms conventional baseline methods by up to 10\% in efficiency and 20\% in precision. The results underscore the feasibility and robustness of our system, which achieves state-of-the-art performance and offers a practical solution for maritime drone operations.
comment: 17 pages, 51 figures
♻ ☆ Light Future: Multimodal Action Frame Prediction via InstructPix2Pix WACV 2026
Predicting future motion trajectories is a critical capability across domains such as robotics, autonomous systems, and human activity forecasting, enabling safer and more intelligent decision-making. This paper proposes a novel, efficient, and lightweight approach for robot action prediction, offering significantly reduced computational cost and inference latency compared to conventional video prediction models. Importantly, it pioneers the adaptation of the InstructPix2Pix model for forecasting future visual frames in robotic tasks, extending its utility beyond static image editing. We implement a deep learning-based visual prediction framework that forecasts what a robot will observe 100 frames (10 seconds) into the future, given a current image and a textual instruction. We repurpose and fine-tune the InstructPix2Pix model to accept both visual and textual inputs, enabling multimodal future frame prediction. Experiments on the RoboTWin dataset (generated based on real-world scenarios) demonstrate that our method achieves superior SSIM and PSNR compared to state-of-the-art baselines in robot action prediction tasks. Unlike conventional video prediction models that require multiple input frames, heavy computation, and slow inference latency, our approach only needs a single image and a text prompt as input. This lightweight design enables faster inference, reduced GPU demands, and flexible multimodal control, particularly valuable for applications like robotics and sports motion trajectory analytics, where motion trajectory precision is prioritized over visual fidelity.
comment: 9 pages including appendix, 4 tables, 8 figures, to be submitted to WACV 2026
♻ ☆ Grounded Vision-Language Interpreter for Integrated Task and Motion Planning
While recent advances in vision-language models have accelerated the development of language-guided robot planners, their black-box nature often lacks safety guarantees and interpretability crucial for real-world deployment. Conversely, classical symbolic planners offer rigorous safety verification but require significant expert knowledge for setup. To bridge the current gap, this paper proposes ViLaIn-TAMP, a hybrid planning framework for enabling verifiable, interpretable, and autonomous robot behaviors. ViLaIn-TAMP comprises three main components: (1) a Vision-Language Interpreter (ViLaIn) adapted from previous work that converts multimodal inputs into structured problem specifications, (2) a modular Task and Motion Planning (TAMP) system that grounds these specifications in actionable trajectory sequences through symbolic and geometric constraint reasoning, and (3) a corrective planning (CP) module which receives concrete feedback on failed solution attempts and feed them with constraints back to ViLaIn to refine the specification. We design challenging manipulation tasks in a cooking domain and evaluate our framework. Experimental results demonstrate that ViLaIn-TAMP outperforms a VLM-as-a-planner baseline by 18% in mean success rate, and that adding the CP module boosts mean success rate by 32%.
comment: Project website: https://omron-sinicx.github.io/ViLaIn-TAMP/
♻ ☆ Talk2Event: Grounded Understanding of Dynamic Scenes from Event Cameras NeurIPS 2025
Event cameras offer microsecond-level latency and robustness to motion blur, making them ideal for understanding dynamic environments. Yet, connecting these asynchronous streams to human language remains an open challenge. We introduce Talk2Event, the first large-scale benchmark for language-driven object grounding in event-based perception. Built from real-world driving data, we provide over 30,000 validated referring expressions, each enriched with four grounding attributes -- appearance, status, relation to viewer, and relation to other objects -- bridging spatial, temporal, and relational reasoning. To fully exploit these cues, we propose EventRefer, an attribute-aware grounding framework that dynamically fuses multi-attribute representations through a Mixture of Event-Attribute Experts (MoEE). Our method adapts to different modalities and scene dynamics, achieving consistent gains over state-of-the-art baselines in event-only, frame-only, and event-frame fusion settings. We hope our dataset and approach will establish a foundation for advancing multimodal, temporally-aware, and language-driven perception in real-world robotics and autonomy.
comment: NeurIPS 2025 Spotlight; 43 pages, 17 figures, 16 tables; Project Page at https://talk2event.github.io
♻ ☆ Towards Predicting Any Human Trajectory In Context NeurIPS 2025
Predicting accurate future trajectories of pedestrians is essential for autonomous systems but remains a challenging task due to the need for adaptability in different environments and domains. A common approach involves collecting scenario-specific data and performing fine-tuning via backpropagation. However, the need to fine-tune for each new scenario is often impractical for deployment on edge devices. To address this challenge, we introduce TrajICL, an In-Context Learning (ICL) framework for pedestrian trajectory prediction that enables adaptation without fine-tuning on the scenario-specific data at inference time without requiring weight updates. We propose a spatio-temporal similarity-based example selection (STES) method that selects relevant examples from previously observed trajectories within the same scene by identifying similar motion patterns at corresponding locations. To further refine this selection, we introduce prediction-guided example selection (PG-ES), which selects examples based on both the past trajectory and the predicted future trajectory, rather than relying solely on the past trajectory. This approach allows the model to account for long-term dynamics when selecting examples. Finally, instead of relying on small real-world datasets with limited scenario diversity, we train our model on a large-scale synthetic dataset to enhance its prediction ability by leveraging in-context examples. Extensive experiments demonstrate that TrajICL achieves remarkable adaptation across both in-domain and cross-domain scenarios, outperforming even fine-tuned approaches across multiple public benchmarks. Project Page: https://fujiry0.github.io/TrajICL-project-page/.
comment: NeurIPS 2025
♻ ☆ Rethinking Bimanual Robotic Manipulation: Learning with Decoupled Interaction Framework
Bimanual robotic manipulation is an emerging and critical topic in the robotics community. Previous works primarily rely on integrated control models that take the perceptions and states of both arms as inputs to directly predict their actions. However, we think bimanual manipulation involves not only coordinated tasks but also various uncoordinated tasks that do not require explicit cooperation during execution, such as grasping objects with the closest hand, which integrated control frameworks ignore to consider due to their enforced cooperation in the early inputs. In this paper, we propose a novel decoupled interaction framework that considers the characteristics of different tasks in bimanual manipulation. The key insight of our framework is to assign an independent model to each arm to enhance the learning of uncoordinated tasks, while introducing a selective interaction module that adaptively learns weights from its own arm to improve the learning of coordinated tasks. Extensive experiments on seven tasks in the RoboTwin dataset demonstrate that: (1) Our framework achieves outstanding performance, with a 23.5% boost over the SOTA method. (2) Our framework is flexible and can be seamlessly integrated into existing methods. (3) Our framework can be effectively extended to multi-agent manipulation tasks, achieving a 28% boost over the integrated control SOTA. (4) The performance boost stems from the decoupled design itself, surpassing the SOTA by 16.5% in success rate with only 1/6 of the model size.
comment: 15 pages, 8 figures
♻ ☆ DiffVLA++: Bridging Cognitive Reasoning and End-to-End Driving through Metric-Guided Alignment
Conventional end-to-end (E2E) driving models are effective at generating physically plausible trajectories, but often fail to generalize to long-tail scenarios due to the lack of essential world knowledge to understand and reason about surrounding environments. In contrast, Vision-Language-Action (VLA) models leverage world knowledge to handle challenging cases, but their limited 3D reasoning capability can lead to physically infeasible actions. In this work we introduce DiffVLA++, an enhanced autonomous driving framework that explicitly bridges cognitive reasoning and E2E planning through metric-guided alignment. First, we build a VLA module directly generating semantically grounded driving trajectories. Second, we design an E2E module with a dense trajectory vocabulary that ensures physical feasibility. Third, and most critically, we introduce a metric-guided trajectory scorer that guides and aligns the outputs of the VLA and E2E modules, thereby integrating their complementary strengths. The experiment on the ICCV 2025 Autonomous Grand Challenge leaderboard shows that DiffVLA++ achieves EPDMS of 49.12.
♻ ☆ RoboTron-Mani: All-in-One Multimodal Large Model for Robotic Manipulation
Recently, robotics has advanced significantly through the integration of larger models and large-scale datasets. However, challenges remain in applying these models to 3D spatial interactions and managing data collection costs. To address these issues, we propose the multimodal robotic manipulation model RoboTron-Mani and the comprehensive dataset RoboData. RoboTron-Mani, on one hand, enhances 3D perception through camera parameters and occupancy supervision. On the other hand, it further incorporates Modality-Isolation-Mask and multimodal decoder blocks based on OpenFlamingo, improving modality fusion and fine-grained perception. RoboData integrats several publicly-available datasets, achieving the first fusion of multi-view images, camera parameters, depth maps, actions, and space alignment, which facilitates comprehensive learning from diverse robotic datasets and offers one complete evaluation system. Trained on RoboData, RoboTron-Mani is the first generalist policy that surpasses expert models, enabling simultaneous evaluation of all tasks across multiple datasets, rather than being limited to specific data or task selections. Specifically, RoboTron-Mani boosts manipulation performance by increasing the average sequence length on CALVIN from 1.7 to 3.5, enabling cross-embodiment generalization, and achieving state-of-the-art results on both simulated and real-world datasets.
♻ ☆ Adv-BMT: Bidirectional Motion Transformer for Safety-Critical Traffic Scenario Generation
Scenario-based testing is essential for validating the performance of autonomous driving (AD) systems. However, such testing is limited by the scarcity of long-tailed, safety-critical scenarios in existing datasets collected in the real world. To tackle the data issue, we propose the Adv-BMT framework, which augments real-world scenarios with diverse and realistic adversarial traffic interactions. The core component of Adv-BMT is a bidirectional motion transformer (BMT) model to perform inverse traffic motion predictions, which takes agent information in the last time step of the scenario as input, and reconstructs the traffic in the inverse of chronological order until the initial time step. The Adv-BMT framework is a two-staged pipeline: it first conducts adversarial initializations and then inverse motion predictions. Different from previous work, we do not need any collision data for pretraining, and are able to generate realistic and diverse collision interactions. Our experimental results validate the quality of generated collision scenarios by Adv-BMT: training in our augmented dataset would reduce episode collision rates by 20%. Demo and code are available at: https://metadriverse.github.io/adv-bmt/.
♻ ☆ MetAdv: A Unified and Interactive Adversarial Testing Platform for Autonomous Driving ACM MM 2025
Evaluating and ensuring the adversarial robustness of autonomous driving (AD) systems is a critical and unresolved challenge. This paper introduces MetAdv, a novel adversarial testing platform that enables realistic, dynamic, and interactive evaluation by tightly integrating virtual simulation with physical vehicle feedback. At its core, MetAdv establishes a hybrid virtual-physical sandbox, within which we design a three-layer closed-loop testing environment with dynamic adversarial test evolution. This architecture facilitates end-to-end adversarial evaluation, ranging from high-level unified adversarial generation, through mid-level simulation-based interaction, to low-level execution on physical vehicles. Additionally, MetAdv supports a broad spectrum of AD tasks, algorithmic paradigms (e.g., modular deep learning pipelines, end-to-end learning, vision-language models). It supports flexible 3D vehicle modeling and seamless transitions between simulated and physical environments, with built-in compatibility for commercial platforms such as Apollo and Tesla. A key feature of MetAdv is its human-in-the-loop capability: besides flexible environmental configuration for more customized evaluation, it enables real-time capture of physiological signals and behavioral feedback from drivers, offering new insights into human-machine trust under adversarial conditions. We believe MetAdv can offer a scalable and unified framework for adversarial assessment, paving the way for safer AD.
comment: ACM MM 2025 Most Popular Demo Award
♻ ☆ ROADWork: A Dataset and Benchmark for Learning to Recognize, Observe, Analyze and Drive Through Work Zones ICCV 2025
Perceiving and autonomously navigating through work zones is a challenging and underexplored problem. Open datasets for this long-tailed scenario are scarce. We propose the ROADWork dataset to learn to recognize, observe, analyze, and drive through work zones. State-of-the-art foundation models fail when applied to work zones. Fine-tuning models on our dataset significantly improves perception and navigation in work zones. With ROADWork dataset, we discover new work zone images with higher precision (+32.5%) at a much higher rate (12.8$\times$) around the world. Open-vocabulary methods fail too, whereas fine-tuned detectors improve performance (+32.2 AP). Vision-Language Models (VLMs) struggle to describe work zones, but fine-tuning substantially improves performance (+36.7 SPICE). Beyond fine-tuning, we show the value of simple techniques. Video label propagation provides additional gains (+2.6 AP) for instance segmentation. While reading work zone signs, composing a detector and text spotter via crop-scaling improves performance +14.2% 1-NED). Composing work zone detections to provide context further reduces hallucinations (+3.9 SPICE) in VLMs. We predict navigational goals and compute drivable paths from work zone videos. Incorporating road work semantics ensures 53.6% goals have angular error (AE) < 0.5 (+9.9 %) and 75.3% pathways have AE < 0.5 (+8.1 %).
comment: ICCV 2025 Accepted Paper
♻ ☆ Human-Exoskeleton Kinematic Calibration to Improve Hand Tracking for Dexterous Teleoperation
Hand exoskeletons are critical tools for dexterous teleoperation and immersive manipulation interfaces, but achieving accurate hand tracking remains a challenge due to user-specific anatomical variability and donning inconsistencies. These issues lead to kinematic misalignments that degrade tracking performance and limit applicability in precision tasks. We propose a subject-specific calibration framework for exoskeleton-based hand tracking that estimates virtual link parameters through residual-weighted optimization. A data-driven approach is introduced to empirically tune cost function weights using motion capture ground truth, enabling accurate and consistent calibration across users. Implemented on the Maestro hand exoskeleton with seven healthy participants, the method achieved substantial reductions in joint and fingertip tracking errors across diverse hand geometries. Qualitative visualizations using a Unity-based virtual hand further demonstrate improved motion fidelity. The proposed framework generalizes to exoskeletons with closed-loop kinematics and minimal sensing, laying the foundation for high-fidelity teleoperation and robot learning applications.
comment: 8 pages, 10 figures, 1 supplementary video, submitted to RA-L
♻ ☆ Enhancing Fatigue Detection through Heterogeneous Multi-Source Data Integration and Cross-Domain Modality Imputation
Fatigue detection for human operators plays a key role in safety critical applications such as aviation, mining, and long haul transport. While numerous studies have demonstrated the effectiveness of high fidelity sensors in controlled laboratory environments, their performance often degrades when ported to real world settings due to noise, lighting conditions, and field of view constraints, thereby limiting their practicality. This paper formalizes a deployment oriented setting for real world fatigue detection, where high quality sensors are often unavailable in practical applications. To address this challenge, we propose leveraging knowledge from heterogeneous source domains, including high fidelity sensors that are difficult to deploy in the field but commonly used in controlled environments, to assist fatigue detection in the real world target domain. Building on this idea, we design a heterogeneous and multiple source fatigue detection framework that adaptively utilizes the available modalities in the target domain while exploiting diverse configurations in the source domains through alignment across domains and modality imputation. Our experiments, conducted using a field deployed sensor setup and two publicly available human fatigue datasets, demonstrate the practicality, robustness, and improved generalization of our approach across subjects and domains. The proposed method achieves consistent gains over strong baselines in sensor constrained scenarios.
comment: 4figures,14pages
♻ ☆ NaviTrace: Evaluating Embodied Navigation of Vision-Language Models
Vision-language models demonstrate unprecedented performance and generalization across a wide range of tasks and scenarios. Integrating these foundation models into robotic navigation systems opens pathways toward building general-purpose robots. Yet, evaluating these models' navigation capabilities remains constrained by costly real-world trials, overly simplified simulations, and limited benchmarks. We introduce NaviTrace, a high-quality Visual Question Answering benchmark where a model receives an instruction and embodiment type (human, legged robot, wheeled robot, bicycle) and must output a 2D navigation trace in image space. Across 1000 scenarios and more than 3000 expert traces, we systematically evaluate eight state-of-the-art VLMs using a newly introduced semantic-aware trace score. This metric combines Dynamic Time Warping distance, goal endpoint error, and embodiment-conditioned penalties derived from per-pixel semantics and correlates with human preferences. Our evaluation reveals consistent gap to human performance caused by poor spatial grounding and goal localization. NaviTrace establishes a scalable and reproducible benchmark for real-world robotic navigation. The benchmark and leaderboard can be found at https://leggedrobotics.github.io/navitrace_webpage/.
comment: 9 pages, 6 figures, under review at IEEE conference
♻ ☆ Deep Learning Warm Starts for Trajectory Optimization on the International Space Station
Trajectory optimization is a cornerstone of modern robot autonomy, enabling systems to compute trajectories and controls in real-time while respecting safety and physical constraints. However, it has seen limited usage in spaceflight applications due to its heavy computational demands that exceed the capability of most flight computers. In this work, we provide results on the first in-space demonstration of using machine learning-based warm starts for accelerating trajectory optimization for the Astrobee free-flying robot onboard the International Space Station (ISS). We formulate a data-driven optimal control approach that trains a neural network to learn the structure of the trajectory generation problem being solved using sequential convex programming (SCP). Onboard, this trained neural network predicts solutions for the trajectory generation problem and relies on using the SCP solver to enforce safety constraints for the system. Our trained network reduces the number of solver iterations required for convergence in cases including rotational dynamics by 60% and in cases with obstacles drawn from the training distribution of the warm start model by 50%. This work represents a significant milestone in the use of learning-based control for spaceflight applications and a stepping stone for future advances in the use of machine learning for autonomous guidance, navigation, & control.
comment: Accepted to 2025 International Conference on Space Robotics (iSpaRo). Presented at RSS 2025 Workshop on Space Robotics
♻ ☆ Generalized Nash Equilibrium Solutions in Dynamic Games With Shared Constraints
In dynamic games with shared constraints, Generalized Nash Equilibria (GNE) are often computed using the normalized solution concept, which assumes identical Lagrange multipliers for shared constraints across all players. While widely used, this approach excludes other potentially valuable GNE. This paper presents a novel method based on the Mixed Complementarity Problem (MCP) formulation to compute non-normalized GNE, expanding the solution space. We also propose a systematic approach for selecting the optimal GNE based on predefined criteria, enhancing practical flexibility. Numerical examples illustrate the methods effectiveness, offering an alternative to traditional normalized solutions.
Robotics 77
☆ TACO: Trajectory-Aware Controller Optimization for Quadrotors ICRA 2026
Controller performance in quadrotor trajectory tracking depends heavily on parameter tuning, yet standard approaches often rely on fixed, manually tuned parameters that sacrifice task-specific performance. We present Trajectory-Aware Controller Optimization (TACO), a framework that adapts controller parameters online based on the upcoming reference trajectory and current quadrotor state. TACO employs a learned predictive model and a lightweight optimization scheme to optimize controller gains in real time with respect to a broad class of trajectories, and can also be used to adapt trajectories to improve dynamic feasibility while respecting smoothness constraints. To enable large-scale training, we also introduce a parallelized quadrotor simulator supporting fast data collection on diverse trajectories. Experiments on a variety of trajectory types show that TACO outperforms conventional, static parameter tuning while operating orders of magnitude faster than black-box optimization baselines, enabling practical real-time deployment on a physical quadrotor. Furthermore, we show that adapting trajectories using TACO significantly reduces the tracking error obtained by the quadrotor.
comment: 8 pages, 6 figures. In submission to ICRA 2026
☆ TurboMap: GPU-Accelerated Local Mapping for Visual SLAM ICRA 2026
This paper presents TurboMap, a GPU-accelerated and CPU-optimized local mapping module for visual SLAM systems. We identify key performance bottlenecks in the local mapping process for visual SLAM and address them through targeted GPU and CPU optimizations. Specifically, we offload map point triangulation and fusion to the GPU, accelerate redundant keyframe culling on the CPU, and integrate a GPU-accelerated solver to speed up local bundle adjustment. Our implementation is built on top of ORB-SLAM3 and leverages CUDA for GPU programming. The experimental results show that TurboMap achieves an average speedup of 1.3x in the EuRoC dataset and 1.6x in the TUM-VI dataset in the local mapping module, on both desktop and embedded platforms, while maintaining the accuracy of the original system.
comment: Submitted to ICRA 2026
☆ Path-Coordinated Continual Learning with Neural Tangent Kernel-Justified Plasticity: A Theoretical Framework with Near State-of-the-Art Performance
Catastrophic forgetting is one of the fundamental issues of continual learning because neural networks forget the tasks learned previously when trained on new tasks. The proposed framework is a new path-coordinated framework of continual learning that unites the Neural Tangent Kernel (NTK) theory of principled plasticity bounds, statistical validation by Wilson confidence intervals, and evaluation of path quality by the use of multiple metrics. Experimental evaluation shows an average accuracy of 66.7% at the cost of 23.4% catastrophic forgetting on Split-CIFAR10, a huge improvement over the baseline and competitive performance achieved, which is very close to state-of-the-art results. Further, it is found out that NTK condition numbers are predictive indicators of learning capacity limits, showing the existence of a critical threshold at condition number $>10^{11}$. It is interesting to note that the proposed strategy shows a tendency of lowering forgetting as the sequence of tasks progresses (27% to 18%), which is a system stabilization. The framework validates 80% of discovered paths with a rigorous statistical guarantee and maintains 90-97% retention on intermediate tasks. The core capacity limits of the continual learning environment are determined in the analysis, and actionable insights to enhance the adaptive regularization are offered.
comment: Under review, IEEE Letters
☆ Stein-based Optimization of Sampling Distributions in Model Predictive Path Integral Control
This paper presents a novel method for Model Predictive Path Integral (MPPI) control that optimizes sample generation towards an optimal trajectory through Stein Variational Gradient Descent (SVGD). MPPI is traditionally reliant on randomly sampled trajectories, often by a Gaussian distribution. The result can lead to sample deprivation, under-representing the space of possible trajectories, and yield suboptimal results. Through introducing SVGD updates in between MPPI environment steps, we present Stein-Optimized Path-Integral Inference (SOPPI), an MPPI/SVGD algorithm that can dynamically update noise distributions at runtime to shape a more optimal representation without an excessive increase in computational requirements. We demonstrate the efficacy of our method systems ranging from a Cart-Pole to a two-dimensional bipedal walking task, indicating improved performance above standard MPPI across a range of hyper-parameters and demonstrate feasibility at lower particle counts. We discuss the applicability of this MPPI/SVGD method to higher degree-of-freedom systems, as well as its potential to new developments in state-of-the-art differentiable simulators.
comment: 8 pages, 6 figures
☆ TRACE: Textual Reasoning for Affordance Coordinate Extraction ICCV 2025
Vision-Language Models (VLMs) struggle to translate high-level instructions into the precise spatial affordances required for robotic manipulation. While visual Chain-of-Thought (CoT) methods exist, they are often computationally intensive. In this work, we introduce TRACE (Textual Reasoning for Affordance Coordinate Extraction), a novel methodology that integrates a textual Chain of Reasoning (CoR) into the affordance prediction process. We use this methodology to create the TRACE dataset, a large-scale collection created via an autonomous pipeline that pairs instructions with explicit textual rationales. By fine-tuning a VLM on this data, our model learns to externalize its spatial reasoning before acting. Our experiments show that our TRACE-tuned model achieves state-of-the-art performance, reaching 48.1% accuracy on the primary Where2Place (W2P) benchmark (a 9.6% relative improvement) and 55.0% on the more challenging W2P(h) subset. Crucially, an ablation study demonstrates that performance scales directly with the amount of reasoning data used, confirming the CoR's effectiveness. Furthermore, analysis of the model's attention maps reveals an interpretable reasoning process where focus shifts dynamically across reasoning steps. This work shows that training VLMs to generate a textual CoR is an effective and robust strategy for enhancing the precision, reliability, and interpretability of VLM-based robot control. Our dataset and code are available at https://github.com/jink-ucla/TRACE
comment: ICCV 2025. *Equal contribution. {\dag}Corresponding author
☆ Hybrid Neural Network-Based Indoor Localisation System for Mobile Robots Using CSI Data in a Robotics Simulator
We present a hybrid neural network model for inferring the position of mobile robots using Channel State Information (CSI) data from a Massive MIMO system. By leveraging an existing CSI dataset, our approach integrates a Convolutional Neural Network (CNN) with a Multilayer Perceptron (MLP) to form a Hybrid Neural Network (HyNN) that estimates 2D robot positions. CSI readings are converted into synthetic images using the TINTO tool. The localisation solution is integrated with a robotics simulator, and the Robot Operating System (ROS), which facilitates its evaluation through heterogeneous test cases, and the adoption of state estimators like Kalman filters. Our contributions illustrate the potential of our HyNN model in achieving precise indoor localisation and navigation for mobile robots in complex environments. The study follows, and proposes, a generalisable procedure applicable beyond the specific use case studied, making it adaptable to different scenarios and datasets.
comment: 13 pages, 7 figures. Conference paper (ROBOVIS 2025)
☆ Fractional Diffusion Bridge Models NeurIPS 2025
We present Fractional Diffusion Bridge Models (FDBM), a novel generative diffusion bridge framework driven by an approximation of the rich and non-Markovian fractional Brownian motion (fBM). Real stochastic processes exhibit a degree of memory effects (correlations in time), long-range dependencies, roughness and anomalous diffusion phenomena that are not captured in standard diffusion or bridge modeling due to the use of Brownian motion (BM). As a remedy, leveraging a recent Markovian approximation of fBM (MA-fBM), we construct FDBM that enable tractable inference while preserving the non-Markovian nature of fBM. We prove the existence of a coupling-preserving generative diffusion bridge and leverage it for future state prediction from paired training data. We then extend our formulation to the Schr\"{o}dinger bridge problem and derive a principled loss function to learn the unpaired data translation. We evaluate FDBM on both tasks: predicting future protein conformations from aligned data, and unpaired image translation. In both settings, FDBM achieves superior performance compared to the Brownian baselines, yielding lower root mean squared deviation (RMSD) of C$_\alpha$ atomic positions in protein structure prediction and lower Fr\'echet Inception Distance (FID) in unpaired image translation.
comment: To appear in NeurIPS 2025 proceedings. This version includes post-camera-ready revisions
☆ GenDexHand: Generative Simulation for Dexterous Hands
Data scarcity remains a fundamental bottleneck for embodied intelligence. Existing approaches use large language models (LLMs) to automate gripper-based simulation generation, but they transfer poorly to dexterous manipulation, which demands more specialized environment design. Meanwhile, dexterous manipulation tasks are inherently more difficult due to their higher degrees of freedom. Massively generating feasible and trainable dexterous hand tasks remains an open challenge. To this end, we present GenDexHand, a generative simulation pipeline that autonomously produces diverse robotic tasks and environments for dexterous manipulation. GenDexHand introduces a closed-loop refinement process that adjusts object placements and scales based on vision-language model (VLM) feedback, substantially improving the average quality of generated environments. Each task is further decomposed into sub-tasks to enable sequential reinforcement learning, reducing training time and increasing success rates. Our work provides a viable path toward scalable training of diverse dexterous hand behaviors in embodied intelligence by offering a simulation-based solution to synthetic data generation. Our website: https://winniechen2002.github.io/GenDexHand/.
☆ MOBIUS: A Multi-Modal Bipedal Robot that can Walk, Crawl, Climb, and Roll
This article presents a Multi-Modal Bipedal Intelligent Urban Scout robot (MOBIUS) capable of walking, crawling, climbing, and rolling. MOBIUS features four limbs--two 6-DoF arms with two-finger grippers for manipulation and climbing, and two 4-DoF legs for locomotion--enabling smooth transitions across diverse terrains without reconfiguration. A hybrid control architecture combines reinforcement learning-based locomotion with model-based predictive and admittance control enhanced for safety by a Reference Governor toward compliant contact interactions. A high-level MIQCP planner autonomously selects locomotion modes to balance stability and energy efficiency. Hardware experiments demonstrate robust gait transitions, dynamic climbing, and full-body load support via pinch grasp. Overall, MOBIUS demonstrates the importance of tight integration between morphology, high-level planning, and control to enable mobile loco-manipulation and grasping, substantially expanding its interaction capabilities, workspace, and traversability.
comment: 23 pages, 20 figures. Collaborative work between the Robotics and Mechanisms Laboratory (RoMeLa) and Mitsubishi Electric Research Laboratories (MERL)
☆ Lightweight Learning from Actuation-Space Demonstrations via Flow Matching for Whole-Body Soft Robotic Grasping
Robotic grasping under uncertainty remains a fundamental challenge due to its uncertain and contact-rich nature. Traditional rigid robotic hands, with limited degrees of freedom and compliance, rely on complex model-based and heavy feedback controllers to manage such interactions. Soft robots, by contrast, exhibit embodied mechanical intelligence: their underactuated structures and passive flexibility of their whole body, naturally accommodate uncertain contacts and enable adaptive behaviors. To harness this capability, we propose a lightweight actuation-space learning framework that infers distributional control representations for whole-body soft robotic grasping, directly from deterministic demonstrations using a flow matching model (Rectified Flow),without requiring dense sensing or heavy control loops. Using only 30 demonstrations (less than 8% of the reachable workspace), the learned policy achieves a 97.5% grasp success rate across the whole workspace, generalizes to grasped-object size variations of +-33%, and maintains stable performance when the robot's dynamic response is directly adjusted by scaling the execution time from 20% to 200%. These results demonstrate that actuation-space learning, by leveraging its passive redundant DOFs and flexibility, converts the body's mechanics into functional control intelligence and substantially reduces the burden on central controllers for this uncertain-rich task.
☆ 3EED: Ground Everything Everywhere in 3D NeurIPS 2025
Visual grounding in 3D is the key for embodied agents to localize language-referred objects in open-world environments. However, existing benchmarks are limited to indoor focus, single-platform constraints, and small scale. We introduce 3EED, a multi-platform, multi-modal 3D grounding benchmark featuring RGB and LiDAR data from vehicle, drone, and quadruped platforms. We provide over 128,000 objects and 22,000 validated referring expressions across diverse outdoor scenes -- 10x larger than existing datasets. We develop a scalable annotation pipeline combining vision-language model prompting with human verification to ensure high-quality spatial grounding. To support cross-platform learning, we propose platform-aware normalization and cross-modal alignment techniques, and establish benchmark protocols for in-domain and cross-platform evaluations. Our findings reveal significant performance gaps, highlighting the challenges and opportunities of generalizable 3D grounding. The 3EED dataset and benchmark toolkit are released to advance future research in language-driven 3D embodied perception.
comment: NeurIPS 2025 DB Track; 29 pages, 17 figures, 10 tables; Project Page at https://project-3eed.github.io/
☆ Unified Diffusion VLA: Vision-Language-Action Model via Joint Discrete Denoising Diffusion Process
Vision-language-action (VLA) models aim to understand natural language instructions and visual observations and to execute corresponding actions as an embodied agent. Recent work integrates future images into the understanding-acting loop, yielding unified VLAs that jointly understand, generate, and act -- reading text and images and producing future images and actions. However, these models either rely on external experts for modality unification or treat image generation and action prediction as separate processes, limiting the benefits of direct synergy between these tasks. Our core philosophy is to optimize generation and action jointly through a synchronous denoising process, where the iterative refinement enables actions to evolve from initialization, under constant and sufficient visual guidance. We ground this philosophy in our proposed Unified Diffusion VLA and Joint Discrete Denoising Diffusion Process (JD3P), which is a joint diffusion process that integrates multiple modalities into a single denoising trajectory to serve as the key mechanism enabling understanding, generation, and acting to be intrinsically synergistic. Our model and theory are built on a unified tokenized space of all modalities and a hybrid attention mechanism. We further propose a two-stage training pipeline and several inference-time techniques that optimize performance and efficiency. Our approach achieves state-of-the-art performance on benchmarks such as CALVIN, LIBERO, and SimplerEnv with 4$\times$ faster inference than autoregressive methods, and we demonstrate its effectiveness through in-depth analysis and real-world evaluations. Our project page is available at https://irpn-eai.github.io/UD-VLA.github.io/.
☆ MARS: Multi-Agent Robotic System with Multimodal Large Language Models for Assistive Intelligence
Multimodal large language models (MLLMs) have shown remarkable capabilities in cross-modal understanding and reasoning, offering new opportunities for intelligent assistive systems, yet existing systems still struggle with risk-aware planning, user personalization, and grounding language plans into executable skills in cluttered homes. We introduce MARS - a Multi-Agent Robotic System powered by MLLMs for assistive intelligence and designed for smart home robots supporting people with disabilities. The system integrates four agents: a visual perception agent for extracting semantic and spatial features from environment images, a risk assessment agent for identifying and prioritizing hazards, a planning agent for generating executable action sequences, and an evaluation agent for iterative optimization. By combining multimodal perception with hierarchical multi-agent decision-making, the framework enables adaptive, risk-aware, and personalized assistance in dynamic indoor environments. Experiments on multiple datasets demonstrate the superior overall performance of the proposed system in risk-aware planning and coordinated multi-agent execution compared with state-of-the-art multimodal models. The proposed approach also highlights the potential of collaborative AI for practical assistive scenarios and provides a generalizable methodology for deploying MLLM-enabled multi-agent systems in real-world environments.
comment: 3 figures, 1 table; under review at Multimedia Systems (Springer)
☆ PixelVLA: Advancing Pixel-level Understanding in Vision-Language-Action Model
Vision-Language-Action models (VLAs) are emerging as powerful tools for learning generalizable visuomotor control policies. However, current VLAs are mostly trained on large-scale image-text-action data and remain limited in two key ways: (i) they struggle with pixel-level scene understanding, and (ii) they rely heavily on textual prompts, which reduces their flexibility in real-world settings. To address these challenges, we introduce PixelVLA, the first VLA model designed to support both pixel-level reasoning and multimodal prompting with text and visual inputs. Our approach is built on a new visuomotor instruction tuning framework that integrates a multiscale pixel-aware encoder with a visual prompting encoder. To train PixelVLA effectively, we further propose a two-stage automated annotation pipeline that generates Pixel-160K, a large-scale dataset with pixel-level annotations derived from existing robot data. Experiments on three standard VLA benchmarks and two VLA model variants show that PixelVLA improves manipulation success rates by 10.1%-17.8% over OpenVLA, while requiring only 1.5% of its pretraining cost. These results demonstrate that PixelVLA can be integrated into existing VLAs to enable more accurate, efficient, and versatile robot control in complex environments. The dataset and code will be released as open source.
comment: 17pages,7 figures, 5 tabels
☆ Phy-Tac: Toward Human-Like Grasping via Physics-Conditioned Tactile Goals
Humans naturally grasp objects with minimal level required force for stability, whereas robots often rely on rigid, over-squeezing control. To narrow this gap, we propose a human-inspired physics-conditioned tactile method (Phy-Tac) for force-optimal stable grasping (FOSG) that unifies pose selection, tactile prediction, and force regulation. A physics-based pose selector first identifies feasible contact regions with optimal force distribution based on surface geometry. Then, a physics-conditioned latent diffusion model (Phy-LDM) predicts the tactile imprint under FOSG target. Last, a latent-space LQR controller drives the gripper toward this tactile imprint with minimal actuation, preventing unnecessary compression. Trained on a physics-conditioned tactile dataset covering diverse objects and contact conditions, the proposed Phy-LDM achieves superior tactile prediction accuracy, while the Phy-Tac outperforms fixed-force and GraspNet-based baselines in grasp stability and force efficiency. Experiments on classical robotic platforms demonstrate force-efficient and adaptive manipulation that bridges the gap between robotic and human grasping.
comment: 9 papges, 10 figures, 3 tables
☆ Discriminately Treating Motion Components Evolves Joint Depth and Ego-Motion Learning
Unsupervised learning of depth and ego-motion, two fundamental 3D perception tasks, has made significant strides in recent years. However, most methods treat ego-motion as an auxiliary task, either mixing all motion types or excluding depth-independent rotational motions in supervision. Such designs limit the incorporation of strong geometric constraints, reducing reliability and robustness under diverse conditions. This study introduces a discriminative treatment of motion components, leveraging the geometric regularities of their respective rigid flows to benefit both depth and ego-motion estimation. Given consecutive video frames, network outputs first align the optical axes and imaging planes of the source and target cameras. Optical flows between frames are transformed through these alignments, and deviations are quantified to impose geometric constraints individually on each ego-motion component, enabling more targeted refinement. These alignments further reformulate the joint learning process into coaxial and coplanar forms, where depth and each translation component can be mutually derived through closed-form geometric relationships, introducing complementary constraints that improve depth robustness. DiMoDE, a general depth and ego-motion joint learning framework incorporating these designs, achieves state-of-the-art performance on multiple public datasets and a newly collected diverse real-world dataset, particularly under challenging conditions. Our source code will be publicly available at mias.group/DiMoDE upon publication.
comment: 18 pages, 14 figures
☆ SE(3)-PoseFlow: Estimating 6D Pose Distributions for Uncertainty-Aware Robotic Manipulation
Object pose estimation is a fundamental problem in robotics and computer vision, yet it remains challenging due to partial observability, occlusions, and object symmetries, which inevitably lead to pose ambiguity and multiple hypotheses consistent with the same observation. While deterministic deep networks achieve impressive performance under well-constrained conditions, they are often overconfident and fail to capture the multi-modality of the underlying pose distribution. To address these challenges, we propose a novel probabilistic framework that leverages flow matching on the SE(3) manifold for estimating 6D object pose distributions. Unlike existing methods that regress a single deterministic output, our approach models the full pose distribution with a sample-based estimate and enables reasoning about uncertainty in ambiguous cases such as symmetric objects or severe occlusions. We achieve state-of-the-art results on Real275, YCB-V, and LM-O, and demonstrate how our sample-based pose estimates can be leveraged in downstream robotic manipulation tasks such as active perception for disambiguating uncertain viewpoints or guiding grasp synthesis in an uncertainty-aware manner.
☆ Floor Plan-Guided Visual Navigation Incorporating Depth and Directional Cues
Guiding an agent to a specific target in indoor environments based solely on RGB inputs and a floor plan is a promising yet challenging problem. Although existing methods have made significant progress, two challenges remain unresolved. First, the modality gap between egocentric RGB observations and the floor plan hinders the integration of visual and spatial information for both local obstacle avoidance and global planning. Second, accurate localization is critical for navigation performance, but remains challenging at deployment in unseen environments due to the lack of explicit geometric alignment between RGB inputs and floor plans. We propose a novel diffusion-based policy, denoted as GlocDiff, which integrates global path planning from the floor plan with local depth-aware features derived from RGB observations. The floor plan offers explicit global guidance, while the depth features provide implicit geometric cues, collectively enabling precise prediction of optimal navigation directions and robust obstacle avoidance. Moreover, GlocDiff introduces noise perturbation during training to enhance robustness against pose estimation errors, and we find that combining this with a relatively stable VO module during inference results in significantly improved navigation performance. Extensive experiments on the FloNa benchmark demonstrate GlocDiff's efficiency and effectiveness in achieving superior navigation performance, and the success of real-world deployments also highlights its potential for widespread practical applications.
☆ MO-SeGMan: Rearrangement Planning Framework for Multi Objective Sequential and Guided Manipulation in Constrained Environments
In this work, we introduce MO-SeGMan, a Multi-Objective Sequential and Guided Manipulation planner for highly constrained rearrangement problems. MO-SeGMan generates object placement sequences that minimize both replanning per object and robot travel distance while preserving critical dependency structures with a lazy evaluation method. To address highly cluttered, non-monotone scenarios, we propose a Selective Guided Forward Search (SGFS) that efficiently relocates only critical obstacles and to feasible relocation points. Furthermore, we adopt a refinement method for adaptive subgoal selection to eliminate unnecessary pick-and-place actions, thereby improving overall solution quality. Extensive evaluations on nine benchmark rearrangement tasks demonstrate that MO-SeGMan generates feasible motion plans in all cases, consistently achieving faster solution times and superior solution quality compared to the baselines. These results highlight the robustness and scalability of the proposed framework for complex rearrangement planning problems.
comment: 8 pages, 8 figures, website:https://sites.google.com/view/mo-segman/
☆ AERMANI-VLM: Structured Prompting and Reasoning for Aerial Manipulation with Vision Language Models
The rapid progress of vision--language models (VLMs) has sparked growing interest in robotic control, where natural language can express the operation goals while visual feedback links perception to action. However, directly deploying VLM-driven policies on aerial manipulators remains unsafe and unreliable since the generated actions are often inconsistent, hallucination-prone, and dynamically infeasible for flight. In this work, we present AERMANI-VLM, the first framework to adapt pretrained VLMs for aerial manipulation by separating high-level reasoning from low-level control, without any task-specific fine-tuning. Our framework encodes natural language instructions, task context, and safety constraints into a structured prompt that guides the model to generate a step-by-step reasoning trace in natural language. This reasoning output is used to select from a predefined library of discrete, flight-safe skills, ensuring interpretable and temporally consistent execution. By decoupling symbolic reasoning from physical action, AERMANI-VLM mitigates hallucinated commands and prevents unsafe behavior, enabling robust task completion. We validate the framework in both simulation and hardware on diverse multi-step pick-and-place tasks, demonstrating strong generalization to previously unseen commands, objects, and environments.
☆ Designing for Distributed Heterogeneous Modularity: On Software Architecture and Deployment of MoonBots SP
This paper presents the software architecture and deployment strategy behind the MoonBot platform: a modular space robotic system composed of heterogeneous components distributed across multiple computers, networks and ultimately celestial bodies. We introduce a principled approach to distributed, heterogeneous modularity, extending modular robotics beyond physical reconfiguration to software, communication and orchestration. We detail the architecture of our system that integrates component-based design, a data-oriented communication model using ROS2 and Zenoh, and a deployment orchestrator capable of managing complex multi-module assemblies. These abstractions enable dynamic reconfiguration, decentralized control, and seamless collaboration between numerous operators and modules. At the heart of this system lies our open-source Motion Stack software, validated by months of field deployment with self-assembling robots, inter-robot cooperation, and remote operation. Our architecture tackles the significant hurdles of modular robotics by significantly reducing integration and maintenance overhead, while remaining scalable and robust. Although tested with space in mind, we propose generalizable patterns for designing robotic systems that must scale across time, hardware, teams and operational environments.
comment: 6 pages, 8 figures. Accepted at ISPARO 2025
☆ FoldPath: End-to-End Object-Centric Motion Generation via Modulated Implicit Paths IROS 2025
Object-Centric Motion Generation (OCMG) is instrumental in advancing automated manufacturing processes, particularly in domains requiring high-precision expert robotic motions, such as spray painting and welding. To realize effective automation, robust algorithms are essential for generating extended, object-aware trajectories across intricate 3D geometries. However, contemporary OCMG techniques are either based on ad-hoc heuristics or employ learning-based pipelines that are still reliant on sensitive post-processing steps to generate executable paths. We introduce FoldPath, a novel, end-to-end, neural field based method for OCMG. Unlike prior deep learning approaches that predict discrete sequences of end-effector waypoints, FoldPath learns the robot motion as a continuous function, thus implicitly encoding smooth output paths. This paradigm shift eliminates the need for brittle post-processing steps that concatenate and order the predicted discrete waypoints. Particularly, our approach demonstrates superior predictive performance compared to recently proposed learning-based methods, and attains generalization capabilities even in real industrial settings, where only a limited amount of 70 expert samples are provided. We validate FoldPath through comprehensive experiments in a realistic simulation environment and introduce new, rigorous metrics designed to comprehensively evaluate long-horizon robotic paths, thus advancing the OCMG task towards practical maturity.
comment: Accepted at 2025 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2025)
☆ CaRLi-V: Camera-RADAR-LiDAR Point-Wise 3D Velocity Estimation
Accurate point-wise velocity estimation in 3D is crucial for robot interaction with non-rigid, dynamic agents, such as humans, enabling robust performance in path planning, collision avoidance, and object manipulation in dynamic environments. To this end, this paper proposes a novel RADAR, LiDAR, and camera fusion pipeline for point-wise 3D velocity estimation named CaRLi-V. This pipeline leverages raw RADAR measurements to create a novel RADAR representation, the velocity cube, which densely represents radial velocities within the RADAR's field-of-view. By combining the velocity cube for radial velocity extraction, optical flow for tangential velocity estimation, and LiDAR for point-wise range measurements through a closed-form solution, our approach can produce 3D velocity estimates for a dense array of points. Developed as an open-source ROS2 package, CaRLi-V has been field-tested against a custom dataset and proven to produce low velocity error metrics relative to ground truth, enabling point-wise velocity estimation for robotic applications.
☆ EREBUS: End-to-end Robust Event Based Underwater Simulation ICRA
The underwater domain presents a vast array of challenges for roboticists and computer vision researchers alike, such as poor lighting conditions and high dynamic range scenes. In these adverse conditions, traditional vision techniques struggle to adapt and lead to suboptimal performance. Event-based cameras present an attractive solution to this problem, mitigating the issues of traditional cameras by tracking changes in the footage on a frame-by-frame basis. In this paper, we introduce a pipeline which can be used to generate realistic synthetic data of an event-based camera mounted to an AUV (Autonomous Underwater Vehicle) in an underwater environment for training vision models. We demonstrate the effectiveness of our pipeline using the task of rock detection with poor visibility and suspended particulate matter, but the approach can be generalized to other underwater tasks.
comment: Accepted to ICRA AQUA2SIM Workshop 2025, 6 pages, 3 figures, conference paper
☆ CM-LIUW-Odometry: Robust and High-Precision LiDAR-Inertial-UWB-Wheel Odometry for Extreme Degradation Coal Mine Tunnels IROS 2025
Simultaneous Localization and Mapping (SLAM) in large-scale, complex, and GPS-denied underground coal mine environments presents significant challenges. Sensors must contend with abnormal operating conditions: GPS unavailability impedes scene reconstruction and absolute geographic referencing, uneven or slippery terrain degrades wheel odometer accuracy, and long, feature-poor tunnels reduce LiDAR effectiveness. To address these issues, we propose CoalMine-LiDAR-IMU-UWB-Wheel-Odometry (CM-LIUW-Odometry), a multimodal SLAM framework based on the Iterated Error-State Kalman Filter (IESKF). First, LiDAR-inertial odometry is tightly fused with UWB absolute positioning constraints to align the SLAM system with a global coordinate. Next, wheel odometer is integrated through tight coupling, enhanced by nonholonomic constraints (NHC) and vehicle lever arm compensation, to address performance degradation in areas beyond UWB measurement range. Finally, an adaptive motion mode switching mechanism dynamically adjusts the robot's motion mode based on UWB measurement range and environmental degradation levels. Experimental results validate that our method achieves superior accuracy and robustness in real-world underground coal mine scenarios, outperforming state-of-the-art approaches. We open source our code of this work on Github to benefit the robotics community.
comment: Accepted by IROS 2025
☆ Lateral Velocity Model for Vehicle Parking Applications
Automated parking requires accurate localization for quick and precise maneuvering in tight spaces. While the longitudinal velocity can be measured using wheel encoders, the estimation of the lateral velocity remains a key challenge due to the absence of dedicated sensors in consumer-grade vehicles. Existing approaches often rely on simplified vehicle models, such as the zero-slip model, which assumes no lateral velocity at the rear axle. It is well established that this assumption does not hold during low-speed driving and researchers thus introduce additional heuristics to account for differences. In this work, we analyze real-world data from parking scenarios and identify a systematic deviation from the zero-slip assumption. We provide explanations for the observed effects and then propose a lateral velocity model that better captures the lateral dynamics of the vehicle during parking. The model improves estimation accuracy, while relying on only two parameters, making it well-suited for integration into consumer-grade applications.
comment: This manuscript has been submitted to Vehicle System Dynamics for possible publication
☆ Model to Model: Understanding the Venus Flytrap Snapping Mechanism and Transferring it to a 3D-printed Bistable Soft Robotic Demonstrator
The Venus flytrap (Dionaea muscipula) does not only serve as the textbook model for a carnivorous plant, but also has long intrigued both botanists and engineers with its rapidly closing leaf trap. The trap closure is triggered by two consecutive touches of a potential prey, after which the lobes rapidly switch from their concave open-state to their convex close-state and catch the prey within 100-500 ms after being triggered. This transformation from concave to convex is initiated by changes in turgor pressure and the release of stored elastic energy from prestresses in the concave state, which accelerate this movement, leading to inversion of the lobes bi-axial curvature. Possessing two low-energy states, the leaves can be characterized as bistable systems. With our research, we seek to deepen the understanding of Venus flytrap motion mechanics and apply its principles to the design of an artificial bistable lobe actuator. We identified geometrical characteristics, such as dimensional ratios and the thickness gradient in the lobe, and transferred these to two 3D-printed bistable actuator models. One actuator parallels the simulated geometry of a Venus flytrap leaf, the other is a lobe model designed with CAD. Both models display concave-convex bi-stability and snap close. These demonstrators are the first step in the development of an artificial Venus flytrap that mimics the mechanical behavior of the biological model and can be used as a soft fast gripper.
comment: Conference Proceedings Paper Living machines 2025
☆ Design and development of an electronics-free earthworm robot
Soft robotic systems have gained widespread attention due to their inherent flexibility, adaptability, and safety, making them well-suited for varied applications. Among bioinspired designs, earthworm locomotion has been extensively studied for its efficient peristaltic motion, enabling movement in confined and unstructured environments. Existing earthworm-inspired robots primarily utilize pneumatic actuation due to its high force-to-weight ratio and ease of implementation. However, these systems often rely on bulky, power-intensive electronic control units, limiting their practicality. In this work, we present an electronics-free, earthworm-inspired pneumatic robot utilizing a modified Pneumatic Logic Gate (PLG) design. By integrating preconfigured PLG units with bellow actuators, we achieved a plug-and-play style modular system capable of peristaltic locomotion without external electronic components. The proposed design reduces system complexity while maintaining efficient actuation. We characterize the bellow actuators under different operating conditions and evaluate the robots locomotion performance. Our findings demonstrate that the modified PLG-based control system effectively generates peristaltic wave propagation, achieving autonomous motion with minimal deviation. This study serves as a proof of concept for the development of electronics-free, peristaltic soft robots. The proposed system has potential for applications in hazardous environments, where untethered, adaptable locomotion is critical. Future work will focus on further optimizing the robot design and exploring untethered operation using onboard compressed air sources.
comment: Conference Proceedings Paper Living Machines 2025
☆ Thermo-responsive closing and reopening artificial Venus Flytrap utilizing shape memory elastomers
Despite their often perceived static and slow nature, some plants can move faster than the blink of an eye. The rapid snap closure motion of the Venus flytrap (Dionaea muscipula) has long captivated the interest of researchers and engineers alike, serving as a model for plant-inspired soft machines and robots. The translation of the fast snapping closure has inspired the development of various artificial Venus flytrap (AVF) systems. However, translating both the closing and reopening motion of D. muscipula into an autonomous plant inspired soft machine has yet to be achieved. In this study, we present an AVF that autonomously closes and reopens, utilizing novel thermo-responsive UV-curable shape memory materials for soft robotic systems. The life-sized thermo-responsive AVF exhibits closing and reopening motions triggered in a naturally occurring temperature range. The doubly curved trap lobes, built from shape memory polymers, close at 38{\deg}C, while reopening initiates around 45{\deg}C, employing shape memory elastomer strips as antagonistic actuators to facilitate lobe reopening. This work represents the first demonstration of thermo-responsive closing and reopening in an AVF with programmed sequential motion in response to increasing temperature. This approach marks the next step toward autonomously bidirectional moving soft machines/robots.
comment: Conference Proceedings Paper Living Machines 2025
☆ Embodied Cognition Augmented End2End Autonomous Driving
In recent years, vision-based end-to-end autonomous driving has emerged as a new paradigm. However, popular end-to-end approaches typically rely on visual feature extraction networks trained under label supervision. This limited supervision framework restricts the generality and applicability of driving models. In this paper, we propose a novel paradigm termed $E^{3}AD$, which advocates for comparative learning between visual feature extraction networks and the general EEG large model, in order to learn latent human driving cognition for enhancing end-to-end planning. In this work, we collected a cognitive dataset for the mentioned contrastive learning process. Subsequently, we investigated the methods and potential mechanisms for enhancing end-to-end planning with human driving cognition, using popular driving models as baselines on publicly available autonomous driving datasets. Both open-loop and closed-loop tests are conducted for a comprehensive evaluation of planning performance. Experimental results demonstrate that the $E^{3}AD$ paradigm significantly enhances the end-to-end planning performance of baseline models. Ablation studies further validate the contribution of driving cognition and the effectiveness of comparative learning process. To the best of our knowledge, this is the first work to integrate human driving cognition for improving end-to-end autonomous driving planning. It represents an initial attempt to incorporate embodied cognitive data into end-to-end autonomous driving, providing valuable insights for future brain-inspired autonomous driving systems. Our code will be made available at Github
comment: 24 pages,4 pages
☆ RobustVLA: Robustness-Aware Reinforcement Post-Training for Vision-Language-Action Models
Vision-Language-Action (VLA) models have recently emerged as powerful general-purpose policies for robotic manipulation, benefiting from large-scale multi-modal pre-training. However, they often fail to generalize reliably in out-of-distribution deployments, where unavoidable disturbances such as observation noise, sensor errors, or actuation perturbations become prevalent. While recent Reinforcement Learning (RL)-based post-training provides a practical means to adapt pre-trained VLA models, existing methods mainly emphasize reward maximization and overlook robustness to environmental uncertainty. In this work, we introduce RobustVLA, a lightweight online RL post-training method designed to explicitly enhance the resilience of VLA models. Through a systematic robustness analysis, we identify two key regularizations: Jacobian regularization, which mitigates sensitivity to observation noise, and smoothness regularization, which stabilizes policies under action perturbations. Extensive experiments across diverse robotic environments demonstrate that RobustVLA significantly outperforms prior state-of-the-art methods in robustness and reliability. Our results highlight the importance of principled robustness-aware RL post-training as a key step toward improving the reliability and robustness of VLA models.
☆ A High-Speed Capable Spherical Robot
This paper designs a new spherical robot structure capable of supporting high-speed motion at up to 10 m/s. Building upon a single-pendulum-driven spherical robot, the design incorporates a momentum wheel with an axis aligned with the secondary pendulum, creating a novel spherical robot structure. Practical experiments with the physical prototype have demonstrated that this new spherical robot can achieve stable high-speed motion through simple decoupled control, which was unattainable with the original structure. The spherical robot designed for high-speed motion not only increases speed but also significantly enhances obstacle-crossing performance and terrain robustness.
comment: 5 pages
☆ Lyapunov Stability Learning with Nonlinear Control via Inductive Biases
Finding a control Lyapunov function (CLF) in a dynamical system with a controller is an effective way to guarantee stability, which is a crucial issue in safety-concerned applications. Recently, deep learning models representing CLFs have been applied into a learner-verifier framework to identify satisfiable candidates. However, the learner treats Lyapunov conditions as complex constraints for optimisation, which is hard to achieve global convergence. It is also too complicated to implement these Lyapunov conditions for verification. To improve this framework, we treat Lyapunov conditions as inductive biases and design a neural CLF and a CLF-based controller guided by this knowledge. This design enables a stable optimisation process with limited constraints, and allows end-to-end learning of both the CLF and the controller. Our approach achieves a higher convergence rate and larger region of attraction (ROA) in learning the CLF compared to existing methods among abundant experiment cases. We also thoroughly reveal why the success rate decreases with previous methods during learning.
comment: Accepted by IEEE Robio 2025
☆ Contact Map Transfer with Conditional Diffusion Model for Generalizable Dexterous Grasp Generation
Dexterous grasp generation is a fundamental challenge in robotics, requiring both grasp stability and adaptability across diverse objects and tasks. Analytical methods ensure stable grasps but are inefficient and lack task adaptability, while generative approaches improve efficiency and task integration but generalize poorly to unseen objects and tasks due to data limitations. In this paper, we propose a transfer-based framework for dexterous grasp generation, leveraging a conditional diffusion model to transfer high-quality grasps from shape templates to novel objects within the same category. Specifically, we reformulate the grasp transfer problem as the generation of an object contact map, incorporating object shape similarity and task specifications into the diffusion process. To handle complex shape variations, we introduce a dual mapping mechanism, capturing intricate geometric relationship between shape templates and novel objects. Beyond the contact map, we derive two additional object-centric maps, the part map and direction map, to encode finer contact details for more stable grasps. We then develop a cascaded conditional diffusion model framework to jointly transfer these three maps, ensuring their intra-consistency. Finally, we introduce a robust grasp recovery mechanism, identifying reliable contact points and optimizing grasp configurations efficiently. Extensive experiments demonstrate the superiority of our proposed method. Our approach effectively balances grasp quality, generation efficiency, and generalization performance across various tasks. Project homepage: https://cmtdiffusion.github.io/
☆ Design and Fabrication of Origami-Inspired Knitted Fabrics for Soft Robotics
Soft robots employing compliant materials and deformable structures offer great potential for wearable devices that are comfortable and safe for human interaction. However, achieving both structural integrity and compliance for comfort remains a significant challenge. In this study, we present a novel fabrication and design method that combines the advantages of origami structures with the material programmability and wearability of knitted fabrics. We introduce a general design method that translates origami patterns into knit designs by programming both stitch and material patterns. The method creates folds in preferred directions while suppressing unintended buckling and bending by selectively incorporating heat fusible yarn to create rigid panels around compliant creases. We experimentally quantify folding moments and show that stitch patterning enhances folding directionality while the heat fusible yarn (1) keeps geometry consistent by reducing edge curl and (2) prevents out-of-plane deformations by stiffening panels. We demonstrate the framework through the successful reproduction of complex origami tessellations, including Miura-ori, Yoshimura, and Kresling patterns, and present a wearable knitted Kaleidocycle robot capable of locomotion. The combination of structural reconfigurability, material programmability, and potential for manufacturing scalability highlights knitted origami as a promising platform for next-generation wearable robotics.
☆ Improving Needle Penetration via Precise Rotational Insertion Using Iterative Learning Control
Achieving precise control of robotic tool paths is often challenged by inherent system misalignments, unmodeled dynamics, and actuation inaccuracies. This work introduces an Iterative Learning Control (ILC) strategy to enable precise rotational insertion of a tool during robotic surgery, improving penetration efficacy and safety compared to straight insertion tested in subretinal injection. A 4 degree of freedom (DOF) robot manipulator is used, where misalignment of the fourth joint complicates the simple application of needle rotation, motivating an ILC approach that iteratively adjusts joint commands based on positional feedback. The process begins with calibrating the forward kinematics for the chosen surgical tool to achieve higher accuracy, followed by successive ILC iterations guided by Optical Coherence Tomography (OCT) volume scans to measure the error and refine control inputs. Experimental results, tested on subretinal injection tasks on ex vivo pig eyes, show that the optimized trajectory resulted in higher success rates in tissue penetration and subretinal injection compared to straight insertion, demonstrating the effectiveness of ILC in overcoming misalignment challenges. This approach offers potential applications for other high precision robot tasks requiring controlled insertions as well.
comment: 10 pages, 10 figures
☆ Don't Just Search, Understand: Semantic Path Planning Agent for Spherical Tensegrity Robots in Unknown Environments
Endowed with inherent dynamical properties that grant them remarkable ruggedness and adaptability, spherical tensegrity robots stand as prototypical examples of hybrid softrigid designs and excellent mobile platforms. However, path planning for these robots in unknown environments presents a significant challenge, requiring a delicate balance between efficient exploration and robust planning. Traditional path planners, which treat the environment as a geometric grid, often suffer from redundant searches and are prone to failure in complex scenarios due to their lack of semantic understanding. To overcome these limitations, we reframe path planning in unknown environments as a semantic reasoning task. We introduce a Semantic Agent for Tensegrity robots (SATPlanner) driven by a Large Language Model (LLM). SATPlanner leverages high-level environmental comprehension to generate efficient and reliable planning strategies.At the core of SATPlanner is an Adaptive Observation Window mechanism, inspired by the "fast" and "slow" thinking paradigms of LLMs. This mechanism dynamically adjusts the perceptual field of the agent: it narrows for rapid traversal of open spaces and expands to reason about complex obstacle configurations. This allows the agent to construct a semantic belief of the environment, enabling the search space to grow only linearly with the path length (O(L)) while maintaining path quality. We extensively evaluate SATPlanner in 1,000 simulation trials, where it achieves a 100% success rate, outperforming other real-time planning algorithms. Critically, SATPlanner reduces the search space by 37.2% compared to the A* algorithm while achieving comparable, near-optimal path lengths. Finally, the practical feasibility of SATPlanner is validated on a physical spherical tensegrity robot prototype.
comment: 8 pages, 5 figures
☆ High-Precision Surgical Robotic System for Intraocular Procedures
Despite the extensive demonstration of robotic systems for both cataract and vitreoretinal procedures, existing technologies or mechanisms still possess insufficient accuracy, precision, and degrees of freedom for instrument manipulation or potentially automated tool exchange during surgical procedures. A new robotic system that focuses on improving tooltip accuracy, tracking performance, and smooth instrument exchange mechanism is therefore designed and manufactured. Its tooltip accuracy, precision, and mechanical capability of maintaining small incision through remote center of motion were externally evaluated using an optical coherence tomography (OCT) system. Through robot calibration and precise coordinate registration, the accuracy of tooltip positioning was measured to be 0.053$\pm$0.031 mm, and the overall performance was demonstrated on an OCT-guided automated cataract lens extraction procedure with deep learning-based pre-operative anatomical modeling and real-time supervision.
☆ Embodiment Transfer Learning for Vision-Language-Action Models
Vision-language-action (VLA) models have significantly advanced robotic learning, enabling training on large-scale, cross-embodiment data and fine-tuning for specific robots. However, state-of-the-art autoregressive VLAs struggle with multi-robot collaboration. We introduce embodiment transfer learning, denoted as ET-VLA, a novel framework for efficient and effective transfer of pre-trained VLAs to multi-robot. ET-VLA's core is Synthetic Continued Pretraining (SCP), which uses synthetically generated data to warm up the model for the new embodiment, bypassing the need for real human demonstrations and reducing data collection costs. SCP enables the model to learn correct actions and precise action token numbers. Following SCP, the model is fine-tuned on target embodiment data. To further enhance the model performance on multi-embodiment, we present the Embodied Graph-of-Thought technique, a novel approach that formulates each sub-task as a node, that allows the VLA model to distinguish the functionalities and roles of each embodiment during task execution. Our work considers bimanual robots, a simple version of multi-robot to verify our approaches. We validate the effectiveness of our method on both simulation benchmarks and real robots covering three different bimanual embodiments. In particular, our proposed ET-VLA \space can outperform OpenVLA on six real-world tasks over 53.2%. We will open-source all codes to support the community in advancing VLA models for robot learning.
☆ Saliency-Guided Domain Adaptation for Left-Hand Driving in Autonomous Steering
Domain adaptation is required for automated driving models to generalize well across diverse road conditions. This paper explores a training method for domain adaptation to adapt PilotNet, an end-to-end deep learning-based model, for left-hand driving conditions using real-world Australian highway data. Four training methods were evaluated: (1) a baseline model trained on U.S. right-hand driving data, (2) a model trained on flipped U.S. data, (3) a model pretrained on U.S. data and then fine-tuned on Australian highways, and (4) a model pretrained on flipped U.S. data and then finetuned on Australian highways. This setup examines whether incorporating flipped data enhances the model adaptation by providing an initial left-hand driving alignment. The paper compares model performance regarding steering prediction accuracy and attention, using saliency-based analysis to measure attention shifts across significant road regions. Results show that pretraining on flipped data alone worsens prediction stability due to misaligned feature representations, but significantly improves adaptation when followed by fine-tuning, leading to lower prediction error and stronger focus on left-side cues. To validate this approach across different architectures, the same experiments were done on ResNet, which confirmed similar adaptation trends. These findings emphasize the importance of preprocessing techniques, such as flipped-data pretraining, followed by fine-tuning to improve model adaptation with minimal retraining requirements.
☆ Tackling the Kidnapped Robot Problem via Sparse Feasible Hypothesis Sampling and Reliable Batched Multi-Stage Inference
This paper addresses the Kidnapped Robot Problem (KRP), a core localization challenge of relocalizing a robot in a known map without prior pose estimate when localization loss or at SLAM initialization. For this purpose, a passive 2-D global relocalization framework is proposed. It estimates the global pose efficiently and reliably from a single LiDAR scan and an occupancy grid map while the robot remains stationary, thereby enhancing the long-term autonomy of mobile robots. The proposed framework casts global relocalization as a non-convex problem and solves it via the multi-hypothesis scheme with batched multi-stage inference and early termination, balancing completeness and efficiency. The Rapidly-exploring Random Tree (RRT), under traversability constraints, asymptotically covers the reachable space to generate sparse, uniformly distributed feasible positional hypotheses, fundamentally reducing the sampling space. The hypotheses are preliminarily ordered by the proposed Scan Mean Absolute Difference (SMAD), a coarse beam-error level metric that facilitates the early termination by prioritizing high-likelihood candidates. The SMAD computation is optimized for non-panoramic scans. And the Translation-Affinity Scan-to-Map Alignment Metric (TAM) is proposed for reliable orientation selection at hypothesized positions and accurate final pose evaluation to mitigate degradation in conventional likelihood-field metrics under translational uncertainty induced by sparse hypotheses, as well as non-panoramic LiDAR scan and environmental changes. Real-world experiments on a resource-constrained mobile robot with non-panoramic LiDAR scan demonstrate that the proposed framework outperforms existing methods in both global relocalization success rate and computational efficiency.
comment: 10 pages, 8 figures. This work has been submitted to the IEEE for possible publication
☆ Closed-loop Control of Steerable Balloon Endoscopes for Robot-assisted Transcatheter Intracardiac Procedures
To move away from open-heart surgery towards safer transcatheter procedures, there is a growing need for improved imaging techniques and robotic solutions to enable simple, accurate tool navigation. Common imaging modalities, such as fluoroscopy and ultrasound, have limitations that can be overcome using cardioscopy, i.e., direct optical visualization inside the beating heart. We present a cardioscope designed as a steerable balloon. As a balloon, it can be collapsed to pass through the vasculature and subsequently inflated inside the heart for visualization and tool delivery through an integrated working channel. Through careful design of balloon wall thickness, a single input, balloon inflation pressure, is used to independently control two outputs, balloon diameter (corresponding to field of view diameter) and balloon bending angle (enabling precise working channel positioning). This balloon technology can be tuned to produce cardioscopes designed for a range of intracardiac tasks. To illustrate this approach, a balloon design is presented for the specific task of aortic leaflet laceration. Image-based closed-loop control of bending angle is also demonstrated as a means of enabling stable orientation control during tool insertion and removal.
comment: 8 pages, 11 figures
☆ LiDAR-VGGT: Cross-Modal Coarse-to-Fine Fusion for Globally Consistent and Metric-Scale Dense Mapping
Reconstructing large-scale colored point clouds is an important task in robotics, supporting perception, navigation, and scene understanding. Despite advances in LiDAR inertial visual odometry (LIVO), its performance remains highly sensitive to extrinsic calibration. Meanwhile, 3D vision foundation models, such as VGGT, suffer from limited scalability in large environments and inherently lack metric scale. To overcome these limitations, we propose LiDAR-VGGT, a novel framework that tightly couples LiDAR inertial odometry with the state-of-the-art VGGT model through a two-stage coarse- to-fine fusion pipeline: First, a pre-fusion module with robust initialization refinement efficiently estimates VGGT poses and point clouds with coarse metric scale within each session. Then, a post-fusion module enhances cross-modal 3D similarity transformation, using bounding-box-based regularization to reduce scale distortions caused by inconsistent FOVs between LiDAR and camera sensors. Extensive experiments across multiple datasets demonstrate that LiDAR-VGGT achieves dense, globally consistent colored point clouds and outperforms both VGGT-based methods and LIVO baselines. The implementation of our proposed novel color point cloud evaluation toolkit will be released as open source.
☆ Scaling Cross-Embodiment World Models for Dexterous Manipulation
Cross-embodiment learning seeks to build generalist robots that operate across diverse morphologies, but differences in action spaces and kinematics hinder data sharing and policy transfer. This raises a central question: Is there any invariance that allows actions to transfer across embodiments? We conjecture that environment dynamics are embodiment-invariant, and that world models capturing these dynamics can provide a unified interface across embodiments. To learn such a unified world model, the crucial step is to design state and action representations that abstract away embodiment-specific details while preserving control relevance. To this end, we represent different embodiments (e.g., human hands and robot hands) as sets of 3D particles and define actions as particle displacements, creating a shared representation for heterogeneous data and control problems. A graph-based world model is then trained on exploration data from diverse simulated robot hands and real human hands, and integrated with model-based planning for deployment on novel hardware. Experiments on rigid and deformable manipulation tasks reveal three findings: (i) scaling to more training embodiments improves generalization to unseen ones, (ii) co-training on both simulated and real data outperforms training on either alone, and (iii) the learned models enable effective control on robots with varied degrees of freedom. These results establish world models as a promising interface for cross-embodiment dexterous manipulation.
☆ An Enhanced Proprioceptive Method for Soft Robots Integrating Bend Sensors and IMUs
This study presents an enhanced proprioceptive method for accurate shape estimation of soft robots using only off-the-shelf sensors, ensuring cost-effectiveness and easy applicability. By integrating inertial measurement units (IMUs) with complementary bend sensors, IMU drift is mitigated, enabling reliable long-term proprioception. A Kalman filter fuses segment tip orientations from both sensors in a mutually compensatory manner, improving shape estimation over single-sensor methods. A piecewise constant curvature model estimates the tip location from the fused orientation data and reconstructs the robot's deformation. Experiments under no loading, external forces, and passive obstacle interactions during 45 minutes of continuous operation showed a root mean square error of 16.96 mm (2.91% of total length), a 56% reduction compared to IMU-only benchmarks. These results demonstrate that our approach not only enables long-duration proprioception in soft robots but also maintains high accuracy and robustness across these diverse conditions.
♻ ☆ Robotic Monitoring of Colorimetric Leaf Sensors for Precision Agriculture ICRA
Common remote sensing modalities (RGB, multispectral, hyperspectral imaging or LiDAR) are often used to indirectly measure crop health and do not directly capture plant stress indicators. Commercially available direct leaf sensors are bulky, powered electronics that are expensive and interfere with crop growth. In contrast, low-cost, passive and bio-degradable leaf sensors offer an opportunity to advance real-time monitoring as they directly interface with the crop surface while not interfering with crop growth. To this end, we co-design a sensor-detector system, where the sensor is a passive colorimetric leaf sensor that directly measures crop health in a precision agriculture setting, and the detector autonomously obtains optical signals from these leaf sensors. The detector comprises a low size weight and power (SWaP) mobile ground robot with an onboard monocular RGB camera and object detector to localize each leaf sensor, as well as a hyperspectral camera with a motorized mirror and halogen light to acquire hyperspectral images. The sensor's crop health-dependent optical signals can be extracted from the hyperspectral images. The proof-of-concept system is demonstrated in row-crop environments both indoors and outdoors where it is able to autonomously navigate, locate and obtain a hyperspectral image of all leaf sensors present, and acquire interpretable spectral resonance with 80 $\%$ accuracy within a required retrieval distance from the sensor.
comment: Revised version. Initial version was accepted to the Novel Approaches for Precision Agriculture and Forestry with Autonomous Robots IEEE ICRA Workshop - 2025
♻ ☆ Kinematically Controllable Cable Robots with Reconfigurable End-effectors
To enlarge the translational workspace of cable-driven robots, one common approach is to increase the number of cables. However, this introduces two challenges: (1) cable interference significantly reduces the rotational workspace, and (2) the solution of tensions in cables becomes non-unique, resulting in difficulties for kinematic control of the robot. In this work, we design structurally simple reconfigurable end-effectors for cable robots. By incorporating a spring, a helical-grooved shaft, and a matching nut, relative linear motions between end-effector components are converted into relative rotations, thereby expanding the rotational workspace of the mechanism. Meanwhile, a bearing is introduced to provide an additional rotational degree of freedom, making the mechanism non-redundant. As a result, the robot's motion can be controlled purely through kinematics without additional tension sensing and control.
comment: 8 pages, 7 figures, Technical Report
♻ ☆ Mixed-Density Diffuser: Efficient Planning with Non-uniform Temporal Resolution
Recent studies demonstrate that diffusion planners benefit from sparse-step planning over single-step planning. Training models to skip steps in their trajectories helps capture long-term dependencies without additional or memory computational cost. However, predicting excessively sparse plans degrades performance. We hypothesize this temporal density threshold is non-uniform across a temporal horizon and that certain parts of a planned trajectory should be more densely planned. We propose Mixed Density Diffuser (MDD), a diffusion planner where the densities throughout the horizon are tunable hyperparameters. MDD achieves a new SOTA across the Maze2D, Franka Kitchen, and Antmaze D4RL task domains.
comment: European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning (ESSAN) (under review)
♻ ☆ MarsLGPR: Mars Rover Localization with Ground Penetrating Radar
In this work, we propose the use of Ground Penetrating Radar (GPR) for rover localization on Mars. Precise pose estimation is an important task for mobile robots exploring planetary surfaces, as they operate in GPS-denied environments. Although visual odometry provides accurate localization, it is computationally expensive and can fail in dim or high-contrast lighting. Wheel encoders can also provide odometry estimation, but are prone to slipping on the sandy terrain encountered on Mars. Although traditionally a scientific surveying sensor, GPR has been used on Earth for terrain classification and localization through subsurface feature matching. The Perseverance rover and the upcoming ExoMars rover have GPR sensors already equipped to aid in the search of water and mineral resources. We propose to leverage GPR to aid in Mars rover localization. Specifically, we develop a novel GPR-based deep learning model that predicts 1D relative pose translation. We fuse our GPR pose prediction method with inertial and wheel encoder data in a filtering framework to output rover localization. We perform experiments in a Mars analog environment and demonstrate that our GPR-based displacement predictions both outperform wheel encoders and improve multi-modal filtering estimates in high-slip environments. Lastly, we present the first dataset aimed at GPR-based localization in Mars analog environments, which will be made publicly available at https://umfieldrobotics.github.io/marslgpr.
comment: IEEE Transactions on Field Robotics (2025)
♻ ☆ Cosmos-Surg-dVRK: World Foundation Model-based Automated Online Evaluation of Surgical Robot Policy Learning
The rise of surgical robots and vision-language-action models has accelerated the development of autonomous surgical policies and efficient assessment strategies. However, evaluating these policies directly on physical robotic platforms such as the da Vinci Research Kit (dVRK) remains hindered by high costs, time demands, reproducibility challenges, and variability in execution. World foundation models (WFM) for physical AI offer a transformative approach to simulate complex real-world surgical tasks, such as soft tissue deformation, with high fidelity. This work introduces Cosmos-Surg-dVRK, a surgical finetune of the Cosmos WFM, which, together with a trained video classifier, enables fully automated online evaluation and benchmarking of surgical policies. We evaluate Cosmos-Surg-dVRK using two distinct surgical datasets. On tabletop suture pad tasks, the automated pipeline achieves strong correlation between online rollouts in Cosmos-Surg-dVRK and policy outcomes on the real dVRK Si platform, as well as good agreement between human labelers and the V-JEPA 2-derived video classifier. Additionally, preliminary experiments with ex-vivo porcine cholecystectomy tasks in Cosmos-Surg-dVRK demonstrate promising alignment with real-world evaluations, highlighting the platform's potential for more complex surgical procedures.
comment: minor metadata and notation fixes; +3 citations
♻ ☆ Integrated Shape-Force Estimation for Continuum Robots: A Virtual-Work and Polynomial-Curvature Framework
Cable-driven continuum robots (CDCRs) are widely used in surgical and inspection tasks that require dexterous manipulation in confined spaces. Existing model-based estimation methods either assume constant curvature or rely on geometry-space interpolants, both of which struggle with accuracy under large deformations and sparse sensing. This letter introduces an integrated shape-force estimation framework that combines cable-tension measurements with tip-pose data to reconstruct backbone shape and estimate external tip force simultaneously. The framework employs polynomial curvature kinematics (PCK) and a virtual-work-based static formulation expressed directly in curvature space, where polynomial modal coefficients serve as generalized coordinates. The proposed method is validated through Cosserat-rod-based simulations and hardware experiments on a torque-cell-enabled CDCR prototype. Results show that the second-order PCK model achieves superior shape and force accuracy, combining a lightweight shape optimization with a closed-form, iteration-free force estimation, offering a compact and robust alternative to prior constant-curvature and geometry-space approaches.
♻ ☆ Interactive Identification of Granular Materials using Force Measurements
Despite the potential the ability to identify granular materials creates for applications such as robotic cooking or earthmoving, granular material identification remains a challenging area, existing methods mostly relying on shaking the materials in closed containers. This work presents an interactive material identification framework that enables robots to identify a wide range of granular materials using only force-torque measurements. Unlike prior works, the proposed approach uses direct interaction with the materials. The approach is evaluated through experiments with a real-world dataset comprising 11 granular materials, which we also make publicly available. Results show that our method can identify a wide range of granular materials with near-perfect accuracy while relying solely on force measurements obtained from direct interaction. Further, our comprehensive data analysis and experiments show that a high-performancefeature space must combine features related to the force signal's time-domain dynamics and frequency spectrum. We account for this by proposing a combination of the raw signal and its high-frequency magnitude histogram as the suggested feature space representation. We show that the proposed feature space outperforms baselines by a significant margin. The code and data set are available at: https://irobotics.aalto.fi/identify_granular/.
comment: Accepted to 2025 IEEE International Conference on Systems, Man, and Cybernetics (SMC)
♻ ☆ Closing the Intent-to-Behavior Gap via Fulfillment Priority Logic
Practitioners designing reinforcement learning policies face a fundamental challenge: translating intended behavioral objectives into representative reward functions. This challenge stems from behavioral intent requiring simultaneous achievement of multiple competing objectives, typically addressed through labor-intensive linear reward composition that yields brittle results. Consider the ubiquitous robotics scenario where performance maximization directly conflicts with energy conservation. Such competitive dynamics are resistant to simple linear reward combinations. In this paper, we present the concept of objective fulfillment upon which we build Fulfillment Priority Logic (FPL). FPL allows practitioners to define logical formula representing their intentions and priorities within multi-objective reinforcement learning. Our novel Balanced Policy Gradient algorithm leverages FPL specifications to achieve up to 500\% better sample efficiency compared to Soft Actor Critic. Notably, this work constitutes the first implementation of non-linear utility scalarization design, specifically for continuous control problems.
♻ ☆ The Difference between the Left and Right Invariant Extended Kalman Filter
The extended Kalman filter (EKF) has been the industry standard for state estimation problems over the past sixty years. The Invariant Extended Kalman Filter (IEKF) is a recent development of the EKF for the class of group-affine systems on Lie groups that has shown superior performance for inertial navigation problems. The IEKF comes in two versions, left- and right- handed respectively, and there is a perception in the robotics community that these filters are different and one should choose the handedness of the IEKF to match handedness of the measurement model for a given filtering problem. In this paper, we revisit these algorithms and demonstrate that the left- and right- IEKF algorithms (with reset step) are identical, that is, the choice of the handedness does not affect the IEKF's performance when the reset step is properly implemented. The reset step was not originally proposed as part of the IEKF, however, we provide simulations to show that the reset step improves asymptotic performance of all versions of the the filter, and should be included in all high performance algorithms. The GNSS-aided inertial navigation system (INS) is used as a motivating example to demonstrate the equivalence of the two filters.
comment: 20 pages, 4 figures, submitted to Control Engineering Practice
♻ ☆ Learning Terrain-Specialized Policies for Adaptive Locomotion in Challenging Environments
Legged robots must exhibit robust and agile locomotion across diverse, unstructured terrains, a challenge exacerbated under blind locomotion settings where terrain information is unavailable. This work introduces a hierarchical reinforcement learning framework that leverages terrain-specialized policies and curriculum learning to enhance agility and tracking performance in complex environments. We validated our method on simulation, where our approach outperforms a generalist policy by up to 16% in success rate and achieves lower tracking errors as the velocity target increases, particularly on low-friction and discontinuous terrains, demonstrating superior adaptability and robustness across mixed-terrain scenarios.
comment: Accepted to the 22nd International Conference on Advanced Robotics (ICAR 2025). 7 pages
♻ ☆ End-to-End Crop Row Navigation via LiDAR-Based Deep Reinforcement Learning
Reliable navigation in under-canopy agricultural environments remains a challenge due to GNSS unreliability, cluttered rows, and variable lighting. To address these limitations, we present an end-to-end learning-based navigation system that maps raw 3D LiDAR data directly to control commands using a deep reinforcement learning policy trained entirely in simulation. Our method includes a voxel-based downsampling strategy that reduces LiDAR input size by 95.83%, enabling efficient policy learning without relying on labeled datasets or manually designed control interfaces. The policy was validated in simulation, achieving a 100% success rate in straight-row plantations and showing a gradual decline in performance as row curvature increased, tested across varying sinusoidal frequencies and amplitudes.
comment: Accepted to the 22nd International Conference on Advanced Robotics (ICAR 2025). 7 pages
♻ ☆ Multi-Objective Planning with Contextual Lexicographic Reward Preferences AAMAS
Autonomous agents are often required to plan under multiple objectives whose preference ordering varies based on context. The agent may encounter multiple contexts during its course of operation, each imposing a distinct lexicographic ordering over the objectives, with potentially different reward functions associated with each context. Existing approaches to multi-objective planning typically consider a single preference ordering over the objectives, across the state space, and do not support planning under multiple objective orderings within an environment. We present Contextual Lexicographic Markov Decision Process (CLMDP), a framework that enables planning under varying lexicographic objective orderings, depending on the context. In a CLMDP, both the objective ordering at a state and the associated reward functions are determined by the context. We employ a Bayesian approach to infer a state-context mapping from expert trajectories. Our algorithm to solve a CLMDP first computes a policy for each objective ordering and then combines them into a single context-aware policy that is valid and cycle-free. The effectiveness of the proposed approach is evaluated in simulation and using a mobile robot.
comment: 9 pages, 5 figures, 2 tables, To appear in Proceedings of the 24th International Conference on Autonomous Agents and Multiagent Systems (AAMAS) 2025
♻ ☆ FlexEvent: Towards Flexible Event-Frame Object Detection at Varying Operational Frequencies NeurIPS 2025
Event cameras offer unparalleled advantages for real-time perception in dynamic environments, thanks to the microsecond-level temporal resolution and asynchronous operation. Existing event detectors, however, are limited by fixed-frequency paradigms and fail to fully exploit the high-temporal resolution and adaptability of event data. To address these limitations, we propose FlexEvent, a novel framework that enables detection at varying frequencies. Our approach consists of two key components: FlexFuse, an adaptive event-frame fusion module that integrates high-frequency event data with rich semantic information from RGB frames, and FlexTune, a frequency-adaptive fine-tuning mechanism that generates frequency-adjusted labels to enhance model generalization across varying operational frequencies. This combination allows our method to detect objects with high accuracy in both fast-moving and static scenarios, while adapting to dynamic environments. Extensive experiments on large-scale event camera datasets demonstrate that our approach surpasses state-of-the-art methods, achieving significant improvements in both standard and high-frequency settings. Notably, our method maintains robust performance when scaling from 20 Hz to 90 Hz and delivers accurate detection up to 180 Hz, proving its effectiveness in extreme conditions. Our framework sets a new benchmark for event-based object detection and paves the way for more adaptable, real-time vision systems.
comment: NeurIPS 2025; 28 pages, 14 figures, 10 tables; Code at https://flexevent.github.io/
♻ ☆ If They Disagree, Will You Conform? Exploring the Role of Robots' Value Awareness in a Decision-Making Task
This study investigates whether the opinions of robotic agents can influence human decision-making when robots display value awareness (i.e., the capability of understanding human preferences and prioritizing them in decision-making). We designed an experiment in which participants interacted with two Furhat robots - one programmed to be Value-Aware and the other Non-Value-Aware - during a labeling task for images representing human values. Results indicate that participants distinguished the Value-Aware robot from the Non-Value-Aware one. Although their explicit choices did not indicate a clear preference for one robot over the other, participants directed their gaze more toward the Value-Aware robot. Additionally, the Value-Aware robot was perceived as more loyal, suggesting that value awareness in a social robot may enhance its perceived commitment to the group. Finally, when both robots disagreed with the participant, conformity occurred in about one out of four trials, and participants took longer to confirm their responses, suggesting that two robots expressing dissent may introduce hesitation in decision-making. On one hand, this highlights the potential risk that robots, if misused, could manipulate users for unethical purposes. On the other hand, it reinforces the idea that social robots could encourage reflection in ambiguous situations and help users avoid scams.
comment: Pre-print version
♻ ☆ DW-A-PRM: A Dynamic Weighted Planner
Robot path planning plays a pivotal role in enabling autonomous systems to navigate safely and efficiently in complex and uncertain environments. Despite extensive research on classical graph-based methods and sampling-based planners, achieving an optimal balance between global optimality, computational efficiency, and adaptability to dynamic environments remains an open challenge. To address this issue, this paper proposes a hybrid path planning framework, which integrates heuristic-driven search with probabilistic roadmap construction under a dynamic weighting scheme. By coupling the global guidance of A* with the stochastic exploration of PRM, the method achieves a synergistic balance between search optimality and computational tractability. Comprehensive experiments in diverse simulated environments demonstrate that the proposed method consistently yields smoother and shorter paths while significantly reducing computational overhead compared with conventional approach and other hybrid planners. These results highlight the potential of the proposed framework as an effective and generalizable solution for real-time robotic navigation in complex environments.
♻ ☆ RL-100: Performant Robotic Manipulation with Real-World Reinforcement Learning
Real-world robotic manipulation in homes and factories demands reliability, efficiency, and robustness that approach or surpass skilled human operators. We present RL-100, a real-world reinforcement learning training framework built on diffusion visuomotor policies trained by supervised learning. RL-100 introduces a three-stage pipeline. First, imitation learning leverages human priors. Second, iterative offline reinforcement learning uses an Offline Policy Evaluation procedure, abbreviated OPE, to gate PPO-style updates that are applied in the denoising process for conservative and reliable improvement. Third, online reinforcement learning eliminates residual failure modes. An additional lightweight consistency distillation head compresses the multi-step sampling process in diffusion into a single-step policy, enabling high-frequency control with an order-of-magnitude reduction in latency while preserving task performance. The framework is task-, embodiment-, and representation-agnostic and supports both 3D point clouds and 2D RGB inputs, a variety of robot platforms, and both single-step and action-chunk policies. We evaluate RL-100 on seven real-robot tasks spanning dynamic rigid-body control, such as Push-T and Agile Bowling, fluids and granular pouring, deformable cloth folding, precise dexterous unscrewing, and multi-stage orange juicing. RL-100 attains 100\% success across evaluated trials for a total of 900 out of 900 episodes, including up to 250 out of 250 consecutive trials on one task. The method achieves near-human teleoperation or better time efficiency and demonstrates multi-hour robustness with uninterrupted operation lasting up to two hours.
comment: https://lei-kun.github.io/RL-100/
♻ ☆ Neuro-Symbolic Imitation Learning: Discovering Symbolic Abstractions for Skill Learning ICRA
Imitation learning is a popular method for teaching robots new behaviors. However, most existing methods focus on teaching short, isolated skills rather than long, multi-step tasks. To bridge this gap, imitation learning algorithms must not only learn individual skills but also an abstract understanding of how to sequence these skills to perform extended tasks effectively. This paper addresses this challenge by proposing a neuro-symbolic imitation learning framework. Using task demonstrations, the system first learns a symbolic representation that abstracts the low-level state-action space. The learned representation decomposes a task into easier subtasks and allows the system to leverage symbolic planning to generate abstract plans. Subsequently, the system utilizes this task decomposition to learn a set of neural skills capable of refining abstract plans into actionable robot commands. Experimental results in three simulated robotic environments demonstrate that, compared to baselines, our neuro-symbolic approach increases data efficiency, improves generalization capabilities, and facilitates interpretability.
comment: IEEE International Conference on Robotics and Automation (ICRA) 2025
♻ ☆ Infinite-Horizon Value Function Approximation for Model Predictive Control
Model Predictive Control has emerged as a popular tool for robots to generate complex motions. However, the real-time requirement has limited the use of hard constraints and large preview horizons, which are necessary to ensure safety and stability. In practice, practitioners have to carefully design cost functions that can imitate an infinite horizon formulation, which is tedious and often results in local minima. In this work, we study how to approximate the infinite horizon value function of constrained optimal control problems with neural networks using value iteration and trajectory optimization. Furthermore, we experimentally demonstrate how using this value function approximation as a terminal cost provides global stability to the model predictive controller. The approach is validated on two toy problems and a real-world scenario with online obstacle avoidance on an industrial manipulator where the value function is conditioned to the goal and obstacle.
♻ ☆ UniVLA: Learning to Act Anywhere with Task-centric Latent Actions
A generalist robot should perform effectively across various environments. However, most existing approaches heavily rely on scaling action-annotated data to enhance their capabilities. Consequently, they are often limited to single physical specification and struggle to learn transferable knowledge across different embodiments and environments. To confront these limitations, we propose UniVLA, a new framework for learning cross-embodiment vision-language-action (VLA) policies. Our key innovation is to derive task-centric action representations from videos with a latent action model. This enables us to exploit extensive data across a wide spectrum of embodiments and perspectives. To mitigate the effect of task-irrelevant dynamics, we incorporate language instructions and establish a latent action model within the DINO feature space. Learned from internet-scale videos, the generalist policy can be deployed to various robots through efficient latent action decoding. We obtain state-of-the-art results across multiple manipulation and navigation benchmarks, as well as real-robot deployments. UniVLA achieves superior performance over OpenVLA with less than 1/20 of pretraining compute and 1/10 of downstream data. Continuous performance improvements are observed as heterogeneous data, even including human videos, are incorporated into the training pipeline. The results underscore UniVLA's potential to facilitate scalable and efficient robot policy learning.
comment: Accepted to RSS 2025. Code is available at https://github.com/OpenDriveLab/UniVLA
♻ ☆ A Helping (Human) Hand in Kinematic Structure Estimation ICRA25
Visual uncertainties such as occlusions, lack of texture, and noise present significant challenges in obtaining accurate kinematic models for safe robotic manipulation. We introduce a probabilistic real-time approach that leverages the human hand as a prior to mitigate these uncertainties. By tracking the constrained motion of the human hand during manipulation and explicitly modeling uncertainties in visual observations, our method reliably estimates an object's kinematic model online. We validate our approach on a novel dataset featuring challenging objects that are occluded during manipulation and offer limited articulations for perception. The results demonstrate that by incorporating an appropriate prior and explicitly accounting for uncertainties, our method produces accurate estimates, outperforming two recent baselines by 195% and 140%, respectively. Furthermore, we demonstrate that our approach's estimates are precise enough to allow a robot to manipulate even small objects safely.
comment: Accepted at ICRA25; 8 pages + 7 figures; For supplementary material, see https://www.tu.berlin/robotics/papers/helpinghands
♻ ☆ VO-DP: Semantic-Geometric Adaptive Diffusion Policy for Vision-Only Robotic Manipulation
In the context of imitation learning, visuomotor-based diffusion policy learning is one of the main directions in robotic manipulation. Most of these approaches rely on point clouds as observation inputs and construct scene representations through point clouds feature learning, which enables them to achieve remarkable accuracy. However, the existing literature lacks an in-depth exploration of vision-only solutions that have significant potential. In this paper, we propose a Vision-Only and single-view Diffusion Policy learning method (VO-DP) that leverages pretrained visual foundation models to achieve effective fusion of semantic and geometric features. We utilize intermediate features from VGGT incorporating semantic features from DINOv2 and geometric features from Alternating Attention blocks. Features are fused via cross-attention and spatially compressed with a CNN to form the input to the policy head. Extensive experiments demonstrate that VO-DP not only outperforms the vision-only baseline DP significantly but also exhibits distinct performance trends against the point cloud-based method DP3: in simulation tasks, VO-DP achieves an average success rate of 64.6% on par with DP3 64.0% and far higher than DP 34.8%, while in real-world tasks, it reaches 87.9%, outperforming both DP3 67.5% and DP 11.2% by a notable margin. Further robustness evaluations confirm that VO-DP remains highly stable under varying conditions including color, size, background, and lighting. Lastly, we open-source a training library for robotic manipulation. Built on Accelerate, this library supports multi-machine and multi-GPU parallel training, as well as mixed precision training. It is compatible with visuomotor policies such as DP, DP3 and VO-DP, and also supports the RoboTwin simulator.
♻ ☆ Bellman Diffusion Models
Diffusion models have seen tremendous success as generative architectures. Recently, they have been shown to be effective at modelling policies for offline reinforcement learning and imitation learning. We explore using diffusion as a model class for the successor state measure (SSM) of a policy. We find that enforcing the Bellman flow constraints leads to a simple Bellman update on the diffusion step distribution.
♻ ☆ MARFT: Multi-Agent Reinforcement Fine-Tuning
LLM-based Multi-Agent Systems have demonstrated remarkable capabilities in addressing complex, agentic tasks, from generating high-quality presentation slides to even conducting sophisticated scientific research. Meanwhile, RL has been widely recognized for its effectiveness in enhancing agent intelligence, but limited research has investigated the fine-tuning of LaMAS using foundational RL techniques. Moreover, the direct application of MARL methods to LaMAS introduces significant challenges, stemming from the unique characteristics and mechanisms inherent to LaMAS. To address these challenges, this article presents a comprehensive study of LLM-based MARL and proposes a novel paradigm termed Multi-Agent Reinforcement Fine-Tuning (MARFT). We introduce a brand-new MG called Flex-MG, which aligns with the LaMAS optimization in real-world applications and a universal algorithmic framework tailored specifically for LaMAS, outlining the conceptual foundations, key distinctions, and practical implementation strategies. We review the evolution from RL to RFT, setting the stage for a parallel analysis in the multi-agent domain. In the context of LaMAS, we elucidate critical differences between MARL and MARFT. These differences motivate a transition toward a LaMAS-oriented formulation of RFT. Central to this work is a robust and scalable MARFT framework. We detail the core algorithm and provide a complete, open-source implementation to facilitate adoption and further research. The latter sections of the paper explore real-world application perspectives and opening challenges in MARFT. By bridging theoretical underpinnings with practical methodologies, this work serves as a roadmap for researchers seeking to advance MARFT toward resilient and adaptive solutions in agentic systems. Our implementation of the proposed framework is publicly available at: https://github.com/jwliao-ai/MARFT.
comment: 42 pages
♻ ☆ A Time-dependent Risk-aware distributed Multi-Agent Path Finder based on A* IROS 2025
Multi-Agent Path-Finding (MAPF) focuses on the collaborative planning of paths for multiple agents within shared spaces, aiming for collision-free navigation. Conventional planning methods often overlook the presence of other agents, which can result in conflicts. In response, this article introduces the A$^*_+$T algorithm, a distributed approach that improves coordination among agents by anticipating their positions based on their movement speeds. The algorithm also considers dynamic obstacles, assessing potential collisions with respect to observed speeds and trajectories, thereby facilitating collision-free path planning in environments populated by other agents and moving objects. It incorporates a risk layer surrounding both dynamic and static entities, enhancing its utility in real-world applications. Each agent functions autonomously while being mindful of the paths chosen by others, effectively addressing the complexities inherent in multi-agent situations. The performance of A$^*_+$T has been rigorously tested in the Gazebo simulation environment and benchmarked against established approaches such as CBS, ECBS, and SIPP. Furthermore, the algorithm has shown competence in single-agent experiments, with results demonstrating its effectiveness in managing dynamic obstacles and affirming its practical relevance across various scenarios.
comment: 8 pages, 10 figures, 2 tabels, submited to IROS 2025
♻ ☆ DTAA: A Detect, Track and Avoid Architecture for navigation in spaces with Multiple Velocity Objects
Proactive collision avoidance measures are imperative in environments where humans and robots coexist. Moreover, the introduction of high quality legged robots into workplaces highlighted the crucial role of a robust, fully autonomous safety solution for robots to be viable in shared spaces or in co-existence with humans. This article establishes for the first time ever an innovative Detect-Track-and-Avoid Architecture (DTAA) to enhance safety and overall mission performance. The proposed novel architectyre has the merit ot integrating object detection using YOLOv8, utilizing Ultralytics embedded object tracking, and state estimation of tracked objects through Kalman filters. Moreover, a novel heuristic clustering is employed to facilitate active avoidance of multiple closely positioned objects with similar velocities, creating sets of unsafe spaces for the Nonlinear Model Predictive Controller (NMPC) to navigate around. The NMPC identifies the most hazardous unsafe space, considering not only their current positions but also their predicted future locations. In the sequel, the NMPC calculates maneuvers to guide the robot along a path planned by D$^{*}_{+}$ towards its intended destination, while maintaining a safe distance to all identified obstacles. The efficacy of the novelly suggested DTAA framework is being validated by Real-life experiments featuring a Boston Dynamics Spot robot that demonstrates the robot's capability to consistently maintain a safe distance from humans in dynamic subterranean, urban indoor, and outdoor environments.
♻ ☆ Spatiotemporal Calibration for Laser Vision Sensor in Hand-eye System Based on Straight-line Constraint
Laser vision sensors (LVS) are critical perception modules for industrial robots, facilitating real-time acquisition of workpiece geometric data in welding applications. However, the camera communication delay will lead to a temporal desynchronization between captured images and the robot motions. Additionally, hand-eye extrinsic parameters may vary during prolonged measurement. To address these issues, we introduce a measurement model of LVS considering the effect of the camera's time-offset and propose a teaching-free spatiotemporal calibration method utilizing line constraints. This method involves a robot equipped with an LVS repeatedly scanning straight-line fillet welds using S-shaped trajectories. Regardless of the robot's orientation changes, all measured welding positions are constrained to a straight-line, represented by Plucker coordinates. Moreover, a nonlinear optimization model based on straight-line constraints is established. Subsequently, the Levenberg-Marquardt algorithm (LMA) is employed to optimize parameters, including time-offset, hand-eye extrinsic parameters, and straight-line parameters. The feasibility and accuracy of the proposed approach are quantitatively validated through experiments on curved weld scanning. We open-sourced the code, dataset, and simulation report at https://anonymous.4open.science/r/LVS_ST_CALIB-015F/README.md.
comment: Submitted to IEEE RAL
♻ ☆ Dual-Regularized Riccati Recursions for Interior-Point Optimal Control
We derive closed-form extensions of Riccati's recursions (both sequential and parallel) for solving dual-regularized LQR problems. We show how these methods can be used to solve general constrained, non-convex, discrete-time optimal control problems via a regularized interior point method, while guaranteeing that each primal step is a descent direction of an Augmented Barrier-Lagrangian merit function. We provide MIT-licensed implementations of our methods in C++ and JAX.
♻ ☆ From Grounding to Manipulation: Case Studies of Foundation Model Integration in Embodied Robotic Systems EMNLP 2025
Foundation models (FMs) are increasingly used to bridge language and action in embodied agents, yet the operational characteristics of different FM integration strategies remain under-explored -- particularly for complex instruction following and versatile action generation in changing environments. This paper examines three paradigms for building robotic systems: end-to-end vision-language-action (VLA) models that implicitly integrate perception and planning, and modular pipelines incorporating either vision-language models (VLMs) or multimodal large language models (LLMs). We evaluate these paradigms through two focused case studies: a complex instruction grounding task assessing fine-grained instruction understanding and cross-modal disambiguation, and an object manipulation task targeting skill transfer via VLA finetuning. Our experiments in zero-shot and few-shot settings reveal trade-offs in generalization and data efficiency. By exploring performance limits, we distill design implications for developing language-driven physical agents and outline emerging challenges and opportunities for FM-powered robotics in real-world conditions.
comment: EMNLP 2025 camera ready
♻ ☆ Simultaneous System Identification and Model Predictive Control with No Dynamic Regret
We provide an algorithm for the simultaneous system identification and model predictive control of nonlinear systems. The algorithm has finite-time near-optimality guarantees and asymptotically converges to the optimal (non-causal) controller. Particularly, the algorithm enjoys sublinear dynamic regret, defined herein as the suboptimality against an optimal clairvoyant controller that knows how the unknown disturbances and system dynamics will adapt to its actions. The algorithm is self-supervised and applies to control-affine systems with unknown dynamics and disturbances that can be expressed in reproducing kernel Hilbert spaces. Such spaces can model external disturbances and modeling errors that can even be adaptive to the system's state and control input. For example, they can model wind and wave disturbances to aerial and marine vehicles, or inaccurate model parameters such as inertia of mechanical systems. The algorithm first generates random Fourier features that are used to approximate the unknown dynamics or disturbances. Then, it employs model predictive control based on the current learned model of the unknown dynamics (or disturbances). The model of the unknown dynamics is updated online using least squares based on the data collected while controlling the system. We validate our algorithm in both hardware experiments and physics-based simulations. The simulations include (i) a cart-pole aiming to maintain the pole upright despite inaccurate model parameters, and (ii) a quadrotor aiming to track reference trajectories despite unmodeled aerodynamic drag effects. The hardware experiments include a quadrotor aiming to track a circular trajectory despite unmodeled aerodynamic drag effects, ground effects, and wind disturbances.
comment: IEEE Transactions on Robotics (T-RO). v6 update on stability analysis in Appendix J under relaxed Assumption 1
♻ ☆ iKap: Kinematics-aware Planning with Imperative Learning
Trajectory planning in robotics aims to generate collision-free pose sequences that can be reliably executed. Recently, vision-to-planning systems have gained increasing attention for their efficiency and ability to interpret and adapt to surrounding environments. However, traditional modular systems suffer from increased latency and error propagation, while purely data-driven approaches often overlook the robot's kinematic constraints. This oversight leads to discrepancies between planned trajectories and those that are executable. To address these challenges, we propose iKap, a novel vision-to-planning system that integrates the robot's kinematic model directly into the learning pipeline. iKap employs a self-supervised learning approach and incorporates the state transition model within a differentiable bi-level optimization framework. This integration ensures the network learns collision-free waypoints while satisfying kinematic constraints, enabling gradient back-propagation for end-to-end training. Our experimental results demonstrate that iKap achieves higher success rates and reduced latency compared to the state-of-the-art methods. Besides the complete system, iKap offers a visual-to-planning network that seamlessly works with various controllers, providing a robust solution for robots navigating complex environments.
comment: 6 pages, 6 figures
♻ ☆ Dexterous Contact-Rich Manipulation via the Contact Trust Region
What is a good local description of contact dynamics for contact-rich manipulation, and where can we trust this local description? While many approaches often rely on the Taylor approximation of dynamics with an ellipsoidal trust region, we argue that such approaches are fundamentally inconsistent with the unilateral nature of contact. As a remedy, we present the Contact Trust Region (CTR), which captures the unilateral nature of contact while remaining efficient for computation. With CTR, we first develop a Model-Predictive Control (MPC) algorithm capable of synthesizing local contact-rich plans. Then, we extend this capability to plan globally by stitching together local MPC plans, enabling efficient and dexterous contact-rich manipulation. To verify the performance of our method, we perform comprehensive evaluations, both in high-fidelity simulation and on hardware, on two contact-rich systems: a planar IiwaBimanual system and a 3D AllegroHand system. On both systems, our method offers a significantly lower-compute alternative to existing RL-based approaches to contact-rich manipulation. In particular, our Allegro in-hand manipulation policy, in the form of a roadmap, takes fewer than 10 minutes to build offline on a standard laptop using just its CPU, with online inference taking just a few seconds. Experiment data, video and code are available at ctr.theaiinstitute.com.
♻ ☆ Adaptive Multirobot Virtual Structure Control using Dual Quaternions
This paper presents a control strategy based on dual quaternions for the coordinated formation flying of small UAV groups. A virtual structure is employed to define the desired formation, enabling unified control of its position, orientation, and shape. This abstraction makes formation management easier by allowing a low-level controller to compute individual UAV commands efficiently. The proposed controller integrates a pose control module with a geometry-based adaptive strategy, ensuring precise and robust task execution. The effectiveness of the approach is demonstrated through both simulation and experimental results.
Robotics 18
☆ SLAP: Shortcut Learning for Abstract Planning
Long-horizon decision-making with sparse rewards and continuous states and actions remains a fundamental challenge in AI and robotics. Task and motion planning (TAMP) is a model-based framework that addresses this challenge by planning hierarchically with abstract actions (options). These options are manually defined, limiting the agent to behaviors that we as human engineers know how to program (pick, place, move). In this work, we propose Shortcut Learning for Abstract Planning (SLAP), a method that leverages existing TAMP options to automatically discover new ones. Our key idea is to use model-free reinforcement learning (RL) to learn shortcuts in the abstract planning graph induced by the existing options in TAMP. Without any additional assumptions or inputs, shortcut learning leads to shorter solutions than pure planning, and higher task success rates than flat and hierarchical RL. Qualitatively, SLAP discovers dynamic physical improvisations (e.g., slap, wiggle, wipe) that differ significantly from the manually-defined ones. In experiments in four simulated robotic environments, we show that SLAP solves and generalizes to a wide range of tasks, reducing overall plan lengths by over 50% and consistently outperforming planning and RL baselines.
☆ Deployable Vision-driven UAV River Navigation via Human-in-the-loop Preference Alignment ICRA 2026
Rivers are critical corridors for environmental monitoring and disaster response, where Unmanned Aerial Vehicles (UAVs) guided by vision-driven policies can provide fast, low-cost coverage. However, deployment exposes simulation-trained policies with distribution shift and safety risks and requires efficient adaptation from limited human interventions. We study human-in-the-loop (HITL) learning with a conservative overseer who vetoes unsafe or inefficient actions and provides statewise preferences by comparing the agent's proposal with a corrective override. We introduce Statewise Hybrid Preference Alignment for Robotics (SPAR-H), which fuses direct preference optimization on policy logits with a reward-based pathway that trains an immediate-reward estimator from the same preferences and updates the policy using a trust-region surrogate. With five HITL rollouts collected from a fixed novice policy, SPAR-H achieves the highest final episodic reward and the lowest variance across initial conditions among tested methods. The learned reward model aligns with human-preferred actions and elevates nearby non-intervened choices, supporting stable propagation of improvements. We benchmark SPAR-H against imitation learning (IL), direct preference variants, and evaluative reinforcement learning (RL) in the HITL setting, and demonstrate real-world feasibility of continual preference alignment for UAV river following. Overall, dual statewise preferences empirically provide a practical route to data-efficient online adaptation in riverine navigation.
comment: Submitted to ICRA 2026
☆ AquaROM: shape optimization pipeline for soft swimmers using parametric reduced order models
The efficient optimization of actuated soft structures, particularly under complex nonlinear forces, remains a critical challenge in advancing robotics. Simulations of nonlinear structures, such as soft-bodied robots modeled using the finite element method (FEM), often demand substantial computational resources, especially during optimization. To address this challenge, we propose a novel optimization algorithm based on a tensorial parametric reduced order model (PROM). Our algorithm leverages dimensionality reduction and solution approximation techniques to facilitate efficient solving of nonlinear constrained optimization problems. The well-structured tensorial approach enables the use of analytical gradients within a specifically chosen reduced order basis (ROB), significantly enhancing computational efficiency. To showcase the performance of our method, we apply it to optimizing soft robotic swimmer shapes. These actuated soft robots experience hydrodynamic forces, subjecting them to both internal and external nonlinear forces, which are incorporated into our optimization process using a data-free ROB for fast and accurate computations. This approach not only reduces computational complexity but also unlocks new opportunities to optimize complex nonlinear systems in soft robotics, paving the way for more efficient design and control.
☆ GauDP: Reinventing Multi-Agent Collaboration through Gaussian-Image Synergy in Diffusion Policies NeurIPS 2025
Recently, effective coordination in embodied multi-agent systems has remained a fundamental challenge, particularly in scenarios where agents must balance individual perspectives with global environmental awareness. Existing approaches often struggle to balance fine-grained local control with comprehensive scene understanding, resulting in limited scalability and compromised collaboration quality. In this paper, we present GauDP, a novel Gaussian-image synergistic representation that facilitates scalable, perception-aware imitation learning in multi-agent collaborative systems. Specifically, GauDP constructs a globally consistent 3D Gaussian field from decentralized RGB observations, then dynamically redistributes 3D Gaussian attributes to each agent's local perspective. This enables all agents to adaptively query task-critical features from the shared scene representation while maintaining their individual viewpoints. This design facilitates both fine-grained control and globally coherent behavior without requiring additional sensing modalities (e.g., 3D point cloud). We evaluate GauDP on the RoboFactory benchmark, which includes diverse multi-arm manipulation tasks. Our method achieves superior performance over existing image-based methods and approaches the effectiveness of point-cloud-driven methods, while maintaining strong scalability as the number of agents increases.
comment: Accepted by NeurIPS 2025. Project page: https://ziyeeee.github.io/gaudp.io/
☆ Breaking the Latency Barrier: Synergistic Perception and Control for High-Frequency 3D Ultrasound Servoing
Real-time tracking of dynamic targets amidst large-scale, high-frequency disturbances remains a critical unsolved challenge in Robotic Ultrasound Systems (RUSS), primarily due to the end-to-end latency of existing systems. This paper argues that breaking this latency barrier requires a fundamental shift towards the synergistic co-design of perception and control. We realize it in a novel framework with two tightly-coupled contributions: (1) a Decoupled Dual-Stream Perception Network that robustly estimates 3D translational state from 2D images at high frequency, and (2) a Single-Step Flow Policy that generates entire action sequences in one inference pass, bypassing the iterative bottleneck of conventional policies. This synergy enables a closed-loop control frequency exceeding 60Hz. On a dynamic phantom, our system not only tracks complex 3D trajectories with a mean error below 6.5mm but also demonstrates robust re-acquisition from over 170mm displacement. Furthermore, it can track targets at speeds of 102mm/s, achieving a terminal error below 1.7mm. Moreover, in-vivo experiments on a human volunteer validate the framework's effectiveness and robustness in a realistic clinical setting. Our work presents a RUSS holistically architected to unify high-bandwidth tracking with large-scale repositioning, a critical step towards robust autonomy in dynamic clinical environments.
☆ URDF-Anything: Constructing Articulated Objects with 3D Multimodal Language Model NeurIPS 2025
Constructing accurate digital twins of articulated objects is essential for robotic simulation training and embodied AI world model building, yet historically requires painstaking manual modeling or multi-stage pipelines. In this work, we propose \textbf{URDF-Anything}, an end-to-end automatic reconstruction framework based on a 3D multimodal large language model (MLLM). URDF-Anything utilizes an autoregressive prediction framework based on point-cloud and text multimodal input to jointly optimize geometric segmentation and kinematic parameter prediction. It implements a specialized $[SEG]$ token mechanism that interacts directly with point cloud features, enabling fine-grained part-level segmentation while maintaining consistency with the kinematic parameter predictions. Experiments on both simulated and real-world datasets demonstrate that our method significantly outperforms existing approaches regarding geometric segmentation (mIoU 17\% improvement), kinematic parameter prediction (average error reduction of 29\%), and physical executability (surpassing baselines by 50\%). Notably, our method exhibits excellent generalization ability, performing well even on objects outside the training set. This work provides an efficient solution for constructing digital twins for robotic simulation, significantly enhancing the sim-to-real transfer capability.
comment: Accepted to the 39th Conference on Neural Information Processing Systems (NeurIPS 2025)
☆ pacSTL: PAC-Bounded Signal Temporal Logic from Data-Driven Reachability Analysis
Real-world robotic systems must comply with safety requirements in the presence of uncertainty. To define and measure requirement adherence, Signal Temporal Logic (STL) offers a mathematically rigorous and expressive language. However, standard STL cannot account for uncertainty. We address this problem by presenting pacSTL, a framework that combines Probably Approximately Correct (PAC) bounded set predictions with an interval extension of STL through optimization problems on the atomic proposition level. pacSTL provides PAC-bounded robustness intervals on the specification level that can be utilized in monitoring. We demonstrate the effectiveness of this approach through maritime navigation and analyze the efficiency and scalability of pacSTL through simulation and real-world experimentation on model vessels.
☆ Fast-SmartWay: Panoramic-Free End-to-End Zero-Shot Vision-and-Language Navigation
Recent advances in Vision-and-Language Navigation in Continuous Environments (VLN-CE) have leveraged multimodal large language models (MLLMs) to achieve zero-shot navigation. However, existing methods often rely on panoramic observations and two-stage pipelines involving waypoint predictors, which introduce significant latency and limit real-world applicability. In this work, we propose Fast-SmartWay, an end-to-end zero-shot VLN-CE framework that eliminates the need for panoramic views and waypoint predictors. Our approach uses only three frontal RGB-D images combined with natural language instructions, enabling MLLMs to directly predict actions. To enhance decision robustness, we introduce an Uncertainty-Aware Reasoning module that integrates (i) a Disambiguation Module for avoiding local optima, and (ii) a Future-Past Bidirectional Reasoning mechanism for globally coherent planning. Experiments on both simulated and real-robot environments demonstrate that our method significantly reduces per-step latency while achieving competitive or superior performance compared to panoramic-view baselines. These results demonstrate the practicality and effectiveness of Fast-SmartWay for real-world zero-shot embodied navigation.
☆ Maestro: Orchestrating Robotics Modules with Vision-Language Models for Zero-Shot Generalist Robots
Today's best-explored routes towards generalist robots center on collecting ever larger "observations-in actions-out" robotics datasets to train large end-to-end models, copying a recipe that has worked for vision-language models (VLMs). We pursue a road less traveled: building generalist policies directly around VLMs by augmenting their general capabilities with specific robot capabilities encapsulated in a carefully curated set of perception, planning, and control modules. In Maestro, a VLM coding agent dynamically composes these modules into a programmatic policy for the current task and scenario. Maestro's architecture benefits from a streamlined closed-loop interface without many manually imposed structural constraints, and a comprehensive and diverse tool repertoire. As a result, it largely surpasses today's VLA models for zero-shot performance on challenging manipulation skills. Further, Maestro is easily extensible to incorporate new modules, easily editable to suit new embodiments such as a quadruped-mounted arm, and even easily adapts from minimal real-world experiences through local code edits.
comment: Project website: https://maestro-robot.github.io
☆ Heuristic Step Planning for Learning Dynamic Bipedal Locomotion: A Comparative Study of Model-Based and Model-Free Approaches
This work presents an extended framework for learning-based bipedal locomotion that incorporates a heuristic step-planning strategy guided by desired torso velocity tracking. The framework enables precise interaction between a humanoid robot and its environment, supporting tasks such as crossing gaps and accurately approaching target objects. Unlike approaches based on full or simplified dynamics, the proposed method avoids complex step planners and analytical models. Step planning is primarily driven by heuristic commands, while a Raibert-type controller modulates the foot placement length based on the error between desired and actual torso velocity. We compare our method with a model-based step-planning approach -- the Linear Inverted Pendulum Model (LIPM) controller. Experimental results demonstrate that our approach attains comparable or superior accuracy in maintaining target velocity (up to 80%), significantly greater robustness on uneven terrain (over 50% improvement), and improved energy efficiency. These results suggest that incorporating complex analytical, model-based components into the training architecture may be unnecessary for achieving stable and robust bipedal walking, even in unstructured environments.
☆ Real-Time Learning of Predictive Dynamic Obstacle Models for Robotic Motion Planning ICRA
Autonomous systems often must predict the motions of nearby agents from partial and noisy data. This paper asks and answers the question: "can we learn, in real-time, a nonlinear predictive model of another agent's motions?" Our online framework denoises and forecasts such dynamics using a modified sliding-window Hankel Dynamic Mode Decomposition (Hankel-DMD). Partial noisy measurements are embedded into a Hankel matrix, while an associated Page matrix enables singular-value hard thresholding (SVHT) to estimate the effective rank. A Cadzow projection enforces structured low-rank consistency, yielding a denoised trajectory and local noise variance estimates. From this representation, a time-varying Hankel-DMD lifted linear predictor is constructed for multi-step forecasts. The residual analysis provides variance-tracking signals that can support downstream estimators and risk-aware planning. We validate the approach in simulation under Gaussian and heavy-tailed noise, and experimentally on a dynamic crane testbed. Results show that the method achieves stable variance-aware denoising and short-horizon prediction suitable for integration into real-time control frameworks.
comment: 10 pages, 6 figures, submitted to IEEE International Conference on Robotics and Automation (ICRA) 2025
☆ When Semantics Connect the Swarm: LLM-Driven Fuzzy Control for Cooperative Multi-Robot Underwater Coverage
Underwater multi-robot cooperative coverage remains challenging due to partial observability, limited communication, environmental uncertainty, and the lack of access to global localization. To address these issues, this paper presents a semantics-guided fuzzy control framework that couples Large Language Models (LLMs) with interpretable control and lightweight coordination. Raw multimodal observations are compressed by the LLM into compact, human-interpretable semantic tokens that summarize obstacles, unexplored regions, and Objects Of Interest (OOIs) under uncertain perception. A fuzzy inference system with pre-defined membership functions then maps these tokens into smooth and stable steering and gait commands, enabling reliable navigation without relying on global positioning. Then, we further coordinate multiple robots by introducing semantic communication that shares intent and local context in linguistic form, enabling agreement on who explores where while avoiding redundant revisits. Extensive simulations in unknown reef-like environments show that, under limited sensing and communication, the proposed framework achieves robust OOI-oriented navigation and cooperative coverage with improved efficiency and adaptability, narrowing the gap between semantic cognition and distributed underwater control in GPS-denied, map-free conditions.
comment: This paper has been submitted to IEEE Transactions on Mobile Computing
☆ Model-free source seeking of exponentially convergent unicycle: theoretical and robotic experimental results
This paper introduces a novel model-free, real-time unicycle-based source seeking design. This design steers autonomously the unicycle dynamic system towards the extremum point of an objective function or physical/scaler signal that is unknown expression-wise, but accessible via measurements. A key contribution of this paper is that the introduced design converges exponentially to the extremum point of objective functions (or scaler signals) that behave locally like a higher-degree power functions (e.g., fourth degree polynomial function) as opposed to locally quadratic objective functions, the usual case in literature. We provide theoretical and simulation results to support out theoretical results. Also, for the first time in the literature, we provide experimental robotic results that demonstrate the effectiveness of the proposed design and its exponential convergence ability.
♻ ☆ Dropping the D: RGB-D SLAM Without the Depth Sensor
We present DropD-SLAM, a real-time monocular SLAM system that achieves RGB-D-level accuracy without relying on depth sensors. The system replaces active depth input with three pretrained vision modules: a monocular metric depth estimator, a learned keypoint detector, and an instance segmentation network. Dynamic objects are suppressed using dilated instance masks, while static keypoints are assigned predicted depth values and backprojected into 3D to form metrically scaled features. These are processed by an unmodified RGB-D SLAM back end for tracking and mapping. On the TUM RGB-D benchmark, DropD-SLAM attains 7.4 cm mean ATE on static sequences and 1.8 cm on dynamic sequences, matching or surpassing state-of-the-art RGB-D methods while operating at 22 FPS on a single GPU. These results suggest that modern pretrained vision models can replace active depth sensors as reliable, real-time sources of metric scale, marking a step toward simpler and more cost-effective SLAM systems.
♻ ☆ Robust Trajectory Generation and Control for Quadrotor Motion Planning with Field-of-View Control Barrier Certification
Many approaches to multi-robot coordination are susceptible to failure due to communication loss and uncertainty in estimation. We present a real-time communication-free distributed navigation algorithm certified by control barrier functions, that models and controls the onboard sensing behavior to keep neighbors in the limited field of view for position estimation. The approach is robust to temporary tracking loss and directly synthesizes control to stabilize visual contact through control Lyapunov-barrier functions. The main contributions of this paper are a continuous-time robust trajectory generation and control method certified by control barrier functions for distributed multi-robot systems and a discrete optimization procedure, namely, MPC-CBF, to approximate the certified controller. In addition, we propose a linear surrogate of high-order control barrier function constraints and use sequential quadratic programming to solve MPC-CBF efficiently.
comment: 8 pages, 8 figures, 3 tables, accepted to RA-L 2025
♻ ☆ MOSPA: Human Motion Generation Driven by Spatial Audio NeurIPS 2025
Enabling virtual humans to dynamically and realistically respond to diverse auditory stimuli remains a key challenge in character animation, demanding the integration of perceptual modeling and motion synthesis. Despite its significance, this task remains largely unexplored. Most previous works have primarily focused on mapping modalities like speech, audio, and music to generate human motion. As of yet, these models typically overlook the impact of spatial features encoded in spatial audio signals on human motion. To bridge this gap and enable high-quality modeling of human movements in response to spatial audio, we introduce the first comprehensive Spatial Audio-Driven Human Motion (SAM) dataset, which contains diverse and high-quality spatial audio and motion data. For benchmarking, we develop a simple yet effective diffusion-based generative framework for human MOtion generation driven by SPatial Audio, termed MOSPA, which faithfully captures the relationship between body motion and spatial audio through an effective fusion mechanism. Once trained, MOSPA can generate diverse, realistic human motions conditioned on varying spatial audio inputs. We perform a thorough investigation of the proposed dataset and conduct extensive experiments for benchmarking, where our method achieves state-of-the-art performance on this task. Our code and model are publicly available at https://github.com/xsy27/Mospa-Acoustic-driven-Motion-Generation
comment: NeurIPS 2025 (Spotlight)
♻ ☆ Co-MTP: A Cooperative Trajectory Prediction Framework with Multi-Temporal Fusion for Autonomous Driving ICRA 2025
Vehicle-to-everything technologies (V2X) have become an ideal paradigm to extend the perception range and see through the occlusion. Exiting efforts focus on single-frame cooperative perception, however, how to capture the temporal cue between frames with V2X to facilitate the prediction task even the planning task is still underexplored. In this paper, we introduce the Co-MTP, a general cooperative trajectory prediction framework with multi-temporal fusion for autonomous driving, which leverages the V2X system to fully capture the interaction among agents in both history and future domains to benefit the planning. In the history domain, V2X can complement the incomplete history trajectory in single-vehicle perception, and we design a heterogeneous graph transformer to learn the fusion of the history feature from multiple agents and capture the history interaction. Moreover, the goal of prediction is to support future planning. Thus, in the future domain, V2X can provide the prediction results of surrounding objects, and we further extend the graph transformer to capture the future interaction among the ego planning and the other vehicles' intentions and obtain the final future scenario state under a certain planning action. We evaluate the Co-MTP framework on the real-world dataset V2X-Seq, and the results show that Co-MTP achieves state-of-the-art performance and that both history and future fusion can greatly benefit prediction.
comment: 8 pages, 3 figures, ICRA 2025
♻ ☆ Event-RGB Fusion for Spacecraft Pose Estimation Under Harsh Lighting
Spacecraft pose estimation is crucial for autonomous in-space operations, such as rendezvous, docking and on-orbit servicing. Vision-based pose estimation methods, which typically employ RGB imaging sensors, is a compelling solution for spacecraft pose estimation, but are challenged by harsh lighting conditions, which produce imaging artifacts such as glare, over-exposure, blooming and lens flare. Due to their much higher dynamic range, neuromorphic or event sensors are more resilient to extreme lighting conditions. However, event sensors generally have lower spatial resolution and suffer from reduced signal-to-noise ratio during periods of low relative motion. This work addresses these individual sensor limitations by introducing a sensor fusion approach combining RGB and event sensors. A beam-splitter prism was employed to achieve precise optical and temporal alignment. Then, a RANSAC-based technique was developed to fuse the information from the RGB and event channels to achieve pose estimation that leveraged the strengths of the two modalities. The pipeline was complemented by dropout uncertainty estimation to detect extreme conditions that affect either channel. To benchmark the performance of the proposed event-RGB fusion method, we collected a comprehensive real dataset of RGB and event data for satellite pose estimation in a laboratory setting under a variety of challenging illumination conditions. Encouraging results on the dataset demonstrate the efficacy of our event-RGB fusion approach and further supports the usage of event sensors for spacecraft pose estimation. To support community research on this topic, our dataset has been released publicly.
comment: Associated dataset: https://zenodo.org/records/15861758
Robotics 13
☆ Multi-Mapcher: Loop Closure Detection-Free Heterogeneous LiDAR Multi-Session SLAM Leveraging Outlier-Robust Registration for Autonomous Vehicles
As various 3D light detection and ranging (LiDAR) sensors have been introduced to the market, research on multi-session simultaneous localization and mapping (MSS) using heterogeneous LiDAR sensors has been actively conducted. Existing MSS methods mostly rely on loop closure detection for inter-session alignment; however, the performance of loop closure detection can be potentially degraded owing to the differences in the density and field of view (FoV) of the sensors used in different sessions. In this study, we challenge the existing paradigm that relies heavily on loop detection modules and propose a novel MSS framework, called Multi-Mapcher, that employs large-scale map-to-map registration to perform inter-session initial alignment, which is commonly assumed to be infeasible, by leveraging outlier-robust 3D point cloud registration. Next, after finding inter-session loops by radius search based on the assumption that the inter-session initial alignment is sufficiently precise, anchor node-based robust pose graph optimization is employed to build a consistent global map. As demonstrated in our experiments, our approach shows substantially better MSS performance for various LiDAR sensors used to capture the sessions and is faster than state-of-the-art approaches. Our code is available at https://github.com/url-kaist/multi-mapcher.
comment: 13 pages, 12 figures
☆ Improving Robustness to Out-of-Distribution States in Imitation Learning via Deep Koopman-Boosted Diffusion Policy
Integrating generative models with action chunking has shown significant promise in imitation learning for robotic manipulation. However, the existing diffusion-based paradigm often struggles to capture strong temporal dependencies across multiple steps, particularly when incorporating proprioceptive input. This limitation can lead to task failures, where the policy overfits to proprioceptive cues at the expense of capturing the visually derived features of the task. To overcome this challenge, we propose the Deep Koopman-boosted Dual-branch Diffusion Policy (D3P) algorithm. D3P introduces a dual-branch architecture to decouple the roles of different sensory modality combinations. The visual branch encodes the visual observations to indicate task progression, while the fused branch integrates both visual and proprioceptive inputs for precise manipulation. Within this architecture, when the robot fails to accomplish intermediate goals, such as grasping a drawer handle, the policy can dynamically switch to execute action chunks generated by the visual branch, allowing recovery to previously observed states and facilitating retrial of the task. To further enhance visual representation learning, we incorporate a Deep Koopman Operator module that captures structured temporal dynamics from visual inputs. During inference, we use the test-time loss of the generative model as a confidence signal to guide the aggregation of the temporally overlapping predicted action chunks, thereby enhancing the reliability of policy execution. In simulation experiments across six RLBench tabletop tasks, D3P outperforms the state-of-the-art diffusion policy by an average of 14.6\%. On three real-world robotic manipulation tasks, it achieves a 15.0\% improvement. Code: https://github.com/dianyeHuang/D3P.
comment: Accepted by IEEE T-RO
☆ Adaptive and Multi-object Grasping via Deformable Origami Modules
Soft robotics gripper have shown great promise in handling fragile and geometrically complex objects. However, most existing solutions rely on bulky actuators, complex control strategies, or advanced tactile sensing to achieve stable and reliable grasping performance. In this work, we present a multi-finger hybrid gripper featuring passively deformable origami modules that generate constant force and torque output. Each finger composed of parallel origami modules is driven by a 1-DoF actuator mechanism, enabling passive shape adaptability and stable grasping force without active sensing or feedback control. More importantly, we demonstrate an interesting capability in simultaneous multi-object grasping, which allows stacked objects of varied shape and size to be picked, transported and placed independently at different states, significantly improving manipulation efficiency compared to single-object grasping. These results highlight the potential of origami-based compliant structures as scalable modules for adaptive, stable and efficient multi-object manipulation in domestic and industrial pick-and-place scenarios.
☆ Descriptive Model-based Learning and Control for Bipedal Locomotion
Bipedal balance is challenging due to its multi-phase, hybrid nature and high-dimensional state space. Traditional balance control approaches for bipedal robots rely on low-dimensional models for locomotion planning and reactive control, constraining the full robot to behave like these simplified models. This involves tracking preset reference paths for the Center of Mass and upper body obtained through low-dimensional models, often resulting in inefficient walking patterns with bent knees. However, we observe that bipedal balance is inherently low-dimensional and can be effectively described with simple state and action descriptors in a low-dimensional state space. This allows the robot's motion to evolve freely in its high-dimensional state space, only constraining its projection in the low-dimensional state space. In this work, we propose a novel control approach that avoids prescribing a low-dimensional model to the full model. Instead, our control framework uses a descriptive model with the minimum degrees of freedom necessary to maintain balance, allowing the remaining degrees of freedom to evolve freely in the high-dimensional space. This results in an efficient human-like walking gait and improved robustness.
comment: 8 pages, 15 figures
☆ OmniTrack++: Omnidirectional Multi-Object Tracking by Learning Large-FoV Trajectory Feedback CVPR 2025
This paper investigates Multi-Object Tracking (MOT) in panoramic imagery, which introduces unique challenges including a 360{\deg} Field of View (FoV), resolution dilution, and severe view-dependent distortions. Conventional MOT methods designed for narrow-FoV pinhole cameras generalize unsatisfactorily under these conditions. To address panoramic distortion, large search space, and identity ambiguity under a 360{\deg} FoV, OmniTrack++ adopts a feedback-driven framework that progressively refines perception with trajectory cues. A DynamicSSM block first stabilizes panoramic features, implicitly alleviating geometric distortion. On top of normalized representations, FlexiTrack Instances use trajectory-informed feedback for flexible localization and reliable short-term association. To ensure long-term robustness, an ExpertTrack Memory consolidates appearance cues via a Mixture-of-Experts design, enabling recovery from fragmented tracks and reducing identity drift. Finally, a Tracklet Management module adaptively switches between end-to-end and tracking-by-detection modes according to scene dynamics, offering a balanced and scalable solution for panoramic MOT. To support rigorous evaluation, we establish the EmboTrack benchmark, a comprehensive dataset for panoramic MOT that includes QuadTrack, captured with a quadruped robot, and BipTrack, collected with a bipedal wheel-legged robot. Together, these datasets span wide-angle environments and diverse motion patterns, providing a challenging testbed for real-world panoramic perception. Extensive experiments on JRDB and EmboTrack demonstrate that OmniTrack++ achieves state-of-the-art performance, yielding substantial HOTA improvements of +25.5% on JRDB and +43.07% on QuadTrack over the original OmniTrack. Datasets and code will be made publicly available at https://github.com/xifen523/OmniTrack.
comment: Extended version of CVPR 2025 paper arXiv:2503.04565. Datasets and code will be made publicly available at https://github.com/xifen523/OmniTrack
☆ Design and Development of a Modular Bucket Drum Excavator for Lunar ISRU
In-Situ Resource Utilization (ISRU) is one of the key technologies for enabling sustainable access to the Moon. The ability to excavate lunar regolith is the first step in making lunar resources accessible and usable. This work presents the development of a bucket drum for the modular robotic system MoonBot, as part of the Japanese Moonshot program. A 3D-printed prototype made of PLA was manufactured to evaluate its efficiency through a series of sandbox tests. The resulting tool weighs 4.8 kg and has a volume of 14.06 L. It is capable of continuous excavation at a rate of 777.54 kg/h with a normalized energy consumption of 0.022 Wh/kg. In batch operation, the excavation rate is 172.02 kg/h with a normalized energy consumption of 0.86 Wh per kilogram of excavated material. The obtained results demonstrate the successful implementation of the concept. A key advantage of the developed tool is its compatibility with the modular MoonBot robotic platform, which enables flexible and efficient mission planning. Further improvements may include the integration of sensors and an autonomous control system to enhance the excavation process.
comment: 6 pages, 4 figures. Accepted at IEEE iSpaRo 2025
☆ Bootstrap Off-policy with World Model NeurIPS 2025
Online planning has proven effective in reinforcement learning (RL) for improving sample efficiency and final performance. However, using planning for environment interaction inevitably introduces a divergence between the collected data and the policy's actual behaviors, degrading both model learning and policy improvement. To address this, we propose BOOM (Bootstrap Off-policy with WOrld Model), a framework that tightly integrates planning and off-policy learning through a bootstrap loop: the policy initializes the planner, and the planner refines actions to bootstrap the policy through behavior alignment. This loop is supported by a jointly learned world model, which enables the planner to simulate future trajectories and provides value targets to facilitate policy improvement. The core of BOOM is a likelihood-free alignment loss that bootstraps the policy using the planner's non-parametric action distribution, combined with a soft value-weighted mechanism that prioritizes high-return behaviors and mitigates variability in the planner's action quality within the replay buffer. Experiments on the high-dimensional DeepMind Control Suite and Humanoid-Bench show that BOOM achieves state-of-the-art results in both training stability and final performance. The code is accessible at https://github.com/molumitu/BOOM_MBRL.
comment: NeurIPS 2025
☆ iFlyBot-VLA Technical Report
We introduce iFlyBot-VLA, a large-scale Vision-Language-Action (VLA) model trained under a novel framework. The main contributions are listed as follows: (1) a latent action model thoroughly trained on large-scale human and robotic manipulation videos; (2) a dual-level action representation framework that jointly supervises both the Vision-Language Model (VLM) and the action expert during training; (3) a mixed training strategy that combines robot trajectory data with general QA and spatial QA datasets, effectively enhancing the 3D perceptual and reasoning capabilities of the VLM backbone. Specifically, the VLM is trained to predict two complementary forms of actions: latent actions, derived from our latent action model pretrained on cross-embodiment manipulation data, which capture implicit high-level intentions; and structured discrete action tokens, obtained through frequency-domain transformations of continuous control signals, which encode explicit low-level dynamics. This dual supervision aligns the representation spaces of language, vision, and action, enabling the VLM to directly contribute to action generation. Experimental results on the LIBERO Franka benchmark demonstrate the superiority of our frame-work, while real-world evaluations further show that iFlyBot-VLA achieves competitive success rates across diverse and challenging manipulation tasks. Furthermore, we plan to open-source a portion of our self-constructed dataset to support future research in the community
☆ Runge-Kutta Approximations for Direct Coning Compensation Applying Lie Theory
The integration of gyroscope measurements is an essential task for most navigation systems. Modern vehicles typically use strapdown systems, such that gyro integration requires coning compensation to account for the sensor's rotation during the integration. Many coning compensation algorithms have been developed and a few are reviewed. This work introduces a new class of coning correction algorithm built directly from the classical Runge-Kutta integration routines. A simple case is shown to collapse to one of the most popular coning algorithms and a clear procedure for generating higher-order algorithms is presented.
♻ ☆ RoboOmni: Proactive Robot Manipulation in Omni-modal Context
Recent advances in Multimodal Large Language Models (MLLMs) have driven rapid progress in Vision-Language-Action (VLA) models for robotic manipulation. Although effective in many scenarios, current approaches largely rely on explicit instructions, whereas in real-world interactions, humans rarely issue instructions directly. Effective collaboration requires robots to infer user intentions proactively. In this work, we introduce cross-modal contextual instructions, a new setting where intent is derived from spoken dialogue, environmental sounds, and visual cues rather than explicit commands. To address this new setting, we present RoboOmni, a Perceiver-Thinker-Talker-Executor framework based on end-to-end omni-modal LLMs that unifies intention recognition, interaction confirmation, and action execution. RoboOmni fuses auditory and visual signals spatiotemporally for robust intention recognition, while supporting direct speech interaction. To address the absence of training data for proactive intention recognition in robotic manipulation, we build OmniAction, comprising 140k episodes, 5k+ speakers, 2.4k event sounds, 640 backgrounds, and six contextual instruction types. Experiments in simulation and real-world settings show that RoboOmni surpasses text- and ASR-based baselines in success rate, inference speed, intention recognition, and proactive assistance.
♻ ☆ Beyond the Uncanny Valley: A Mixed-Method Investigation of Anthropomorphism in Protective Responses to Robot Abuse
Robots with anthropomorphic features are increasingly shaping how humans perceive and morally engage with them. Our research investigates how different levels of anthropomorphism influence protective responses to robot abuse, extending the Computers as Social Actors (CASA) and uncanny valley theories into a moral domain. In an experiment, we invite 201 participants to view videos depicting abuse toward a robot with low (Spider), moderate (Two-Foot), or high (Humanoid) anthropomorphism. To provide a comprehensive analysis, we triangulate three modalities: self-report surveys measuring emotions and uncanniness, physiological data from automated facial expression analysis, and qualitative reflections. Findings indicate that protective responses are not linear. The moderately anthropomorphic Two-Foot robot, rated highest in eeriness and "spine-tingling" sensations consistent with the uncanny valley, elicited the strongest physiological anger expressions. Self-reported anger and guilt are significantly higher for both the Two-Foot and Humanoid robots compared to the Spider. Qualitative findings further reveal that as anthropomorphism increases, moral reasoning shifts from technical assessments of property damage to condemnation of the abuser's character, while governance proposals expand from property law to calls for quasi-animal rights and broader societal responsibility. These results suggest that the uncanny valley does not dampen moral concern but paradoxically heightens protective impulses, offering critical implications for robot design, policy, and future legal frameworks.
♻ ☆ Knolling Bot: Teaching Robots the Human Notion of Tidiness NeurIPS 2025
For robots to truly collaborate and assist humans, they must understand not only logic and instructions, but also the subtle emotions, aesthetics, and feelings that define our humanity. Human art and aesthetics are among the most elusive concepts-often difficult even for people to articulate-and without grasping these fundamentals, robots will be unable to help in many spheres of daily life. Consider the long-promised robotic butler: automating domestic chores demands more than motion planning. It requires an internal model of cleanliness and tidiness-a challenge largely unexplored by AI. To bridge this gap, we propose an approach that equips domestic robots to perform simple tidying tasks via knolling, the practice of arranging scattered items into neat, space-efficient layouts. Unlike the uniformity of industrial settings, household environments feature diverse objects and highly subjective notions of tidiness. Drawing inspiration from NLP, we treat knolling as a sequential prediction problem and employ a transformer based model to forecast each object's placement. Our method learns a generalizable concept of tidiness, generates diverse solutions adaptable to varying object sets, and incorporates human preferences for personalized arrangements. This work represents a step forward in building robots that internalize human aesthetic sense and can genuinely co-create in our living spaces.
comment: Accepted at the 39th Conference on Neural Information Processing Systems (NeurIPS 2025) Creative AI Track
♻ ☆ MindJourney: Test-Time Scaling with World Models for Spatial Reasoning
Spatial reasoning in 3D space is central to human cognition and indispensable for embodied tasks such as navigation and manipulation. However, state-of-the-art vision-language models (VLMs) struggle frequently with tasks as simple as anticipating how a scene will look after an egocentric motion: they perceive 2D images but lack an internal model of 3D dynamics. We therefore propose MindJourney, a test-time scaling framework that grants a VLM with this missing capability by coupling it to a controllable world model based on video diffusion. The VLM iteratively sketches a concise camera trajectory, while the world model synthesizes the corresponding view at each step. The VLM then reasons over this multi-view evidence gathered during the interactive exploration. Without any fine-tuning, our MindJourney achieves over an average 7.7% performance boost on the representative spatial reasoning benchmark SAT, showing that pairing VLMs with world models for test-time scaling offers a simple, plug-and-play route to robust 3D reasoning. Meanwhile, our method also improves upon the test-time inference VLMs trained through reinforcement learning, which demonstrates the potential of our method that utilizes world models for test-time scaling.
comment: Project Page: https://umass-embodied-agi.github.io/MindJourney
Robotics 27
☆ Whole-Body Proprioceptive Morphing: A Modular Soft Gripper for Robust Cross-Scale Grasping
Biological systems, such as the octopus, exhibit masterful cross-scale manipulation by adaptively reconfiguring their entire form, a capability that remains elusive in robotics. Conventional soft grippers, while compliant, are mostly constrained by a fixed global morphology, and prior shape-morphing efforts have been largely confined to localized deformations, failing to replicate this biological dexterity. Inspired by this natural exemplar, we introduce the paradigm of collaborative, whole-body proprioceptive morphing, realized in a modular soft gripper architecture. Our design is a distributed network of modular self-sensing pneumatic actuators that enables the gripper to intelligently reconfigure its entire topology, achieving multiple morphing states that are controllable to form diverse polygonal shapes. By integrating rich proprioceptive feedback from embedded sensors, our system can seamlessly transition from a precise pinch to a large envelope grasp. We experimentally demonstrate that this approach expands the grasping envelope and enhances generalization across diverse object geometries (standard and irregular) and scales (up to 10$\times$), while also unlocking novel manipulation modalities such as multi-object and internal hook grasping. This work presents a low-cost, easy-to-fabricate, and scalable framework that fuses distributed actuation with integrated sensing, offering a new pathway toward achieving biological levels of dexterity in robotic manipulation.
☆ Dual-Stream Diffusion for World-Model Augmented Vision-Language-Action Model
Recently, augmenting Vision-Language-Action models (VLAs) with world modeling has shown promise in improving robotic policy learning. However, it remains challenging to jointly predict next-state observations and action sequences because of the inherent difference between the two modalities. To address this, we propose DUal-STream diffusion (DUST), a world-model augmented VLA framework that handles the modality conflict and enhances the performance of VLAs across diverse tasks. Specifically, we propose a multimodal diffusion transformer architecture that explicitly maintains separate modality streams while still enabling cross-modal knowledge sharing. In addition, we introduce independent noise perturbations for each modality and a decoupled flow-matching loss. This design enables the model to learn the joint distribution in a bidirectional manner while avoiding the need for a unified latent space. Based on the decoupling of modalities during training, we also introduce a joint sampling method that supports test-time scaling, where action and vision tokens evolve asynchronously at different rates. Through experiments on simulated benchmarks such as RoboCasa and GR-1, DUST achieves up to 6% gains over baseline methods, while our test-time scaling approach provides an additional 2-5% boost. On real-world tasks with the Franka Research 3, DUST improves success rates by 13%, confirming its effectiveness beyond simulation. Furthermore, pre-training on action-free videos from BridgeV2 yields significant transfer gains on RoboCasa, underscoring DUST's potential for large-scale VLA pretraining.
comment: 20 pages, 10 figures
☆ Toward Accurate Long-Horizon Robotic Manipulation: Language-to-Action with Foundation Models via Scene Graphs
This paper presents a framework that leverages pre-trained foundation models for robotic manipulation without domain-specific training. The framework integrates off-the-shelf models, combining multimodal perception from foundation models with a general-purpose reasoning model capable of robust task sequencing. Scene graphs, dynamically maintained within the framework, provide spatial awareness and enable consistent reasoning about the environment. The framework is evaluated through a series of tabletop robotic manipulation experiments, and the results highlight its potential for building robotic manipulation systems directly on top of off-the-shelf foundation models.
☆ EBT-Policy: Energy Unlocks Emergent Physical Reasoning Capabilities
Implicit policies parameterized by generative models, such as Diffusion Policy, have become the standard for policy learning and Vision-Language-Action (VLA) models in robotics. However, these approaches often suffer from high computational cost, exposure bias, and unstable inference dynamics, which lead to divergence under distribution shifts. Energy-Based Models (EBMs) address these issues by learning energy landscapes end-to-end and modeling equilibrium dynamics, offering improved robustness and reduced exposure bias. Yet, policies parameterized by EBMs have historically struggled to scale effectively. Recent work on Energy-Based Transformers (EBTs) demonstrates the scalability of EBMs to high-dimensional spaces, but their potential for solving core challenges in physically embodied models remains underexplored. We introduce a new energy-based architecture, EBT-Policy, that solves core issues in robotic and real-world settings. Across simulated and real-world tasks, EBT-Policy consistently outperforms diffusion-based policies, while requiring less training and inference computation. Remarkably, on some tasks it converges within just two inference steps, a 50x reduction compared to Diffusion Policy's 100. Moreover, EBT-Policy exhibits emergent capabilities not seen in prior models, such as zero-shot recovery from failed action sequences using only behavior cloning and without explicit retry training. By leveraging its scalar energy for uncertainty-aware inference and dynamic compute allocation, EBT-Policy offers a promising path toward robust, generalizable robot behavior under distribution shifts.
comment: 9 pages, 6 figures, 4 tables
☆ Preliminary Prototyping of Avoidance Behaviors Triggered by a User's Physical Approach to a Robot
Human-robot interaction frequently involves physical proximity or contact. In human-human settings, people flexibly accept, reject, or tolerate such approaches depending on the relationship and context. We explore the design of a robot's rejective internal state and corresponding avoidance behaviors, such as withdrawing or pushing away, when a person approaches. We model the accumulation and decay of discomfort as a function of interpersonal distance, and implement tolerance (endurance) and limit-exceeding avoidance driven by the Dominance axis of the PAD affect model. The behaviors and their intensities are realized on an arm robot. Results illustrate a coherent pipeline from internal state parameters to graded endurance motions and, once a limit is crossed, to avoidance actions.
comment: Workshop on Socially Aware and Cooperative Intelligent Systems in HAI 2025
☆ Learning Soft Robotic Dynamics with Active Exploration
Soft robots offer unmatched adaptability and safety in unstructured environments, yet their compliant, high-dimensional, and nonlinear dynamics make modeling for control notoriously difficult. Existing data-driven approaches often fail to generalize, constrained by narrowly focused task demonstrations or inefficient random exploration. We introduce SoftAE, an uncertainty-aware active exploration framework that autonomously learns task-agnostic and generalizable dynamics models of soft robotic systems. SoftAE employs probabilistic ensemble models to estimate epistemic uncertainty and actively guides exploration toward underrepresented regions of the state-action space, achieving efficient coverage of diverse behaviors without task-specific supervision. We evaluate SoftAE on three simulated soft robotic platforms -- a continuum arm, an articulated fish in fluid, and a musculoskeletal leg with hybrid actuation -- and on a pneumatically actuated continuum soft arm in the real world. Compared with random exploration and task-specific model-based reinforcement learning, SoftAE produces more accurate dynamics models, enables superior zero-shot control on unseen tasks, and maintains robustness under sensing noise, actuation delays, and nonlinear material effects. These results demonstrate that uncertainty-driven active exploration can yield scalable, reusable dynamics models across diverse soft robotic morphologies, representing a step toward more autonomous, adaptable, and data-efficient control in compliant robots.
☆ Towards a Multi-Embodied Grasping Agent
Multi-embodiment grasping focuses on developing approaches that exhibit generalist behavior across diverse gripper designs. Existing methods often learn the kinematic structure of the robot implicitly and face challenges due to the difficulty of sourcing the required large-scale data. In this work, we present a data-efficient, flow-based, equivariant grasp synthesis architecture that can handle different gripper types with variable degrees of freedom and successfully exploit the underlying kinematic model, deducing all necessary information solely from the gripper and scene geometry. Unlike previous equivariant grasping methods, we translated all modules from the ground up to JAX and provide a model with batching capabilities over scenes, grippers, and grasps, resulting in smoother learning, improved performance and faster inference time. Our dataset encompasses grippers ranging from humanoid hands to parallel yaw grippers and includes 25,000 scenes and 20 million grasps.
comment: 9 pages, 3 figures
☆ Modified-Emergency Index (MEI): A Criticality Metric for Autonomous Driving in Lateral Conflict
Effective, reliable, and efficient evaluation of autonomous driving safety is essential to demonstrate its trustworthiness. Criticality metrics provide an objective means of assessing safety. However, as existing metrics primarily target longitudinal conflicts, accurately quantifying the risks of lateral conflicts - prevalent in urban settings - remains challenging. This paper proposes the Modified-Emergency Index (MEI), a metric designed to quantify evasive effort in lateral conflicts. Compared to the original Emergency Index (EI), MEI refines the estimation of the time available for evasive maneuvers, enabling more precise risk quantification. We validate MEI on a public lateral conflict dataset based on Argoverse-2, from which we extract over 1,500 high-quality AV conflict cases, including more than 500 critical events. MEI is then compared with the well-established ACT and the widely used PET metrics. Results show that MEI consistently outperforms them in accurately quantifying criticality and capturing risk evolution. Overall, these findings highlight MEI as a promising metric for evaluating urban conflicts and enhancing the safety assessment framework for autonomous driving. The open-source implementation is available at https://github.com/AutoChengh/MEI.
☆ A Modular and Scalable System Architecture for Heterogeneous UAV Swarms Using ROS 2 and PX4-Autopilot
In this paper a modular and scalable architecture for heterogeneous swarm-based Counter Unmanned Aerial Systems (C-UASs) built on PX4-Autopilot and Robot Operating System 2 (ROS 2) framework is presented. The proposed architecture emphasizes seamless integration of hardware components by introducing independent ROS 2 nodes for each component of a Unmanned Aerial Vehicle (UAV). Communication between swarm participants is abstracted in software, allowing the use of various technologies without architectural changes. Key functionalities are supported, e.g. leader following and formation flight to maneuver the swarm. The system also allows computer vision algorithms to be integrated for the detection and tracking of UAVs. Additionally, a ground station control is integrated for the coordination of swarm operations. Swarm-based Unmanned Aerial System (UAS) architecture is verified within a Gazebo simulation environment but also in real-world demonstrations.
☆ Vectorized Online POMDP Planning ICRA 2026
Planning under partial observability is an essential capability of autonomous robots. The Partially Observable Markov Decision Process (POMDP) provides a powerful framework for planning under partial observability problems, capturing the stochastic effects of actions and the limited information available through noisy observations. POMDP solving could benefit tremendously from massive parallelization of today's hardware, but parallelizing POMDP solvers has been challenging. They rely on interleaving numerical optimization over actions with the estimation of their values, which creates dependencies and synchronization bottlenecks between parallel processes that can quickly offset the benefits of parallelization. In this paper, we propose Vectorized Online POMDP Planner (VOPP), a novel parallel online solver that leverages a recent POMDP formulation that analytically solves part of the optimization component, leaving only the estimation of expectations for numerical computation. VOPP represents all data structures related to planning as a collection of tensors and implements all planning steps as fully vectorized computations over this representation. The result is a massively parallel solver with no dependencies and synchronization bottlenecks between parallel computations. Experimental results indicate that VOPP is at least 20X more efficient in computing near-optimal solutions compared to an existing state-of-the-art parallel online solver.
comment: 8 pages, 3 figures. Submitted to ICRA 2026
☆ Hybrid Gripper Finger Enabling In-Grasp Friction Modulation Using Inflatable Silicone Pockets ICRA 2026
Grasping objects with diverse mechanical properties, such as heavy, slippery, or fragile items, remains a significant challenge in robotics. Conventional grippers often rely on applying high normal forces, which can cause damage to objects. To address this limitation, we present a hybrid gripper finger that combines a rigid structural shell with a soft, inflatable silicone pocket. The gripper finger can actively modulate its surface friction by controlling the internal air pressure of the silicone pocket. Results from fundamental experiments indicate that increasing the internal pressure results in a proportional increase in the effective coefficient of friction. This enables the gripper to stably lift heavy and slippery objects without increasing the gripping force and to handle fragile or deformable objects, such as eggs, fruits, and paper cups, with minimal damage by increasing friction rather than applying excessive force. The experimental results demonstrate that the hybrid gripper finger with adaptable friction provides a robust and safer alternative to relying solely on high normal forces, thereby enhancing the gripper flexibility in handling delicate, fragile, and diverse objects.
comment: Submitted to ICRA 2026
☆ MobiDock: Design and Control of A Modular Self Reconfigurable Bimanual Mobile Manipulator via Robotic Docking ICRA2026
Multi-robot systems, particularly mobile manipulators, face challenges in control coordination and dynamic stability when working together. To address this issue, this study proposes MobiDock, a modular self-reconfigurable mobile manipulator system that allows two independent robots to physically connect and form a unified mobile bimanual platform. This process helps transform a complex multi-robot control problem into the management of a simpler, single system. The system utilizes an autonomous docking strategy based on computer vision with AprilTag markers and a new threaded screw-lock mechanism. Experimental results show that the docked configuration demonstrates better performance in dynamic stability and operational efficiency compared to two independently cooperating robots. Specifically, the unified system has lower Root Mean Square (RMS) Acceleration and Jerk values, higher angular precision, and completes tasks significantly faster. These findings confirm that physical reconfiguration is a powerful design principle that simplifies cooperative control, improving stability and performance for complex tasks in real-world environments.
comment: ICRA2026 submited
☆ Confined Space Underwater Positioning Using Collaborative Robots
Positioning of underwater robots in confined and cluttered spaces remains a key challenge for field operations. Existing systems are mostly designed for large, open-water environments and struggle in industrial settings due to poor coverage, reliance on external infrastructure, and the need for feature-rich surroundings. Multipath effects from continuous sound reflections further degrade signal quality, reducing accuracy and reliability. Accurate and easily deployable positioning is essential for repeatable autonomous missions; however, this requirement has created a technological bottleneck limiting underwater robotic deployment. This paper presents the Collaborative Aquatic Positioning (CAP) system, which integrates collaborative robotics and sensor fusion to overcome these limitations. Inspired by the "mother-ship" concept, the surface vehicle acts as a mobile leader to assist in positioning a submerged robot, enabling localization even in GPS-denied and highly constrained environments. The system is validated in a large test tank through repeatable autonomous missions using CAP's position estimates for real-time trajectory control. Experimental results demonstrate a mean Euclidean distance (MED) error of 70 mm, achieved in real time without requiring fixed infrastructure, extensive calibration, or environmental features. CAP leverages advances in mobile robot sensing and leader-follower control to deliver a step change in accurate, practical, and infrastructure-free underwater localization.
comment: 31 pages including appendix, 24 figures
☆ WildfireX-SLAM: A Large-scale Low-altitude RGB-D Dataset for Wildfire SLAM and Beyond
3D Gaussian splatting (3DGS) and its subsequent variants have led to remarkable progress in simultaneous localization and mapping (SLAM). While most recent 3DGS-based SLAM works focus on small-scale indoor scenes, developing 3DGS-based SLAM methods for large-scale forest scenes holds great potential for many real-world applications, especially for wildfire emergency response and forest management. However, this line of research is impeded by the absence of a comprehensive and high-quality dataset, and collecting such a dataset over real-world scenes is costly and technically infeasible. To this end, we have built a large-scale, comprehensive, and high-quality synthetic dataset for SLAM in wildfire and forest environments. Leveraging the Unreal Engine 5 Electric Dreams Environment Sample Project, we developed a pipeline to easily collect aerial and ground views, including ground-truth camera poses and a range of additional data modalities from unmanned aerial vehicle. Our pipeline also provides flexible controls on environmental factors such as light, weather, and types and conditions of wildfire, supporting the need for various tasks covering forest mapping, wildfire emergency response, and beyond. The resulting pilot dataset, WildfireX-SLAM, contains 5.5k low-altitude RGB-D aerial images from a large-scale forest map with a total size of 16 km2. On top of WildfireX-SLAM, a thorough benchmark is also conducted, which not only reveals the unique challenges of 3DGS-based SLAM in the forest but also highlights potential improvements for future works. The dataset and code will be publicly available. Project page: https://zhicongsun.github.io/wildfirexslam.
comment: This paper has been accepted by MMM 2026
☆ Learning Generalizable Visuomotor Policy through Dynamics-Alignment
Behavior cloning methods for robot learning suffer from poor generalization due to limited data support beyond expert demonstrations. Recent approaches leveraging video prediction models have shown promising results by learning rich spatiotemporal representations from large-scale datasets. However, these models learn action-agnostic dynamics that cannot distinguish between different control inputs, limiting their utility for precise manipulation tasks and requiring large pretraining datasets. We propose a Dynamics-Aligned Flow Matching Policy (DAP) that integrates dynamics prediction into policy learning. Our method introduces a novel architecture where policy and dynamics models provide mutual corrective feedback during action generation, enabling self-correction and improved generalization. Empirical validation demonstrates generalization performance superior to baseline methods on real-world robotic manipulation tasks, showing particular robustness in OOD scenarios including visual distractions and lighting variations.
comment: 9 pages, 6 figures
♻ ☆ RObotic MAnipulation Network (ROMAN) -- Hybrid Hierarchical Learning for Solving Complex Sequential Tasks
Solving long sequential tasks poses a significant challenge in embodied artificial intelligence. Enabling a robotic system to perform diverse sequential tasks with a broad range of manipulation skills is an active area of research. In this work, we present a Hybrid Hierarchical Learning framework, the Robotic Manipulation Network (ROMAN), to address the challenge of solving multiple complex tasks over long time horizons in robotic manipulation. ROMAN achieves task versatility and robust failure recovery by integrating behavioural cloning, imitation learning, and reinforcement learning. It consists of a central manipulation network that coordinates an ensemble of various neural networks, each specialising in distinct re-combinable sub-tasks to generate their correct in-sequence actions for solving complex long-horizon manipulation tasks. Experimental results show that by orchestrating and activating these specialised manipulation experts, ROMAN generates correct sequential activations for accomplishing long sequences of sophisticated manipulation tasks and achieving adaptive behaviours beyond demonstrations, while exhibiting robustness to various sensory noises. These results demonstrate the significance and versatility of ROMAN's dynamic adaptability featuring autonomous failure recovery capabilities, and highlight its potential for various autonomous manipulation tasks that demand adaptive motor skills.
comment: To appear in Nature Machine Intelligence. Includes the main and supplementary manuscript. Total of 70 pages, with a total of 9 Figures and 17 Tables
♻ ☆ GenSwarm: Scalable Multi-Robot Code-Policy Generation and Deployment via Language Models
The development of control policies for multi-robot systems traditionally follows a complex and labor-intensive process, often lacking the flexibility to adapt to dynamic tasks. This has motivated research on methods to automatically create control policies. However, these methods require iterative processes of manually crafting and refining objective functions, thereby prolonging the development cycle. This work introduces \textit{GenSwarm}, an end-to-end system that leverages large language models to automatically generate and deploy control policies for multi-robot tasks based on simple user instructions in natural language. As a multi-language-agent system, GenSwarm achieves zero-shot learning, enabling rapid adaptation to altered or unseen tasks. The white-box nature of the code policies ensures strong reproducibility and interpretability. With its scalable software and hardware architectures, GenSwarm supports efficient policy deployment on both simulated and real-world multi-robot systems, realizing an instruction-to-execution end-to-end functionality that could prove valuable for robotics specialists and non-specialists alike.The code of the proposed GenSwarm system is available online: https://github.com/WindyLab/GenSwarm.
comment: This article has been accepted for publication in npj Robotics
♻ ☆ A Study on Human-Swarm Interaction: A Framework for Assessing Situation Awareness and Task Performance
This paper introduces a framework for human swarm interaction studies that measures situation awareness in dynamic environments. A tablet-based interface was developed for a user study by implementing the concepts introduced in the framework, where operators guided a robotic swarm in a single-target search task, marking hazardous cells unknown to the swarm. Both subjective and objective situation awareness measures were used, with task performance evaluated based on how close the robots were to the target. The framework enabled a structured investigation of the role of situation awareness in human swarm interaction, leading to key findings such as improved task performance across attempts, showing the interface was learnable, centroid active robot position proved to be a useful task performance metric for assessing situation awareness, perception and projection played a key role in task performance, highlighting their importance in interface design and objective situation awareness influenced both subjective situation awareness and task performance, emphasizing the need for interfaces that emphasise objective situation awareness. These findings validate our framework as a structured approach for integrating situation awareness concepts into human swarm interaction studies, offering a systematic way to assess situation awareness and task performance. The framework can be applied to other swarming studies to evaluate interface learnability, identify meaningful task performance metrics, and refine interface designs to enhance situation awareness, ultimately improving human swarm interaction in dynamic environments.
comment: 10 pages, 8 figures, 2 tables, 2 equations
♻ ☆ Uncertainty-Based Smooth Policy Regularisation for Reinforcement Learning with Few Demonstrations
In reinforcement learning with sparse rewards, demonstrations can accelerate learning, but determining when to imitate them remains challenging. We propose Smooth Policy Regularisation from Demonstrations (SPReD), a framework that addresses the fundamental question: when should an agent imitate a demonstration versus follow its own policy? SPReD uses ensemble methods to explicitly model Q-value distributions for both demonstration and policy actions, quantifying uncertainty for comparisons. We develop two complementary uncertainty-aware methods: a probabilistic approach estimating the likelihood of demonstration superiority, and an advantage-based approach scaling imitation by statistical significance. Unlike prevailing methods (e.g. Q-filter) that make binary imitation decisions, SPReD applies continuous, uncertainty-proportional regularisation weights, reducing gradient variance during training. Despite its computational simplicity, SPReD achieves remarkable gains in experiments across eight robotics tasks, outperforming existing approaches by up to a factor of 14 in complex tasks while maintaining robustness to demonstration quality and quantity. Our code is available at https://github.com/YujieZhu7/SPReD.
♻ ☆ A Tactile Feedback Approach to Path Recovery after High-Speed Impacts for Collision-Resilient Drones
Aerial robots are a well-established solution for exploration, monitoring, and inspection, thanks to their superior maneuverability and agility. However, in many environments, they risk crashing and sustaining damage after collisions. Traditional methods focus on avoiding obstacles entirely, but these approaches can be limiting, particularly in cluttered spaces or on weight-and compute-constrained platforms such as drones. This paper presents a novel approach to enhance drone robustness and autonomy by developing a path recovery and adjustment method for a high-speed collision-resilient aerial robot equipped with lightweight, distributed tactile sensors. The proposed system explicitly models collisions using pre-collision velocities, rates and tactile feedback to predict post-collision dynamics, improving state estimation accuracy. Additionally, we introduce a computationally efficient vector-field-based path representation that guarantees convergence to a user-specified path, while naturally avoiding known obstacles. Post-collision, contact point locations are incorporated into the vector field as a repulsive potential, enabling the drone to avoid obstacles while naturally returning to its path. The effectiveness of this method is validated through Monte Carlo simulations and demonstrated on a physical prototype, showing successful path following, collision recovery, and adjustment at speeds up to 3.7 m/s.
♻ ☆ SafeAgentBench: A Benchmark for Safe Task Planning of Embodied LLM Agents
With the integration of large language models (LLMs), embodied agents have strong capabilities to understand and plan complicated natural language instructions. However, a foreseeable issue is that those embodied agents can also flawlessly execute some hazardous tasks, potentially causing damages in the real world. Existing benchmarks predominantly overlook critical safety risks, focusing solely on planning performance, while a few evaluate LLMs' safety awareness only on non-interactive image-text data. To address this gap, we present SafeAgentBench -- the first comprehensive benchmark for safety-aware task planning of embodied LLM agents in interactive simulation environments, covering both explicit and implicit hazards. SafeAgentBench includes: (1) an executable, diverse, and high-quality dataset of 750 tasks, rigorously curated to cover 10 potential hazards and 3 task types; (2) SafeAgentEnv, a universal embodied environment with a low-level controller, supporting multi-agent execution with 17 high-level actions for 9 state-of-the-art baselines; and (3) reliable evaluation methods from both execution and semantic perspectives. Experimental results show that, although agents based on different design frameworks exhibit substantial differences in task success rates, their overall safety awareness remains weak. The most safety-conscious baseline achieves only a 10% rejection rate for detailed hazardous tasks. Moreover, simply replacing the LLM driving the agent does not lead to notable improvements in safety awareness. Dataset and codes are available in https://github.com/shengyin1224/SafeAgentBench and https://huggingface.co/datasets/safeagentbench/SafeAgentBench.
comment: 28 pages, 19 tables, 15 figures
♻ ☆ From Canada to Japan: How 10,000 km Affect User Perception in Robot Teleoperation
Robot teleoperation (RTo) has emerged as a viable alternative to local control, particularly when human intervention is still necessary. This research aims to study the distance effect on user perception in RTo, exploring the potential of teleoperated robots for older adult care. We propose an evaluation of non-expert users' perception of long-distance RTo, examining how their perception changes before and after interaction, as well as comparing it to that of locally operated robots. We have designed a specific protocol consisting of multiple questionnaires, along with a dedicated software architecture using the Robotics Operating System (ROS) and Unity. The results revealed no statistically significant differences between the local and remote robot conditions, suggesting that robots may be a viable alternative to traditional local control.
comment: Author preprint - Accepted for Humanoids 2025
♻ ☆ A Practical-Driven Framework for Transitioning Drive-by-Wire to Autonomous Driving Systems: A Case Study with a Chrysler Pacifica Hybrid Vehicle
Transitioning from a Drive-by-Wire (DBW) system to a fully autonomous driving system (ADS) involves multiple stages of development and demands robust positioning and sensing capabilities. This paper presents a practice-driven framework for facilitating the DBW-to-ADS transition using a 2022 Chrysler Pacifica Hybrid Minivan equipped with cameras, LiDAR, GNSS, and onboard computing hardware configured with the Robot Operating System (ROS) and Autoware.AI. The implementation showcases offline autonomous operations utilizing pre-recorded LiDAR and camera data, point clouds, and vector maps, enabling effective localization and path planning within a structured test environment. The study addresses key challenges encountered during the transition, particularly those related to wireless-network-assisted sensing and positioning. It offers practical solutions for overcoming software incompatibility constraints, sensor synchronization issues, and limitations in real-time perception. Furthermore, the integration of sensing, data fusion, and automation is emphasized as a critical factor in supporting autonomous driving systems in map generation, simulation, and training. Overall, the transition process outlined in this work aims to provide actionable strategies for researchers pursuing DBW-to-ADS conversion. It offers direction for incorporating real-time perception, GNSS-LiDAR-camera integration, and fully ADS-equipped autonomous vehicle operations, thus contributing to the advancement of robust autonomous vehicle technologies.
comment: This updated version includes further implementation details and experimental validation. Accepted for presentation at The 22nd International Conference on Automation Technology (AUTOMATION 2025), Taipei, Taiwan, November 2025
♻ ☆ Panoramic Out-of-Distribution Segmentation for Autonomous Driving
Panoramic imaging enables capturing 360{\deg} images with an ultra-wide Field-of-View (FoV) for dense omnidirectional perception, which is critical to applications, such as autonomous driving and augmented reality, etc. However, current panoramic semantic segmentation methods fail to identify outliers, and pinhole Out-of-distribution Segmentation (OoS) models perform unsatisfactorily in the panoramic domain due to background clutter and pixel distortions. To address these issues, we introduce a new task, Panoramic Out-of-distribution Segmentation (PanOoS), with the aim of achieving comprehensive and safe scene understanding. Furthermore, we propose the first solution, POS, which adapts to the characteristics of panoramic images through text-guided prompt distribution learning. Specifically, POS integrates a disentanglement strategy designed to materialize the cross-domain generalization capability of CLIP. The proposed Prompt-based Restoration Attention (PRA) optimizes semantic decoding by prompt guidance and self-adaptive correction, while Bilevel Prompt Distribution Learning (BPDL) refines the manifold of per-pixel mask embeddings via semantic prototype supervision. Besides, to compensate for the scarcity of PanOoS datasets, we establish two benchmarks: DenseOoS, which features diverse outliers in complex environments, and QuadOoS, captured by a quadruped robot with a panoramic annular lens system. Extensive experiments demonstrate superior performance of POS, with AuPRC improving by 34.25% and FPR95 decreasing by 21.42% on DenseOoS, outperforming state-of-the-art pinhole-OoS methods. Moreover, POS achieves leading closed-set segmentation capabilities and advances the development of panoramic understanding. Code and datasets will be available at https://github.com/MengfeiD/PanOoS.
comment: Code and datasets will be available at https://github.com/MengfeiD/PanOoS
♻ ☆ Sim2Real Diffusion: Leveraging Foundation Vision Language Models for Adaptive Automated Driving
Simulation-based design, optimization, and validation of autonomous vehicles have proven to be crucial for their improvement over the years. Nevertheless, the ultimate measure of effectiveness is their successful transition from simulation to reality (sim2real). However, existing sim2real transfer methods struggle to address the autonomy-oriented requirements of balancing: (i) conditioned domain adaptation, (ii) robust performance with limited examples, (iii) modularity in handling multiple domain representations, and (iv) real-time performance. To alleviate these pain points, we present a unified framework for learning cross-domain adaptive representations through conditional latent diffusion for sim2real transferable automated driving. Our framework offers options to leverage: (i) alternate foundation models, (ii) a few-shot fine-tuning pipeline, and (iii) textual as well as image prompts for mapping across given source and target domains. It is also capable of generating diverse high-quality samples when diffusing across parameter spaces such as times of day, weather conditions, seasons, and operational design domains. We systematically analyze the presented framework and report our findings in terms of performance benchmarks and ablation studies. Additionally, we demonstrate its serviceability for autonomous driving using behavioral cloning case studies. Our experiments indicate that the proposed framework is capable of bridging the perceptual sim2real gap by over 40%.
comment: Accepted in IEEE Robotics and Automation Letters (RA-L)
♻ ☆ Faster Model Predictive Control via Self-Supervised Initialization Learning
Model Predictive Control (MPC) is widely used in robot control by optimizing a sequence of control outputs over a finite-horizon. Computational approaches for MPC include deterministic methods (e.g., iLQR and COBYLA), as well as sampling-based methods (e.g., MPPI and CEM). However, complex system dynamics and non-convex or non-differentiable cost terms often lead to prohibitive optimization times that limit real-world deployment. Prior efforts to accelerate MPC have limitations on: (i) reusing previous solutions fails under sharp state changes and (ii) pure imitation learning does not target compute efficiency directly and suffers from suboptimality in the training data. To address these, We propose a warm-start framework that learns a policy to generate high-quality initial guesses for MPC solver. The policy is first trained via behavior cloning from expert MPC rollouts and then fine-tuned online with reinforcement learning to directly minimize MPC optimization time. We empirically validate that our approach improves both deterministic and sampling-based MPC methods, achieving up to 21.6% faster optimization and 34.1% more tracking accuracy for deterministic MPC in Formula 1 track path-tracking domain, and improving safety by 100%, path efficiency by 12.8%, and steering smoothness by 7.2% for sampling-based MPC in obstacle-rich navigation domain. These results demonstrate that our framework not only accelerates MPC but also improves overall control performance. Furthermore, it can be applied to a broader range of control algorithms that benefit from good initial guesses.
♻ ☆ Curvature-Aware Calibration of Tactile Sensors for Accurate Force Estimation on Non-Planar Surfaces
Flexible tactile sensors are increasingly used in real-world applications such as robotic grippers, prosthetic hands, wearable gloves, and assistive devices, where they need to conform to curved and irregular surfaces. However, most existing tactile sensors are calibrated only on flat substrates, and their accuracy and consistency degrade once mounted on curved geometries. This limitation restricts their reliability in practical use. To address this challenge, we develop a calibration model for a widely used resistive tactile sensor design that enables accurate force estimation on one-dimensional curved surfaces. We then train a neural network (a multilayer perceptron) to predict local curvature from baseline sensor outputs recorded under no applied load, achieving an R2 score of 0.91. The proposed approach is validated on five daily objects with varying curvatures under forces from 2 N to 8 N. Results show that the curvature-aware calibration maintains consistent force accuracy across all surfaces, while flat-surface calibration underestimates force as curvature increases. Our results demonstrate that curvature-aware modeling improves the accuracy, consistency, and reliability of flexible tactile sensors, enabling dependable performance across real-world applications.
comment: This work has been submitted to the IEEE for possible publication
Robotics 56
☆ Running VLAs at Real-time Speed
In this paper, we show how to run pi0-level multi-view VLA at 30Hz frame rate and at most 480Hz trajectory frequency using a single consumer GPU. This enables dynamic and real-time tasks that were previously believed to be unattainable by large VLA models. To achieve it, we introduce a bag of strategies to eliminate the overheads in model inference. The real-world experiment shows that the pi0 policy with our strategy achieves a 100% success rate in grasping a falling pen task. Based on the results, we further propose a full streaming inference framework for real-time robot control of VLA. Code is available at https://github.com/Dexmal/realtime-vla.
comment: Code is available at https://github.com/Dexmal/realtime-vla
☆ Hybrid Consistency Policy: Decoupling Multi-Modal Diversity and Real-Time Efficiency in Robotic Manipulation
In visuomotor policy learning, diffusion-based imitation learning has become widely adopted for its ability to capture diverse behaviors. However, approaches built on ordinary and stochastic denoising processes struggle to jointly achieve fast sampling and strong multi-modality. To address these challenges, we propose the Hybrid Consistency Policy (HCP). HCP runs a short stochastic prefix up to an adaptive switch time, and then applies a one-step consistency jump to produce the final action. To align this one-jump generation, HCP performs time-varying consistency distillation that combines a trajectory-consistency objective to keep neighboring predictions coherent and a denoising-matching objective to improve local fidelity. In both simulation and on a real robot, HCP with 25 SDE steps plus one jump approaches the 80-step DDPM teacher in accuracy and mode coverage while significantly reducing latency. These results show that multi-modality does not require slow inference, and a switch time decouples mode retention from speed. It yields a practical accuracy efficiency trade-off for robot policies.
☆ Heuristic Adaptation of Potentially Misspecified Domain Support for Likelihood-Free Inference in Stochastic Dynamical Systems
In robotics, likelihood-free inference (LFI) can provide the domain distribution that adapts a learnt agent in a parametric set of deployment conditions. LFI assumes an arbitrary support for sampling, which remains constant as the initial generic prior is iteratively refined to more descriptive posteriors. However, a potentially misspecified support can lead to suboptimal, yet falsely certain, posteriors. To address this issue, we propose three heuristic LFI variants: EDGE, MODE, and CENTRE. Each interprets the posterior mode shift over inference steps in its own way and, when integrated into an LFI step, adapts the support alongside posterior inference. We first expose the support misspecification issue and evaluate our heuristics using stochastic dynamical benchmarks. We then evaluate the impact of heuristic support adaptation on parameter inference and policy learning for a dynamic deformable linear object (DLO) manipulation task. Inference results in a finer length and stiffness classification for a parametric set of DLOs. When the resulting posteriors are used as domain distributions for sim-based policy learning, they lead to more robust object-centric agent performance.
☆ Hybrid DQN-TD3 Reinforcement Learning for Autonomous Navigation in Dynamic Environments
This paper presents a hierarchical path-planning and control framework that combines a high-level Deep Q-Network (DQN) for discrete sub-goal selection with a low-level Twin Delayed Deep Deterministic Policy Gradient (TD3) controller for continuous actuation. The high-level module selects behaviors and sub-goals; the low-level module executes smooth velocity commands. We design a practical reward shaping scheme (direction, distance, obstacle avoidance, action smoothness, collision penalty, time penalty, and progress), together with a LiDAR-based safety gate that prevents unsafe motions. The system is implemented in ROS + Gazebo (TurtleBot3) and evaluated with PathBench metrics, including success rate, collision rate, path efficiency, and re-planning efficiency, in dynamic and partially observable environments. Experiments show improved success rate and sample efficiency over single-algorithm baselines (DQN or TD3 alone) and rule-based planners, with better generalization to unseen obstacle configurations and reduced abrupt control changes. Code and evaluation scripts are available at the project repository.
comment: 6 pages, 5 figures; ROS+Gazebo (TurtleBot3) implementation; evaluation with PathBench metrics; code (primary): https://github.com/MayaCHEN-github/HierarchicalRL-robot-navigation; mirror (for reproducibility): https://github.com/ShowyHe/DRL-robot-navigation
☆ REALMS2 -- Resilient Exploration And Lunar Mapping System 2 -- A Comprehensive Approach IROS 2025
The European Space Agency (ESA) and the European Space Resources Innovation Centre (ESRIC) created the Space Resources Challenge to invite researchers and companies to propose innovative solutions for Multi-Robot Systems (MRS) space prospection. This paper proposes the Resilient Exploration And Lunar Mapping System 2 (REALMS2), a MRS framework for planetary prospection and mapping. Based on Robot Operating System version 2 (ROS 2) and enhanced with Visual Simultaneous Localisation And Mapping (vSLAM) for map generation, REALMS2 uses a mesh network for a robust ad hoc network. A single graphical user interface (GUI) controls all the rovers, providing a simple overview of the robotic mission. This system is designed for heterogeneous multi-robot exploratory missions, tackling the challenges presented by extraterrestrial environments. REALMS2 was used during the second field test of the ESA-ESRIC Challenge and allowed to map around 60% of the area, using three homogeneous rovers while handling communication delays and blackouts.
comment: 8 Pages, 8 Figures, Submitted and Accepted to IROS 2025
☆ A Sliding-Window Filter for Online Continuous-Time Continuum Robot State Estimation
Stochastic state estimation methods for continuum robots (CRs) often struggle to balance accuracy and computational efficiency. While several recent works have explored sliding-window formulations for CRs, these methods are limited to simplified, discrete-time approximations and do not provide stochastic representations. In contrast, current stochastic filter methods must run at the speed of measurements, limiting their full potential. Recent works in continuous-time estimation techniques for CRs show a principled approach to addressing this runtime constraint, but are currently restricted to offline operation. In this work, we present a sliding-window filter (SWF) for continuous-time state estimation of CRs that improves upon the accuracy of a filter approach while enabling continuous-time methods to operate online, all while running at faster-than-real-time speeds. This represents the first stochastic SWF specifically designed for CRs, providing a promising direction for future research in this area.
comment: 8 pages, 6 figures. Submitted to IEEE-RAS International Conference on Soft Robotics 2026
☆ Spiking Patches: Asynchronous, Sparse, and Efficient Tokens for Event Cameras
We propose tokenization of events and present a tokenizer, Spiking Patches, specifically designed for event cameras. Given a stream of asynchronous and spatially sparse events, our goal is to discover an event representation that preserves these properties. Prior works have represented events as frames or as voxels. However, while these representations yield high accuracy, both frames and voxels are synchronous and decrease the spatial sparsity. Spiking Patches gives the means to preserve the unique properties of event cameras and we show in our experiments that this comes without sacrificing accuracy. We evaluate our tokenizer using a GNN, PCN, and a Transformer on gesture recognition and object detection. Tokens from Spiking Patches yield inference times that are up to 3.4x faster than voxel-based tokens and up to 10.4x faster than frames. We achieve this while matching their accuracy and even surpassing in some cases with absolute improvements up to 3.8 for gesture recognition and up to 1.4 for object detection. Thus, tokenization constitutes a novel direction in event-based vision and marks a step towards methods that preserve the properties of event cameras.
☆ FLYINGTRUST: A Benchmark for Quadrotor Navigation Across Scenarios and Vehicles
Visual navigation algorithms for quadrotors often exhibit a large variation in performance when transferred across different vehicle platforms and scene geometries, which increases the cost and risk of field deployment. To support systematic early-stage evaluation, we introduce FLYINGTRUST, a high-fidelity, configurable benchmarking framework that measures how platform kinodynamics and scenario structure jointly affect navigation robustness. FLYINGTRUST models vehicle capability with two compact, physically interpretable indicators: maximum thrust-to-weight ratio and axis-wise maximum angular acceleration. The benchmark pairs a diverse scenario library with a heterogeneous set of real and virtual platforms and prescribes a standardized evaluation protocol together with a composite scoring method that balances scenario importance, platform importance and performance stability. We use FLYINGTRUST to compare representative optimization-based and learning-based navigation approaches under identical conditions, performing repeated trials per platform-scenario combination and reporting uncertainty-aware metrics. The results reveal systematic patterns: navigation success depends predictably on platform capability and scene geometry, and different algorithms exhibit distinct preferences and failure modes across the evaluated conditions. These observations highlight the practical necessity of incorporating both platform capability and scenario structure into algorithm design, evaluation, and selection, and they motivate future work on methods that remain robust across diverse platforms and scenarios.
☆ Proxemics and Permeability of the Pedestrian Group
People tend to walk in groups, and interactions with those groups have a significant impact on crowd behavior and pedestrian traffic dynamics. Social norms can be seen as unwritten rules regulating people interactions in social settings. This article studies people interactions with groups and the emergence of group proxemics. Group zones, zone occupancy counts and people clearance from the group are studied using naturalistic data. Analysis indicate potential presence of three different zones in addition to the public zone. People tend to remain in the public zone and only progressively get closer to groups, and those closer approaches happen in a low frequency and for brief periods of time.
☆ Adaptive Inverse Kinematics Framework for Learning Variable-Length Tool Manipulation in Robotics
Conventional robots possess a limited understanding of their kinematics and are confined to preprogrammed tasks, hindering their ability to leverage tools efficiently. Driven by the essential components of tool usage - grasping the desired outcome, selecting the most suitable tool, determining optimal tool orientation, and executing precise manipulations - we introduce a pioneering framework. Our novel approach expands the capabilities of the robot's inverse kinematics solver, empowering it to acquire a sequential repertoire of actions using tools of varying lengths. By integrating a simulation-learned action trajectory with the tool, we showcase the practicality of transferring acquired skills from simulation to real-world scenarios through comprehensive experimentation. Remarkably, our extended inverse kinematics solver demonstrates an impressive error rate of less than 1 cm. Furthermore, our trained policy achieves a mean error of 8 cm in simulation. Noteworthy, our model achieves virtually indistinguishable performance when employing two distinct tools of different lengths. This research provides an indication of potential advances in the exploration of all four fundamental aspects of tool usage, enabling robots to master the intricate art of tool manipulation across diverse tasks.
comment: 10 pages, 5 figures. Demonstrates a reinforcement learning framework for adaptive tool manipulation with variable-length extensions
☆ RoboOS-NeXT: A Unified Memory-based Framework for Lifelong, Scalable, and Robust Multi-Robot Collaboration
The proliferation of collaborative robots across diverse tasks and embodiments presents a central challenge: achieving lifelong adaptability, scalable coordination, and robust scheduling in multi-agent systems. Existing approaches, from vision-language-action (VLA) models to hierarchical frameworks, fall short due to their reliance on limited or dividual-agent memory. This fundamentally constrains their ability to learn over long horizons, scale to heterogeneous teams, or recover from failures, highlighting the need for a unified memory representation. To address these limitations, we introduce RoboOS-NeXT, a unified memory-based framework for lifelong, scalable, and robust multi-robot collaboration. At the core of RoboOS-NeXT is the novel Spatio-Temporal-Embodiment Memory (STEM), which integrates spatial scene geometry, temporal event history, and embodiment profiles into a shared representation. This memory-centric design is integrated into a brain-cerebellum framework, where a high-level brain model performs global planning by retrieving and updating STEM, while low-level controllers execute actions locally. This closed loop between cognition, memory, and execution enables dynamic task allocation, fault-tolerant collaboration, and consistent state synchronization. We conduct extensive experiments spanning complex coordination tasks in restaurants, supermarkets, and households. Our results demonstrate that RoboOS-NeXT achieves superior performance across heterogeneous embodiments, validating its effectiveness in enabling lifelong, scalable, and robust multi-robot collaboration. Project website: https://flagopen.github.io/RoboOS/
☆ Efficient Collision-Avoidance Constraints for Ellipsoidal Obstacles in Optimal Control: Application to Path-Following MPC and UAVs
This article proposes a modular optimal control framework for local three-dimensional ellipsoidal obstacle avoidance, exemplarily applied to model predictive path-following control. Static as well as moving obstacles are considered. Central to the approach is a computationally efficient and continuously differentiable condition for detecting collisions with ellipsoidal obstacles. A novel two-stage optimization approach mitigates numerical issues arising from the structure of the resulting optimal control problem. The effectiveness of the approach is demonstrated through simulations and real-world experiments with the Crazyflie quadrotor. This represents the first hardware demonstration of an MPC controller of this kind for UAVs in a three-dimensional task.
☆ Human-in-the-loop Online Rejection Sampling for Robotic Manipulation
Reinforcement learning (RL) is widely used to produce robust robotic manipulation policies, but fine-tuning vision-language-action (VLA) models with RL can be unstable due to inaccurate value estimates and sparse supervision at intermediate steps. In contrast, imitation learning (IL) is easy to train but often underperforms due to its offline nature. In this paper, we propose Hi-ORS, a simple yet effective post-training method that utilizes rejection sampling to achieve both training stability and high robustness. Hi-ORS stabilizes value estimation by filtering out negatively rewarded samples during online fine-tuning, and adopts a reward-weighted supervised training objective to provide dense intermediate-step supervision. For systematic study, we develop an asynchronous inference-training framework that supports flexible online human-in-the-loop corrections, which serve as explicit guidance for learning error-recovery behaviors. Across three real-world tasks and two embodiments, Hi-ORS fine-tunes a pi-base policy to master contact-rich manipulation in just 1.5 hours of real-world training, outperforming RL and IL baselines by a substantial margin in both effectiveness and efficiency. Notably, the fine-tuned policy exhibits strong test-time scalability by reliably executing complex error-recovery behaviors to achieve better performance.
comment: 8 pages
☆ CorVS: Person Identification via Video Trajectory-Sensor Correspondence in a Real-World Warehouse
Worker location data is key to higher productivity in industrial sites. Cameras are a promising tool for localization in logistics warehouses since they also offer valuable environmental contexts such as package status. However, identifying individuals with only visual data is often impractical. Accordingly, several prior studies identified people in videos by comparing their trajectories and wearable sensor measurements. While this approach has advantages such as independence from appearance, the existing methods may break down under real-world conditions. To overcome this challenge, we propose CorVS, a novel data-driven person identification method based on correspondence between visual tracking trajectories and sensor measurements. Firstly, our deep learning model predicts correspondence probabilities and reliabilities for every pair of a trajectory and sensor measurements. Secondly, our algorithm matches the trajectories and sensor measurements over time using the predicted probabilities and reliabilities. We developed a dataset with actual warehouse operations and demonstrated the method's effectiveness for real-world applications.
comment: 7 pages, 3 figures, accepted to IPIN 2025
☆ Towards Reinforcement Learning Based Log Loading Automation
Forestry forwarders play a central role in mechanized timber harvesting by picking up and moving logs from the felling site to a processing area or a secondary transport vehicle. Forwarder operation is challenging and physically and mentally exhausting for the operator who must control the machine in remote areas for prolonged periods of time. Therefore, even partial automation of the process may reduce stress on the operator. This study focuses on continuing previous research efforts in application of reinforcement learning agents in automating log handling process, extending the task from grasping which was studied in previous research to full log loading operation. The resulting agent will be capable to automate a full loading procedure from locating and grappling to transporting and delivering the log to a forestry forwarder bed. To train the agent, a trailer type forestry forwarder simulation model in NVIDIA's Isaac Gym and a virtual environment for a typical log loading scenario were developed. With reinforcement learning agents and a curriculum learning approach, the trained agent may be a stepping stone towards application of reinforcement learning agents in automation of the forestry forwarder. The agent learnt grasping a log in a random position from grapple's random position and transport it to the bed with 94% success rate of the best performing agent.
☆ Cooperative Task Spaces for Multi-Arm Manipulation Control based on Similarity Transformations
Many tasks in human environments require collaborative behavior between multiple kinematic chains, either to provide additional support for carrying big and bulky objects or to enable the dexterity that is required for in-hand manipulation. Since these complex systems often have a very high number of degrees of freedom coordinating their movements is notoriously difficult to model. In this article, we present the derivation of the theoretical foundations for cooperative task spaces of multi-arm robotic systems based on geometric primitives defined using conformal geometric algebra. Based on the similarity transformations of these cooperative geometric primitives, we derive an abstraction of complex robotic systems that enables representing these systems in a way that directly corresponds to single-arm systems. By deriving the associated analytic and geometric Jacobian matrices, we then show the straightforward integration of our approach into classical control techniques rooted in operational space control. We demonstrate this using bimanual manipulators, humanoids and multi-fingered hands in optimal control experiments for reaching desired geometric primitives and in teleoperation experiments using differential kinematics control. We then discuss how the geometric primitives naturally embed nullspace structures into the controllers that can be exploited for introducing secondary control objectives. This work, represents the theoretical foundations of this cooperative manipulation control framework, and thus the experiments are presented in an abstract way, while giving pointers towards potential future applications.
☆ AgriGS-SLAM: Orchard Mapping Across Seasons via Multi-View Gaussian Splatting SLAM
Autonomous robots in orchards require real-time 3D scene understanding despite repetitive row geometry, seasonal appearance changes, and wind-driven foliage motion. We present AgriGS-SLAM, a Visual--LiDAR SLAM framework that couples direct LiDAR odometry and loop closures with multi-camera 3D Gaussian Splatting (3DGS) rendering. Batch rasterization across complementary viewpoints recovers orchard structure under occlusions, while a unified gradient-driven map lifecycle executed between keyframes preserves fine details and bounds memory. Pose refinement is guided by a probabilistic LiDAR-based depth consistency term, back-propagated through the camera projection to tighten geometry-appearance coupling. We deploy the system on a field platform in apple and pear orchards across dormancy, flowering, and harvesting, using a standardized trajectory protocol that evaluates both training-view and novel-view synthesis to reduce 3DGS overfitting in evaluation. Across seasons and sites, AgriGS-SLAM delivers sharper, more stable reconstructions and steadier trajectories than recent state-of-the-art 3DGS-SLAM baselines while maintaining real-time performance on-tractor. While demonstrated in orchard monitoring, the approach can be applied to other outdoor domains requiring robust multimodal perception.
☆ Thor: Towards Human-Level Whole-Body Reactions for Intense Contact-Rich Environments
Humanoids hold great potential for service, industrial, and rescue applications, in which robots must sustain whole-body stability while performing intense, contact-rich interactions with the environment. However, enabling humanoids to generate human-like, adaptive responses under such conditions remains a major challenge. To address this, we propose Thor, a humanoid framework for human-level whole-body reactions in contact-rich environments. Based on the robot's force analysis, we design a force-adaptive torso-tilt (FAT2) reward function to encourage humanoids to exhibit human-like responses during force-interaction tasks. To mitigate the high-dimensional challenges of humanoid control, Thor introduces a reinforcement learning architecture that decouples the upper body, waist, and lower body. Each component shares global observations of the whole body and jointly updates its parameters. Finally, we deploy Thor on the Unitree G1, and it substantially outperforms baselines in force-interaction tasks. Specifically, the robot achieves a peak pulling force of 167.7 N (approximately 48% of the G1's body weight) when moving backward and 145.5 N when moving forward, representing improvements of 68.9% and 74.7%, respectively, compared with the best-performing baseline. Moreover, Thor is capable of pulling a loaded rack (130 N) and opening a fire door with one hand (60 N). These results highlight Thor's effectiveness in enhancing humanoid force-interaction capabilities.
☆ PHUMA: Physically-Grounded Humanoid Locomotion Dataset
Motion imitation is a promising approach for humanoid locomotion, enabling agents to acquire humanlike behaviors. Existing methods typically rely on high-quality motion capture datasets such as AMASS, but these are scarce and expensive, limiting scalability and diversity. Recent studies attempt to scale data collection by converting large-scale internet videos, exemplified by Humanoid-X. However, they often introduce physical artifacts such as floating, penetration, and foot skating, which hinder stable imitation. In response, we introduce PHUMA, a Physically-grounded HUMAnoid locomotion dataset that leverages human video at scale, while addressing physical artifacts through careful data curation and physics-constrained retargeting. PHUMA enforces joint limits, ensures ground contact, and eliminates foot skating, producing motions that are both large-scale and physically reliable. We evaluated PHUMA in two sets of conditions: (i) imitation of unseen motion from self-recorded test videos and (ii) path following with pelvis-only guidance. In both cases, PHUMA-trained policies outperform Humanoid-X and AMASS, achieving significant gains in imitating diverse motions. The code is available at https://davian-robotics.github.io/PHUMA.
☆ Self-localization on a 3D map by fusing global and local features from a monocular camera
Self-localization on a 3D map by using an inexpensive monocular camera is required to realize autonomous driving. Self-localization based on a camera often uses a convolutional neural network (CNN) that can extract local features that are calculated by nearby pixels. However, when dynamic obstacles, such as people, are present, CNN does not work well. This study proposes a new method combining CNN with Vision Transformer, which excels at extracting global features that show the relationship of patches on whole image. Experimental results showed that, compared to the state-of-the-art method (SOTA), the accuracy improvement rate in a CG dataset with dynamic obstacles is 1.5 times higher than that without dynamic obstacles. Moreover, the self-localization error of our method is 20.1% smaller than that of SOTA on public datasets. Additionally, our robot using our method can localize itself with 7.51cm error on average, which is more accurate than SOTA.
☆ Adaptive Trajectory Refinement for Optimization-based Local Planning in Narrow Passages
Trajectory planning for mobile robots in cluttered environments remains a major challenge due to narrow passages, where conventional methods often fail or generate suboptimal paths. To address this issue, we propose the adaptive trajectory refinement algorithm, which consists of two main stages. First, to ensure safety at the path-segment level, a segment-wise conservative collision test is applied, where risk-prone trajectory path segments are recursively subdivided until collision risks are eliminated. Second, to guarantee pose-level safety, pose correction based on penetration direction and line search is applied, ensuring that each pose in the trajectory is collision-free and maximally clear from obstacles. Simulation results demonstrate that the proposed method achieves up to 1.69x higher success rates and up to 3.79x faster planning times than state-of-the-art approaches. Furthermore, real-world experiments confirm that the robot can safely pass through narrow passages while maintaining rapid planning performance.
☆ Kinodynamic Task and Motion Planning using VLM-guided and Interleaved Sampling
Task and Motion Planning (TAMP) integrates high-level task planning with low-level motion feasibility, but existing methods are costly in long-horizon problems due to excessive motion sampling. While LLMs provide commonsense priors, they lack 3D spatial reasoning and cannot ensure geometric or dynamic feasibility. We propose a kinodynamic TAMP framework based on a hybrid state tree that uniformly represents symbolic and numeric states during planning, enabling task and motion decisions to be jointly decided. Kinodynamic constraints embedded in the TAMP problem are verified by an off-the-shelf motion planner and physics simulator, and a VLM guides exploring a TAMP solution and backtracks the search based on visual rendering of the states. Experiments on the simulated domains and in the real world show 32.14% - 1166.67% increased average success rates compared to traditional and LLM-based TAMP planners and reduced planning time on complex problems, with ablations further highlighting the benefits of VLM guidance.
☆ Embodied Intelligence for Advanced Bioinspired Microrobotics: Examples and Insights
The term embodied intelligence (EI) conveys the notion that body morphology, material properties, interaction with the environment, and control strategies can be purposefully integrated into the process of robotic design to generate intelligent behavior; in particular, locomotion and navigation. In this paper, we discuss EI as a design principle for advanced microrobotics, with a particular focus on co-design -- the simultaneous and interdependent development of physical structure and behavioral function. To illustrate the contrast between EI-inspired systems and traditional architectures that decouple sensing, computation, and actuation, we present and discuss a collection of robots developed by the author and his team at the Autonomous Microrobotic Systems Laboratory (AMSL). These robots exhibit intelligent behavior that emerges from their structural dynamics and the physical interaction between their components and with the environment. Platforms such as the Bee++, RoBeetle, SMALLBug, SMARTI, WaterStrider, VLEIBot+, and FRISSHBot exemplify how feedback loops, decision logics, sensing mechanisms, and smart actuation strategies can be embedded into the physical properties of the robotic system itself. Along these lines, we contend that co-design is not only a method for empirical optimization under constraints, but also an enabler of EI, offering a scalable and robust alternative to classical control for robotics at the mm-to-cm-scale.
comment: 8 pages, 7 figures, accepted to ICAR 2025
☆ Exploring Object-Aware Attention Guided Frame Association for RGB-D SLAM
Attention models have recently emerged as a powerful approach, demonstrating significant progress in various fields. Visualization techniques, such as class activation mapping, provide visual insights into the reasoning of convolutional neural networks (CNNs). Using network gradients, it is possible to identify regions where the network pays attention during image recognition tasks. Furthermore, these gradients can be combined with CNN features to localize more generalizable, task-specific attentive (salient) regions within scenes. However, explicit use of this gradient-based attention information integrated directly into CNN representations for semantic object understanding remains limited. Such integration is particularly beneficial for visual tasks like simultaneous localization and mapping (SLAM), where CNN representations enriched with spatially attentive object locations can enhance performance. In this work, we propose utilizing task-specific network attention for RGB-D indoor SLAM. Specifically, we integrate layer-wise attention information derived from network gradients with CNN feature representations to improve frame association performance. Experimental results indicate improved performance compared to baseline methods, particularly for large environments.
comment: double-column 5 pages, 3 figures
☆ Beyond the Uncanny Valley: A Mixed-Method Investigation of Anthropomorphism in Protective Responses to Robot Abuse
Robots with anthropomorphic features are increasingly shaping how humans perceive and morally engage with them. Our research investigates how different levels of anthropomorphism influence protective responses to robot abuse, extending the Computers as Social Actors (CASA) and uncanny valley theories into a moral domain. In an experiment, we invite 201 participants to view videos depicting abuse toward a robot with low (Spider), moderate (Two-Foot), or high (Humanoid) anthropomorphism. To provide a comprehensive analysis, we triangulate three modalities: self-report surveys measuring emotions and uncanniness, physiological data from automated facial expression analysis, and qualitative reflections. Findings indicate that protective responses are not linear. The moderately anthropomorphic Two-Foot robot, rated highest in eeriness and "spine-tingling" sensations consistent with the uncanny valley, elicited the strongest physiological anger expressions. Self-reported anger and guilt are significantly higher for both the Two-Foot and Humanoid robots compared to the Spider. Qualitative findings further reveal that as anthropomorphism increases, moral reasoning shifts from technical assessments of property damage to condemnation of the abuser's character, while governance proposals expand from property law to calls for quasi-animal rights and broader societal responsibility. These results suggest that the uncanny valley does not dampen moral concern but paradoxically heightens protective impulses, offering critical implications for robot design, policy, and future legal frameworks.
☆ I don't Want You to Die: A Shared Responsibility Framework for Safeguarding Child-Robot Companionship
Social robots like Moxie are designed to form strong emotional bonds with children, but their abrupt discontinuation can cause significant struggles and distress to children. When these services end, the resulting harm raises complex questions of who bears responsibility when children's emotional bonds are broken. Using the Moxie shutdown as a case study through a qualitative survey of 72 U.S. participants, our findings show that the responsibility is viewed as a shared duty across the robot company, parents, developers, and government. However, these attributions varied by political ideology and parental status of whether they have children. Participants' perceptions of whether the robot service should continue are highly polarized; supporters propose technical, financial, and governmental pathways for continuity, while opponents cite business realities and risks of unhealthy emotional dependency. Ultimately, this research contributes an empirically grounded shared responsibility framework for safeguarding child-robot companionship by detailing how accountability is distributed and contested, informing concrete design and policy implications to mitigate the emotional harm of robot discontinuation.
☆ Morphology-Aware Graph Reinforcement Learning for Tensegrity Robot Locomotion
Tensegrity robots combine rigid rods and elastic cables, offering high resilience and deployability but posing major challenges for locomotion control due to their underactuated and highly coupled dynamics. This paper introduces a morphology-aware reinforcement learning framework that integrates a graph neural network (GNN) into the Soft Actor-Critic (SAC) algorithm. By representing the robot's physical topology as a graph, the proposed GNN-based policy captures coupling among components, enabling faster and more stable learning than conventional multilayer perceptron (MLP) policies. The method is validated on a physical 3-bar tensegrity robot across three locomotion primitives, including straight-line tracking and bidirectional turning. It shows superior sample efficiency, robustness to noise and stiffness variations, and improved trajectory accuracy. Notably, the learned policies transfer directly from simulation to hardware without fine-tuning, achieving stable real-world locomotion. These results demonstrate the advantages of incorporating structural priors into reinforcement learning for tensegrity robot control.
☆ Accelerating Real-World Overtaking in F1TENTH Racing Employing Reinforcement Learning Methods
While autonomous racing performance in Time-Trial scenarios has seen significant progress and development, autonomous wheel-to-wheel racing and overtaking are still severely limited. These limitations are particularly apparent in real-life driving scenarios where state-of-the-art algorithms struggle to safely or reliably complete overtaking manoeuvres. This is important, as reliable navigation around other vehicles is vital for safe autonomous wheel-to-wheel racing. The F1Tenth Competition provides a useful opportunity for developing wheel-to-wheel racing algorithms on a standardised physical platform. The competition format makes it possible to evaluate overtaking and wheel-to-wheel racing algorithms against the state-of-the-art. This research presents a novel racing and overtaking agent capable of learning to reliably navigate a track and overtake opponents in both simulation and reality. The agent was deployed on an F1Tenth vehicle and competed against opponents running varying competitive algorithms in the real world. The results demonstrate that the agent's training against opponents enables deliberate overtaking behaviours with an overtaking rate of 87% compared 56% for an agent trained just to race.
☆ SpikeATac: A Multimodal Tactile Finger with Taxelized Dynamic Sensing for Dexterous Manipulation
In this work, we introduce SpikeATac, a multimodal tactile finger combining a taxelized and highly sensitive dynamic response (PVDF) with a static transduction method (capacitive) for multimodal touch sensing. Named for its `spiky' response, SpikeATac's 16-taxel PVDF film sampled at 4 kHz provides fast, sensitive dynamic signals to the very onset and breaking of contact. We characterize the sensitivity of the different modalities, and show that SpikeATac provides the ability to stop quickly and delicately when grasping fragile, deformable objects. Beyond parallel grasping, we show that SpikeATac can be used in a learning-based framework to achieve new capabilities on a dexterous multifingered robot hand. We use a learning recipe that combines reinforcement learning from human feedback with tactile-based rewards to fine-tune the behavior of a policy to modulate force. Our hardware platform and learning pipeline together enable a difficult dexterous and contact-rich task that has not previously been achieved: in-hand manipulation of fragile objects. Videos are available at \href{https://roamlab.github.io/spikeatac/}{roamlab.github.io/spikeatac}.
comment: 9 pages, 8 figures, under review
☆ A Multi-Modal Neuro-Symbolic Approach for Spatial Reasoning-Based Visual Grounding in Robotics
Visual reasoning, particularly spatial reasoning, is a challenging cognitive task that requires understanding object relationships and their interactions within complex environments, especially in robotics domain. Existing vision_language models (VLMs) excel at perception tasks but struggle with fine-grained spatial reasoning due to their implicit, correlation-driven reasoning and reliance solely on images. We propose a novel neuro_symbolic framework that integrates both panoramic-image and 3D point cloud information, combining neural perception with symbolic reasoning to explicitly model spatial and logical relationships. Our framework consists of a perception module for detecting entities and extracting attributes, and a reasoning module that constructs a structured scene graph to support precise, interpretable queries. Evaluated on the JRDB-Reasoning dataset, our approach demonstrates superior performance and reliability in crowded, human_built environments while maintaining a lightweight design suitable for robotics and embodied AI applications.
☆ A Hermetic, Transparent Soft Growing Vine Robot System for Pipe Inspection
Rehabilitation of aging pipes requires accurate condition assessment and mapping far into the pipe interiors. Soft growing vine robot systems are particularly promising for navigating confined, sinuous paths such as in pipes, but are currently limited by complex subsystems and a lack of validation in real-world industrial settings. In this paper, we introduce the concept and implementation of a hermetic and transparent vine robot system for visual condition assessment and mapping within non-branching pipes. This design encloses all mechanical and electrical components within the vine robot's soft, airtight, and transparent body, protecting them from environmental interference while enabling visual sensing. Because this approach requires an enclosed mechanism for transporting sensors, we developed, modeled, and tested a passively adapting enclosed tip mount. Finally, we validated the hermetic and transparent vine robot system concept through a real-world condition assessment and mapping task in a wastewater pipe. This work advances the use of soft-growing vine robots in pipe inspection by developing and demonstrating a robust, streamlined, field-validated system suitable for continued development and deployment.
comment: 8 pages, 7 figures
☆ Cooperative Integrated Estimation-Guidance for Simultaneous Interception of Moving Targets
This paper proposes a cooperative integrated estimation-guidance framework for simultaneous interception of a non-maneuvering target using a team of unmanned autonomous vehicles, assuming only a subset of vehicles are equipped with dedicated sensors to measure the target's states. Unlike earlier approaches that focus solely on either estimation or guidance design, the proposed framework unifies both within a cooperative architecture. To circumvent the limitation posed by heterogeneity in target observability, sensorless vehicles estimate the target's state by leveraging information exchanged with neighboring agents over a directed communication topology through a prescribed-time observer. The proposed approach employs true proportional navigation guidance (TPNG), which uses an exact time-to-go formulation and is applicable across a wide spectrum of target motions. Furthermore, prescribed-time observer and controller are employed to achieve convergence to true target's state and consensus in time-to-go within set predefined times, respectively. Simulations demonstrate the effectiveness of the proposed framework under various engagement scenarios.
☆ RepV: Safety-Separable Latent Spaces for Scalable Neurosymbolic Plan Verification
As AI systems migrate to safety-critical domains, verifying that their actions comply with well-defined rules remains a challenge. Formal methods provide provable guarantees but demand hand-crafted temporal-logic specifications, offering limited expressiveness and accessibility. Deep learning approaches enable evaluation of plans against natural-language constraints, yet their opaque decision process invites misclassifications with potentially severe consequences. We introduce RepV, a neurosymbolic verifier that unifies both views by learning a latent space where safe and unsafe plans are linearly separable. Starting from a modest seed set of plans labeled by an off-the-shelf model checker, RepV trains a lightweight projector that embeds each plan, together with a language model-generated rationale, into a low-dimensional space; a frozen linear boundary then verifies compliance for unseen natural-language rules in a single forward pass. Beyond binary classification, RepV provides a probabilistic guarantee on the likelihood of correct verification based on its position in the latent space. This guarantee enables a guarantee-driven refinement of the planner, improving rule compliance without human annotations. Empirical evaluations show that RepV improves compliance prediction accuracy by up to 15% compared to baseline methods while adding fewer than 0.2M parameters. Furthermore, our refinement framework outperforms ordinary fine-tuning baselines across various planning domains. These results show that safety-separable latent spaces offer a scalable, plug-and-play primitive for reliable neurosymbolic plan verification. Code and data are available at: https://repv-project.github.io/.
comment: Code and data are available at: https://repv-project.github.io/
☆ Heterogeneous Robot Collaboration in Unstructured Environments with Grounded Generative Intelligence
Heterogeneous robot teams operating in realistic settings often must accomplish complex missions requiring collaboration and adaptation to information acquired online. Because robot teams frequently operate in unstructured environments -- uncertain, open-world settings without prior maps -- subtasks must be grounded in robot capabilities and the physical world. While heterogeneous teams have typically been designed for fixed specifications, generative intelligence opens the possibility of teams that can accomplish a wide range of missions described in natural language. However, current large language model (LLM)-enabled teaming methods typically assume well-structured and known environments, limiting deployment in unstructured environments. We present SPINE-HT, a framework that addresses these limitations by grounding the reasoning abilities of LLMs in the context of a heterogeneous robot team through a three-stage process. Given language specifications describing mission goals and team capabilities, an LLM generates grounded subtasks which are validated for feasibility. Subtasks are then assigned to robots based on capabilities such as traversability or perception and refined given feedback collected during online operation. In simulation experiments with closed-loop perception and control, our framework achieves nearly twice the success rate compared to prior LLM-enabled heterogeneous teaming approaches. In real-world experiments with a Clearpath Jackal, a Clearpath Husky, a Boston Dynamics Spot, and a high-altitude UAV, our method achieves an 87\% success rate in missions requiring reasoning about robot capabilities and refining subtasks with online feedback. More information is provided at https://zacravichandran.github.io/SPINE-HT.
☆ NaviTrace: Evaluating Embodied Navigation of Vision-Language Models
Vision-language models demonstrate unprecedented performance and generalization across a wide range of tasks and scenarios. Integrating these foundation models into robotic navigation systems opens pathways toward building general-purpose robots. Yet, evaluating these models' navigation capabilities remains constrained by costly real-world trials, overly simplified simulations, and limited benchmarks. We introduce NaviTrace, a high-quality Visual Question Answering benchmark where a model receives an instruction and embodiment type (human, legged robot, wheeled robot, bicycle) and must output a 2D navigation trace in image space. Across 1000 scenarios and more than 3000 expert traces, we systematically evaluate eight state-of-the-art VLMs using a newly introduced semantic-aware trace score. This metric combines Dynamic Time Warping distance, goal endpoint error, and embodiment-conditioned penalties derived from per-pixel semantics and correlates with human preferences. Our evaluation reveals consistent gap to human performance caused by poor spatial grounding and goal localization. NaviTrace establishes a scalable and reproducible benchmark for real-world robotic navigation. The benchmark and leaderboard can be found at https://leggedrobotics.github.io/navitrace_webpage/.
comment: 9 pages, 6 figures, under review at IEEE conference
☆ Design for One, Deploy for Many: Navigating Tree Mazes with Multiple Agents
Maze-like environments, such as cave and pipe networks, pose unique challenges for multiple robots to coordinate, including communication constraints and congestion. To address these challenges, we propose a distributed multi-agent maze traversal algorithm for environments that can be represented by acyclic graphs. It uses a leader-switching mechanism where one agent, assuming a head role, employs any single-agent maze solver while the other agents each choose an agent to follow. The head role gets transferred to neighboring agents where necessary, ensuring it follows the same path as a single agent would. The multi-agent maze traversal algorithm is evaluated in simulations with groups of up to 300 agents, various maze sizes, and multiple single-agent maze solvers. It is compared against strategies that are na\"ive, or assume either global communication or full knowledge of the environment. The algorithm outperforms the na\"ive strategy in terms of makespan and sum-of-fuel. It is superior to the global-communication strategy in terms of makespan but is inferior to it in terms of sum-of-fuel. The findings suggest it is asymptotically equivalent to the full-knowledge strategy with respect to either metric. Moreover, real-world experiments with up to 20 Pi-puck robots confirm the feasibility of the approach.
comment: 7 pages, 7 figures, to be published in MRS 2025
☆ Leveraging Foundation Models for Enhancing Robot Perception and Action
This thesis investigates how foundation models can be systematically leveraged to enhance robotic capabilities, enabling more effective localization, interaction, and manipulation in unstructured environments. The work is structured around four core lines of inquiry, each addressing a fundamental challenge in robotics while collectively contributing to a cohesive framework for semantics-aware robotic intelligence.
comment: Doctoral thesis
♻ ☆ CronusVLA: Towards Efficient and Robust Manipulation via Multi-Frame Vision-Language-Action Modeling
Recent vision-language-action (VLA) models built on pretrained vision-language models (VLMs) have demonstrated strong performance in robotic manipulation. However, these models remain constrained by the single-frame image paradigm and fail to fully leverage the temporal information offered by multi-frame histories, as directly feeding multiple frames into VLM backbones incurs substantial computational overhead and inference latency. We propose CronusVLA, a unified framework that extends single-frame VLA models to the multi-frame paradigm. CronusVLA follows a two-stage process: (1) Single-frame pretraining on large-scale embodied datasets with autoregressive prediction of action tokens, establishing an effective embodied vision-language foundation; (2) Multi-frame post-training, which adapts the prediction of the vision-language backbone from discrete tokens to learnable features, and aggregates historical information via feature chunking. CronusVLA effectively addresses the existing challenges of multi-frame modeling while enhancing performance and observational robustness. To evaluate the robustness under temporal and spatial disturbances, we introduce SimplerEnv-OR, a novel benchmark featuring 24 types of observational disturbances and 120 severity levels. Experiments across three embodiments in simulated and real-world environments demonstrate that CronusVLA achieves leading performance and superior robustness, with a 70.9% success rate on SimplerEnv, a 26.8% improvement over OpenVLA on LIBERO, and the highest robustness score on SimplerEnv-OR. These results highlight the potential of efficient multi-frame adaptation in VLA models for more powerful and robust real-world deployment.
comment: 39 pages, 24 figures
♻ ☆ Agile and Cooperative Aerial Manipulation of a Cable-Suspended Load
Quadrotors can carry slung loads to hard-to-reach locations at high speed. Since a single quadrotor has limited payload capacities, using a team of quadrotors to collaboratively manipulate a heavy object is a scalable and promising solution. However, existing control algorithms for multi-lifting systems only enable low-speed and low-acceleration operations due to the complex dynamic coupling between quadrotors and the load, limiting their use in time-critical missions such as search and rescue. In this work, we present a solution to significantly enhance the agility of cable-suspended multi-lifting systems. Unlike traditional cascaded solutions, we introduce a trajectory-based framework that solves the whole-body kinodynamic motion planning problem online, accounting for the dynamic coupling effects and constraints between the quadrotors and the load. The planned trajectory is provided to the quadrotors as a reference in a receding-horizon fashion and is tracked by an onboard controller that observes and compensates for the cable tension. Real-world experiments demonstrate that our framework can achieve at least eight times greater acceleration than state-of-the-art methods to follow agile trajectories. Our method can even perform complex maneuvers such as flying through narrow passages at high speed. Additionally, it exhibits high robustness against load uncertainties and does not require adding any sensors to the load, demonstrating strong practicality.
comment: 38 pages, 11 figures
♻ ☆ LiGen: GAN-Augmented Spectral Fingerprinting for Indoor Positioning
Accurate and robust indoor localization is critical for smart building applications, yet existing Wi-Fi-based systems are often vulnerable to environmental conditions. This work presents a novel indoor localization system, called LiGen, that leverages the spectral intensity patterns of ambient light as fingerprints, offering a more stable and infrastructure-free alternative to radio signals. To address the limited spectral data, we design a data augmentation framework based on generative adversarial networks (GANs), featuring two variants: PointGAN, which generates fingerprints conditioned on coordinates, and FreeGAN, which uses a weak localization model to label unconditioned samples. Our positioning model, leveraging a Multi-Layer Perceptron (MLP) architecture to train on synthesized data, achieves submeter-level accuracy, outperforming Wi-Fi-based baselines by over 50\%. LiGen also demonstrates strong robustness in cluttered environments. To the best of our knowledge, this is the first system to combine spectral fingerprints with GAN-based data augmentation for indoor localization.
comment: 6 pages, 10 figures
♻ ☆ C-NAV: Towards Self-Evolving Continual Object Navigation in Open World NeurIPS 2025
Embodied agents are expected to perform object navigation in dynamic, open-world environments. However, existing approaches typically rely on static trajectories and a fixed set of object categories during training, overlooking the real-world requirement for continual adaptation to evolving scenarios. To facilitate related studies, we introduce the continual object navigation benchmark, which requires agents to acquire navigation skills for new object categories while avoiding catastrophic forgetting of previously learned knowledge. To tackle this challenge, we propose C-Nav, a continual visual navigation framework that integrates two key innovations: (1) A dual-path anti-forgetting mechanism, which comprises feature distillation that aligns multi-modal inputs into a consistent representation space to ensure representation consistency, and feature replay that retains temporal features within the action decoder to ensure policy consistency. (2) An adaptive sampling strategy that selects diverse and informative experiences, thereby reducing redundancy and minimizing memory overhead. Extensive experiments across multiple model architectures demonstrate that C-Nav consistently outperforms existing approaches, achieving superior performance even compared to baselines with full trajectory retention, while significantly lowering memory requirements. The code will be publicly available at https://bigtree765.github.io/C-Nav-project.
comment: Accepted at NeurIPS 2025
♻ ☆ Loop Closure from Two Views: Revisiting PGO for Scalable Trajectory Estimation through Monocular Priors
(Visual) Simultaneous Localization and Mapping (SLAM) remains a fundamental challenge in enabling autonomous systems to navigate and understand large-scale environments. Traditional SLAM approaches struggle to balance efficiency and accuracy, particularly in large-scale settings where extensive computational resources are required for scene reconstruction and Bundle Adjustment (BA). However, this scene reconstruction, in the form of sparse pointclouds of visual landmarks, is often only used within the SLAM system because navigation and planning methods require different map representations. In this work, we therefore investigate a more scalable Visual SLAM (VSLAM) approach without reconstruction, mainly based on approaches for two-view loop closures. By restricting the map to a sparse keyframed pose graph without dense geometry representations, our `2GO' system achieves efficient optimization with competitive absolute trajectory accuracy. In particular, we find that recent advancements in image matching and monocular depth priors enable very accurate trajectory optimization without BA. We conduct extensive experiments on diverse datasets, including large-scale scenarios, and provide a detailed analysis of the trade-offs between runtime, accuracy, and map size. Our results demonstrate that this streamlined approach supports real-time performance, scales well in map size and trajectory duration, and effectively broadens the capabilities of VSLAM for long-duration deployments to large environments.
♻ ☆ Learning to Insert for Constructive Neural Vehicle Routing Solver NeurIPS 2025
Neural Combinatorial Optimisation (NCO) is a promising learning-based approach for solving Vehicle Routing Problems (VRPs) without extensive manual design. While existing constructive NCO methods typically follow an appending-based paradigm that sequentially adds unvisited nodes to partial solutions, this rigid approach often leads to suboptimal results. To overcome this limitation, we explore the idea of insertion-based paradigm and propose Learning to Construct with Insertion-based Paradigm (L2C-Insert), a novel learning-based method for constructive NCO. Unlike traditional approaches, L2C-Insert builds solutions by strategically inserting unvisited nodes at any valid position in the current partial solution, which can significantly enhance the flexibility and solution quality. The proposed framework introduces three key components: a novel model architecture for precise insertion position prediction, an efficient training scheme for model optimization, and an advanced inference technique that fully exploits the insertion paradigm's flexibility. Extensive experiments on both synthetic and real-world instances of the Travelling Salesman Problem (TSP) and Capacitated Vehicle Routing Problem (CVRP) demonstrate that L2C-Insert consistently achieves superior performance across various problem sizes.
comment: Accepted at NeurIPS 2025
♻ ☆ SAFE: Multitask Failure Detection for Vision-Language-Action Models NeurIPS 2025
While vision-language-action models (VLAs) have shown promising robotic behaviors across a diverse set of manipulation tasks, they achieve limited success rates when deployed on novel tasks out of the box. To allow these policies to safely interact with their environments, we need a failure detector that gives a timely alert such that the robot can stop, backtrack, or ask for help. However, existing failure detectors are trained and tested only on one or a few specific tasks, while generalist VLAs require the detector to generalize and detect failures also in unseen tasks and novel environments. In this paper, we introduce the multitask failure detection problem and propose SAFE, a failure detector for generalist robot policies such as VLAs. We analyze the VLA feature space and find that VLAs have sufficient high-level knowledge about task success and failure, which is generic across different tasks. Based on this insight, we design SAFE to learn from VLA internal features and predict a single scalar indicating the likelihood of task failure. SAFE is trained on both successful and failed rollouts and is evaluated on unseen tasks. SAFE is compatible with different policy architectures. We test it on OpenVLA, $\pi_0$, and $\pi_0$-FAST in both simulated and real-world environments extensively. We compare SAFE with diverse baselines and show that SAFE achieves state-of-the-art failure detection performance and the best trade-off between accuracy and detection time using conformal prediction. More qualitative results and code can be found at the project webpage: https://vla-safe.github.io/
comment: NeurIPS 2025 camera ready. Project Page: https://vla-safe.github.io/
♻ ☆ Towards Predicting Any Human Trajectory In Context NeurIPS 2025
Predicting accurate future trajectories of pedestrians is essential for autonomous systems but remains a challenging task due to the need for adaptability in different environments and domains. A common approach involves collecting scenario-specific data and performing fine-tuning via backpropagation. However, the need to fine-tune for each new scenario is often impractical for deployment on edge devices. To address this challenge, we introduce \paper, an In-Context Learning (ICL) framework for pedestrian trajectory prediction that enables adaptation without fine-tuning on the scenario-specific data at inference time without requiring weight updates. We propose a spatio-temporal similarity-based example selection (STES) method that selects relevant examples from previously observed trajectories within the same scene by identifying similar motion patterns at corresponding locations. To further refine this selection, we introduce prediction-guided example selection (PG-ES), which selects examples based on both the past trajectory and the predicted future trajectory, rather than relying solely on the past trajectory. This approach allows the model to account for long-term dynamics when selecting examples. Finally, instead of relying on small real-world datasets with limited scenario diversity, we train our model on a large-scale synthetic dataset to enhance its prediction ability by leveraging in-context examples. Extensive experiments demonstrate that TrajICL achieves remarkable adaptation across both in-domain and cross-domain scenarios, outperforming even fine-tuned approaches across multiple public benchmarks. Project Page: https://fujiry0.github.io/TrajICL-project-page/.
comment: NeurIPS 2025
♻ ☆ FSR-VLN: Fast and Slow Reasoning for Vision-Language Navigation with Hierarchical Multi-modal Scene Graph
Visual-Language Navigation (VLN) is a fundamental challenge in robotic systems, with broad applications for the deployment of embodied agents in real-world environments. Despite recent advances, existing approaches are limited in long-range spatial reasoning, often exhibiting low success rates and high inference latency, particularly in long-range navigation tasks. To address these limitations, we propose FSR-VLN, a vision-language navigation system that combines a Hierarchical Multi-modal Scene Graph (HMSG) with Fast-to-Slow Navigation Reasoning (FSR). The HMSG provides a multi-modal map representation supporting progressive retrieval, from coarse room-level localization to fine-grained goal view and object identification. Building on HMSG, FSR first performs fast matching to efficiently select candidate rooms, views, and objects, then applies VLM-driven refinement for final goal selection. We evaluated FSR-VLN across four comprehensive indoor datasets collected by humanoid robots, utilizing 87 instructions that encompass a diverse range of object categories. FSR-VLN achieves state-of-the-art (SOTA) performance in all datasets, measured by the retrieval success rate (RSR), while reducing the response time by 82% compared to VLM-based methods on tour videos by activating slow reasoning only when fast intuition fails. Furthermore, we integrate FSR-VLN with speech interaction, planning, and control modules on a Unitree-G1 humanoid robot, enabling natural language interaction and real-time navigation.
comment: 8 pages
♻ ☆ Human-assisted Robotic Policy Refinement via Action Preference Optimization NeurIPS 2025
Establishing a reliable and iteratively refined robotic system is essential for deploying real-world applications. While Vision-Language-Action (VLA) models are widely recognized as the foundation model for such robotic deployment, their reliance on offline expert demonstrations critically limits their capacity for post-deployment refinement. To mitigate this limitation, we introduce Action Preference Optimization (APO), a method designed to refine VLA models by human-assisted preference alignment gathered through interaction with environments. This method begins with a human-robot collaboration framework for reliable failure correction and interaction trajectory collection through human intervention. However, directly leveraging these interaction trajectories for preference optimization is non-trivial due to the challenges of irreversible robotic actions and token distribution mismatch. To solve this, APO proposes an adaptive reweighting algorithm with binary desirability signals derived from interaction, empowering VLA models effectively suppress failure-prone actions while enhancing corrective action adaptation. Ultimately, APO equips VLA models with the crucial capability to learn from failure, paving the way for their iterative refinement and reliable deployment in dynamic environments. The experiments conducted in simulation and real-world scenarios prove superior generalization and robustness of our human-assisted framework across a variety of manipulation tasks. We believe this work could bring insights for efficient and stable optimization of VLA models through human-robot collaboration. The code and dataset are released at https://github.com/GeWu-Lab/Action-Preference-Optimization
comment: Accepted By NeurIPS 2025
♻ ☆ 3D Equivariant Visuomotor Policy Learning via Spherical Projection
Equivariant models have recently been shown to improve the data efficiency of diffusion policy by a significant margin. However, prior work that explored this direction focused primarily on point cloud inputs generated by multiple cameras fixed in the workspace. This type of point cloud input is not compatible with the now-common setting where the primary input modality is an eye-in-hand RGB camera like a GoPro. This paper closes this gap by incorporating into the diffusion policy model a process that projects features from the 2D RGB camera image onto a sphere. This enables us to reason about symmetries in $\mathrm{SO}(3)$ without explicitly reconstructing a point cloud. We perform extensive experiments in both simulation and the real world that demonstrate that our method consistently outperforms strong baselines in terms of both performance and sample efficiency. Our work, Image-to-Sphere Policy ($\textbf{ISP}$), is the first $\mathrm{SO}(3)$-equivariant policy learning framework for robotic manipulation that works using only monocular RGB inputs.
♻ ☆ Falconry-like palm landing by a flapping-wing drone based on the human gesture interaction and distance-aware flight planning
Flapping-wing drones have attracted significant attention due to their biomimetic flight. They are considered more human-friendly due to their characteristics such as low noise and flexible wings, making them suitable for human-drone interactions. However, few studies have explored the practical interaction between humans and flapping-wing drones. On establishing a physical interaction system with flapping-wing drones, we can acquire inspirations from falconers who guide birds of prey to land on their arms. This interaction interprets the human body as a dynamic landing platform, which can be utilized in various scenarios such as crowded or spatially constrained environments. Thus, in this study, we propose a falconry-like interaction system in which a flapping-wing drone performs a palm landing motion on a human hand. To achieve a safe approach toward humans, we design a trajectory planning method that considers both physical and psychological factors of the human safety such as the drone's velocity and distance from the user. We use a commercial flapping platform with our implemented motion planning and conduct experiments to evaluate the palm landing performance and safety. The results demonstrate that our approach enables safe and smooth hand landing interactions. To the best of our knowledge, it is the first time to achieve a contact-based interaction between flapping-wing drones and humans.
comment: 8 pages, 14 figures
♻ ☆ DiffVLA++: Bridging Cognitive Reasoning and End-to-End Driving through Metric-Guided Alignment
Conventional end-to-end (E2E) driving models are effective at generating physically plausible trajectories, but often fail to generalize to long-tail scenarios due to the lack of essential world knowledge to understand and reason about surrounding environments. In contrast, Vision-Language-Action (VLA) models leverage world knowledge to handle challenging cases, but their limited 3D reasoning capability can lead to physically infeasible actions. In this work we introduce DiffVLA++, an enhanced autonomous driving framework that explicitly bridges cognitive reasoning and E2E planning through metric-guided alignment. First, we build a VLA module directly generating semantically grounded driving trajectories. Second, we design an E2E module with a dense trajectory vocabulary that ensures physical feasibility. Third, and most critically, we introduce a metric-guided trajectory scorer that guides and aligns the outputs of the VLA and E2E modules, thereby integrating their complementary strengths. The experiment on the ICCV 2025 Autonomous Grand Challenge leaderboard shows that DiffVLA++ achieves EPDMS of 49.12.
♻ ☆ Online Adaptation for Flying Quadrotors in Tight Formations
The task of flying in tight formations is challenging for teams of quadrotors because the complex aerodynamic wake interactions can destabilize individual team members as well as the team. Furthermore, these aerodynamic effects are highly nonlinear and fast-paced, making them difficult to model and predict. To overcome these challenges, we present L1 KNODE-DW MPC, an adaptive, mixed expert learning based control framework that allows individual quadrotors to accurately track trajectories while adapting to time-varying aerodynamic interactions during formation flights. We evaluate L1 KNODE-DW MPC in two different three-quadrotor formations and show that it outperforms several MPC baselines. Our results show that the proposed framework is capable of enabling the three-quadrotor team to remain vertically aligned in close proximity throughout the flight. These findings show that the L1 adaptive module compensates for unmodeled disturbances most effectively when paired with an accurate dynamics model. A video showcasing our framework and the physical experiments is available here: https://youtu.be/9QX1Q5Ut9Rs
comment: 10 pages, 4 figures
♻ ☆ Mechanical Intelligence-Aware Curriculum Reinforcement Learning for Humanoids with Parallel Actuation
Reinforcement learning (RL) has enabled advances in humanoid robot locomotion, yet most learning frameworks do not account for mechanical intelligence embedded in parallel actuation mechanisms due to limitations in simulator support for closed kinematic chains. This omission can lead to inaccurate motion modeling and suboptimal policies, particularly for robots with high actuation complexity. This paper presents general formulations and simulation methods for three types of parallel mechanisms: a differential pulley, a five-bar linkage, and a four-bar linkage, and trains a parallel-mechanism aware policy through an end-to-end curriculum RL framework for BRUCE, a kid-sized humanoid robot. Unlike prior approaches that rely on simplified serial approximations, we simulate all closed-chain constraints natively using GPU-accelerated MuJoCo (MJX), preserving the hardware's mechanical nonlinear properties during training. We benchmark our RL approach against a model predictive controller (MPC), demonstrating better surface generalization and performance in real-world zero-shot deployment. This work highlights the computational approaches and performance benefits of fully simulating parallel mechanisms in end-to-end learning pipelines for legged humanoids. Project codes with parallel mechanisms: https://github.com/alvister88/og_bruce
comment: Proceeding to the IEEE Humanoid Conference 2025
♻ ☆ Robust Offline Reinforcement Learning with Linearly Structured f-Divergence Regularization ICML 2025
The Robust Regularized Markov Decision Process (RRMDP) is proposed to learn policies robust to dynamics shifts by adding regularization to the transition dynamics in the value function. Existing methods mostly use unstructured regularization, potentially leading to conservative policies under unrealistic transitions. To address this limitation, we propose a novel framework, the $d$-rectangular linear RRMDP ($d$-RRMDP), which introduces latent structures into both transition kernels and regularization. We focus on offline reinforcement learning, where an agent learns policies from a precollected dataset in the nominal environment. We develop the Robust Regularized Pessimistic Value Iteration (R2PVI) algorithm that employs linear function approximation for robust policy learning in $d$-RRMDPs with $f$-divergence based regularization terms on transition kernels. We provide instance-dependent upper bounds on the suboptimality gap of R2PVI policies, demonstrating that these bounds are influenced by how well the dataset covers state-action spaces visited by the optimal robust policy under robustly admissible transitions. We establish information-theoretic lower bounds to verify that our algorithm is near-optimal. Finally, numerical experiments validate that R2PVI learns robust policies and exhibits superior computational efficiency compared to baseline methods.
comment: 41 pages, 3 figures, 2 tables. Published in Proceedings of the 42nd International Conference on Machine Learning (ICML 2025)
♻ ☆ Understanding the Application of Utility Theory in Robotics and Artificial Intelligence: A Survey
As a unifying concept in economics, game theory, and operations research, even in the Robotics and AI field, the utility is used to evaluate the level of individual needs, preferences, and interests. Especially for decision-making and learning in multi-agent/robot systems (MAS/MRS), a suitable utility model can guide agents in choosing reasonable strategies to achieve their current needs and learning to cooperate and organize their behaviors, optimizing the system's utility, building stable and reliable relationships, and guaranteeing each group member's sustainable development, similar to the human society. Although these systems' complex, large-scale, and long-term behaviors are strongly determined by the fundamental characteristics of the underlying relationships, there has been less discussion on the theoretical aspects of mechanisms and the fields of applications in Robotics and AI. This paper introduces a utility-orient needs paradigm to describe and evaluate inter and outer relationships among agents' interactions. Then, we survey existing literature in relevant fields to support it and propose several promising research directions along with some open problems deemed necessary for further investigations.
comment: I am not sure whether withdrawing this paper is suitable. However, right now this paper has significant changes in its topic and author. So, I do not want to lead to any confusion about this paper. In the future, it will have a new version. I hope people will not have issues and confusion about the older one
♻ ☆ Object-Centric Kinodynamic Planning for Nonprehensile Robot Rearrangement Manipulation
Nonprehensile actions such as pushing are crucial for addressing multi-object rearrangement problems. Many traditional methods generate robot-centric actions, which differ from intuitive human strategies and are typically inefficient. To this end, we adopt an object-centric planning paradigm and propose a unified framework for addressing a range of large-scale, physics-intensive nonprehensile rearrangement problems challenged by modeling inaccuracies and real-world uncertainties. By assuming each object can actively move without being driven by robot interactions, our planner first computes desired object motions, which are then realized through robot actions generated online via a closed-loop pushing strategy. Through extensive experiments and in comparison with state-of-the-art baselines in both simulation and on a physical robot, we show that our object-centric planning framework can generate more intuitive and task-effective robot actions with significantly improved efficiency. In addition, we propose a benchmarking protocol to standardize and facilitate future research in nonprehensile rearrangement.
♻ ☆ PoseDiff: A Unified Diffusion Model Bridging Robot Pose Estimation and Video-to-Action Control
We present PoseDiff, a conditional diffusion model that unifies robot state estimation and control within a single framework. At its core, PoseDiff maps raw visual observations into structured robot states-such as 3D keypoints or joint angles-from a single RGB image, eliminating the need for multi-stage pipelines or auxiliary modalities. Building upon this foundation, PoseDiff extends naturally to video-to-action inverse dynamics: by conditioning on sparse video keyframes generated by world models, it produces smooth and continuous long-horizon action sequences through an overlap-averaging strategy. This unified design enables scalable and efficient integration of perception and control. On the DREAM dataset, PoseDiff achieves state-of-the-art accuracy and real-time performance for pose estimation. On Libero-Object manipulation tasks, it substantially improves success rates over existing inverse dynamics modules, even under strict offline settings. Together, these results show that PoseDiff provides a scalable, accurate, and efficient bridge between perception, planning, and control in embodied AI. The video visualization results can be found on the project page: https://haozhuo-zhang.github.io/PoseDiff-project-page/.
comment: The experimental setup and metrics lacks rigor, affecting the fairness of the comparisons
Robotics 55
☆ STITCH 2.0: Extending Augmented Suturing with EKF Needle Estimation and Thread Management
Surgical suturing is a high-precision task that impacts patient healing and scarring. Suturing skill varies widely between surgeons, highlighting the need for robot assistance. Previous robot suturing works, such as STITCH 1.0 [1], struggle to fully close wounds due to inaccurate needle tracking and poor thread management. To address these challenges, we present STITCH 2.0, an elevated augmented dexterity pipeline with seven improvements including: improved EKF needle pose estimation, new thread untangling methods, and an automated 3D suture alignment algorithm. Experimental results over 15 trials find that STITCH 2.0 on average achieves 74.4% wound closure with 4.87 sutures per trial, representing 66% more sutures in 38% less time compared to the previous baseline. When two human interventions are allowed, STITCH 2.0 averages six sutures with 100% wound closure rate. Project website: https://stitch-2.github.io/
comment: Published in RA-L 2025
☆ GET-USE: Learning Generalized Tool Usage for Bimanual Mobile Manipulation via Simulated Embodiment Extensions
The ability to use random objects as tools in a generalizable manner is a missing piece in robots' intelligence today to boost their versatility and problem-solving capabilities. State-of-the-art robotic tool usage methods focused on procedurally generating or crowd-sourcing datasets of tools for a task to learn how to grasp and manipulate them for that task. However, these methods assume that only one object is provided and that it is possible, with the correct grasp, to perform the task; they are not capable of identifying, grasping, and using the best object for a task when many are available, especially when the optimal tool is absent. In this work, we propose GeT-USE, a two-step procedure that learns to perform real-robot generalized tool usage by learning first to extend the robot's embodiment in simulation and then transferring the learned strategies to real-robot visuomotor policies. Our key insight is that by exploring a robot's embodiment extensions (i.e., building new end-effectors) in simulation, the robot can identify the general tool geometries most beneficial for a task. This learned geometric knowledge can then be distilled to perform generalized tool usage tasks by selecting and using the best available real-world object as tool. On a real robot with 22 degrees of freedom (DOFs), GeT-USE outperforms state-of-the-art methods by 30-60% success rates across three vision-based bimanual mobile manipulation tool-usage tasks.
comment: 8 pages, 7 figures
☆ Modeling Collapse of Steered Vine Robots Under Their Own Weight
Soft, vine-inspired growing robots that move by eversion are highly mobile in confined environments, but, when faced with gaps in the environment, they may collapse under their own weight while navigating a desired path. In this work, we present a comprehensive collapse model that can predict the collapse length of steered robots in any shape using true shape information and tail tension. We validate this model by collapsing several unsteered robots without true shape information. The model accurately predicts the trends of those experiments. We then attempt to collapse a robot steered with a single actuator at different orientations. Our models accurately predict collapse when it occurs. Finally, we demonstrate how this could be used in the field by having a robot attempt a gap-crossing task with and without inflating its actuators. The robot needs its actuators inflated to cross the gap without collapsing, which our model supports. Our model has been specifically tested on straight and series pouch motor-actuated robots made of non-stretchable material, but it could be applied to other robot variations. This work enables us to model the robot's collapse behavior in any open environment and understand the parameters it needs to succeed in 3D navigation tasks.
☆ Robotic Assistant: Completing Collaborative Tasks with Dexterous Vision-Language-Action Models
We adapt a pre-trained Vision-Language-Action (VLA) model (Open-VLA) for dexterous human-robot collaboration with minimal language prompting. Our approach adds (i) FiLM conditioning to visual backbones for task-aware perception, (ii) an auxiliary intent head that predicts collaborator hand pose and target cues, and (iii) action-space post-processing that predicts compact deltas (position/rotation) and PCA-reduced finger joints before mapping to full commands. Using a multi-view, teleoperated Franka and Mimic-hand dataset augmented with MediaPipe hand poses, we demonstrate that delta actions are well-behaved and that four principal components explain ~96% of hand-joint variance. Ablations identify action post-processing as the primary performance driver; auxiliary intent helps, FiLM is mixed, and a directional motion loss is detrimental. A real-time stack (~0.3 s latency on one RTX 4090) composes "pick-up" and "pass" into a long-horizon behavior. We surface "trainer overfitting" to specific demonstrators as the key limitation.
☆ Collision avoidance and path finding in a robotic mobile fulfillment system using multi-objective meta-heuristics
Multi-Agent Path Finding (MAPF) has gained significant attention, with most research focusing on minimizing collisions and travel time. This paper also considers energy consumption in the path planning of automated guided vehicles (AGVs). It addresses two main challenges: i) resolving collisions between AGVs and ii) assigning tasks to AGVs. We propose a new collision avoidance strategy that takes both energy use and travel time into account. For task assignment, we present two multi-objective algorithms: Non-Dominated Sorting Genetic Algorithm (NSGA) and Adaptive Large Neighborhood Search (ALNS). Comparative evaluations show that these proposed methods perform better than existing approaches in both collision avoidance and task assignment.
☆ Learning to Plan & Schedule with Reinforcement-Learned Bimanual Robot Skills
Long-horizon contact-rich bimanual manipulation presents a significant challenge, requiring complex coordination involving a mixture of parallel execution and sequential collaboration between arms. In this paper, we introduce a hierarchical framework that frames this challenge as an integrated skill planning & scheduling problem, going beyond purely sequential decision-making to support simultaneous skill invocation. Our approach is built upon a library of single-arm and bimanual primitive skills, each trained using Reinforcement Learning (RL) in GPU-accelerated simulation. We then train a Transformer-based planner on a dataset of skill compositions to act as a high-level scheduler, simultaneously predicting the discrete schedule of skills as well as their continuous parameters. We demonstrate that our method achieves higher success rates on complex, contact-rich tasks than end-to-end RL approaches and produces more efficient, coordinated behaviors than traditional sequential-only planners.
☆ Don't Blind Your VLA: Aligning Visual Representations for OOD Generalization
The growing success of Vision-Language-Action (VLA) models stems from the promise that pretrained Vision-Language Models (VLMs) can endow agents with transferable world knowledge and vision-language (VL) grounding, laying a foundation for action models with broader generalization. Yet when these VLMs are adapted to the action modality, it remains unclear to what extent their original VL representations and knowledge are preserved. In this work, we conduct a systematic study of representation retention during VLA fine-tuning, showing that naive action fine-tuning leads to degradation of visual representations. To characterize and measure these effects, we probe VLA's hidden representations and analyze attention maps, further, we design a set of targeted tasks and methods that contrast VLA models with their counterpart VLMs, isolating changes in VL capabilities induced by action fine-tuning. We further evaluate a range of strategies for aligning visual representations and introduce a simple yet effective method that mitigates degradation and yields improved generalization to out-of-distribution (OOD) scenarios. Taken together, our analysis clarifies the trade-off between action fine-tuning and the degradation of VL representations and highlights practical approaches to recover inherited VL capabilities. Code is publicly available: https://blind-vla-paper.github.io
comment: 13 pages, 6 figures
☆ Incorporating Social Awareness into Control of Unknown Multi-Agent Systems: A Real-Time Spatiotemporal Tubes Approach
This paper presents a decentralized control framework that incorporates social awareness into multi-agent systems with unknown dynamics to achieve prescribed-time reach-avoid-stay tasks in dynamic environments. Each agent is assigned a social awareness index that quantifies its level of cooperation or self-interest, allowing heterogeneous social behaviors within the system. Building on the spatiotemporal tube (STT) framework, we propose a real-time STT framework that synthesizes tubes online for each agent while capturing its social interactions with others. A closed-form, approximation-free control law is derived to ensure that each agent remains within its evolving STT, thereby avoiding dynamic obstacles while also preventing inter-agent collisions in a socially aware manner, and reaching the target within a prescribed time. The proposed approach provides formal guarantees on safety and timing, and is computationally lightweight, model-free, and robust to unknown disturbances. The effectiveness and scalability of the framework are validated through simulation and hardware experiments on a 2D omnidirectional
☆ Using VLM Reasoning to Constrain Task and Motion Planning ICRA 2026
In task and motion planning, high-level task planning is done over an abstraction of the world to enable efficient search in long-horizon robotics problems. However, the feasibility of these task-level plans relies on the downward refinability of the abstraction into continuous motion. When a domain's refinability is poor, task-level plans that appear valid may ultimately fail during motion planning, requiring replanning and resulting in slower overall performance. Prior works mitigate this by encoding refinement issues as constraints to prune infeasible task plans. However, these approaches only add constraints upon refinement failure, expending significant search effort on infeasible branches. We propose VIZ-COAST, a method of leveraging the common-sense spatial reasoning of large pretrained Vision-Language Models to identify issues with downward refinement a priori, bypassing the need to fix these failures during planning. Experiments on two challenging TAMP domains show that our approach is able to extract plausible constraints from images and domain descriptions, drastically reducing planning times and, in some cases, eliminating downward refinement failures altogether, generalizing to a diverse range of instances from the broader domain.
comment: 8 pages, 7 figures, 1 table. Submitted to ICRA 2026
☆ Octopus-like Reaching Motion: A Perspective Inspired by Whipping
The stereotypical reaching motion of the octopus arm has drawn growing attention for its efficient control of a highly deformable body. Previous studies suggest that its characteristic bend propagation may share underlying principles with the dynamics of a whip. This work investigates whether whip-like passive dynamics in water can reproduce the kinematic features observed in biological reaching and their similarities and differences. Platform-based whipping tests were performed in water and air while systematically varying material stiffness and driving speed. Image-based quantification revealed that the Ecoflex Gel 2 arm driven at 150 rpm (motor speed) reproduced curvature propagation similar to that observed in octopus reaching. However, its bend-point velocity decreased monotonically rather than exhibiting the biological bell-shaped profile, confirming that the octopus reaching movement is not merely a passive whipping behavior. The absence of propagation in air further highlights the critical role of the surrounding medium in forming octopus-like reaching motion. This study provides a new perspective for understand biological reaching movement, and offers a potential platform for future hydrodynamic research.
comment: The first two listed authors contributed equally. Yiyuan Zhang is the corresponding author
☆ Combining Moving Mass Actuators and Manoeuvring Models for Underwater Vehicles: A Lagrangian Approach
In this paper, we present a Newton-Euler formulation of the equations of motion for underwater vehicles with an interntal moving mass actuator. Furthermore, the moving mass dynamics are expressed as an extension to the manoeuvring model for underwater vehicles, originally introduced by Fossen (1991). The influence of the moving mass is described in body-frame and included as states in both an additional kinematic equation and as part of the coupled rigid-body kinetics of the underwater vehicle. The Coriolis-centripetal effects are derived from Kirchhoff's equations and the hydrostatics are derived using first principals. The proposed Newton-Euler model is validated through simulation and compared with the traditional Hamiltonian internal moving mass actuator formulation.
comment: \c{opyright} 2025 Alexander Rambech, Ivar Saksvik and Vahid Hassani. Accepted by IFAC for publication under a Creative Commons License CC-BY-NC-ND
☆ SPADE: Sparsity Adaptive Depth Estimator for Zero-Shot, Real-Time, Monocular Depth Estimation in Underwater Environments
Underwater infrastructure requires frequent inspection and maintenance due to harsh marine conditions. Current reliance on human divers or remotely operated vehicles is limited by perceptual and operational challenges, especially around complex structures or in turbid water. Enhancing the spatial awareness of underwater vehicles is key to reducing piloting risks and enabling greater autonomy. To address these challenges, we present SPADE: SParsity Adaptive Depth Estimator, a monocular depth estimation pipeline that combines pre-trained relative depth estimator with sparse depth priors to produce dense, metric scale depth maps. Our two-stage approach first scales the relative depth map with the sparse depth points, then refines the final metric prediction with our proposed Cascade Conv-Deformable Transformer blocks. Our approach achieves improved accuracy and generalisation over state-of-the-art baselines and runs efficiently at over 15 FPS on embedded hardware, promising to support practical underwater inspection and intervention. This work has been submitted to IEEE Journal of Oceanic Engineering Special Issue of AUV 2026.
☆ Solving the Right Problem with Multi-Robot Formations
Formation control simplifies minimizing multi-robot cost functions by encoding a cost function as a shape the robots maintain. However, by reducing complex cost functions to formations, discrepancies arise between maintaining the shape and minimizing the original cost function. For example, a Diamond or Box formation shape is often used for protecting all members of the formation. When more information about the surrounding environment becomes available, a static shape often no longer minimizes the original protection cost. We propose a formation planner to reduce mismatch between a formation and the cost function while still leveraging efficient formation controllers. Our formation planner is a two-step optimization problem that identifies desired relative robot positions. We first solve a constrained problem to estimate non-linear and non-differentiable costs with a weighted sum of surrogate cost functions. We theoretically analyze this problem and identify situations where weights do not need to be updated. The weighted, surrogate cost function is then minimized using relative positions between robots. The desired relative positions are realized using a non-cooperative formation controller derived from Lyapunov's direct approach. We then demonstrate the efficacy of this approach for military-like costs such as protection and obstacle avoidance. In simulations, we show a formation planner can reduce a single cost by over 75%. When minimizing a variety of cost functions simultaneously, using a formation planner with adaptive weights can reduce the cost by 20-40%. Formation planning provides better performance by minimizing a surrogate cost function that closely approximates the original cost function instead of relying on a shape abstraction.
comment: Submitted to SAE WCX 2026
☆ Sim-to-Real Gentle Manipulation of Deformable and Fragile Objects with Stress-Guided Reinforcement Learning
Robotic manipulation of deformable and fragile objects presents significant challenges, as excessive stress can lead to irreversible damage to the object. While existing solutions rely on accurate object models or specialized sensors and grippers, this adds complexity and often lacks generalization. To address this problem, we present a vision-based reinforcement learning approach that incorporates a stress-penalized reward to discourage damage to the object explicitly. In addition, to bootstrap learning, we incorporate offline demonstrations as well as a designed curriculum progressing from rigid proxies to deformables. We evaluate the proposed method in both simulated and real-world scenarios, showing that the policy learned in simulation can be transferred to the real world in a zero-shot manner, performing tasks such as picking up and pushing tofu. Our results show that the learned policies exhibit a damage-aware, gentle manipulation behavior, demonstrating their effectiveness by decreasing the stress applied to fragile objects by 36.5% while achieving the task goals, compared to vanilla RL policies.
comment: Under review
☆ Integrating Legal and Logical Specifications in Perception, Prediction, and Planning for Automated Driving: A Survey of Methods
This survey provides an analysis of current methodologies integrating legal and logical specifications into the perception, prediction, and planning modules of automated driving systems. We systematically explore techniques ranging from logic-based frameworks to computational legal reasoning approaches, emphasizing their capability to ensure regulatory compliance and interpretability in dynamic and uncertain driving environments. A central finding is that significant challenges arise at the intersection of perceptual reliability, legal compliance, and decision-making justifiability. To systematically analyze these challenges, we introduce a taxonomy categorizing existing approaches by their theoretical foundations, architectural implementations, and validation strategies. We particularly focus on methods that address perceptual uncertainty and incorporate explicit legal norms, facilitating decisions that are both technically robust and legally defensible. The review covers neural-symbolic integration methods for perception, logic-driven rule representation, and norm-aware prediction strategies, all contributing toward transparent and accountable autonomous vehicle operation. We highlight critical open questions and practical trade-offs that must be addressed, offering multidisciplinary insights from engineering, logic, and law to guide future developments in legally compliant autonomous driving systems.
comment: Accepted to 2025 IEEE International Automated Vehicle Validation Conference (IAVVC)
☆ Geometric Robot Calibration Using a Calibration Plate
In this paper a new method for geometric robot calibration is introduced, which uses a calibration plate with precisely known distances between its measuring points. The relative measurement between two points on the calibration plate is used to determine predefined error parameters of the system. In comparison to conventional measurement methods, like laser tracker or motion capture systems, the calibration plate provides a more mechanically robust and cheaper alternative, which is furthermore easier to transport due to its small size. The calibration method, the plate design, the mathematical description of the error system as well as the identification of the parameters are described in detail. For identifying the error parameters, the least squares method and a constrained optimization problem are used. The functionality of this method was demonstrated in experiments that led to promising results, correlated with one of a laser tracker calibration. The modeling and identification of the error parameters is done for a gantry machine, but is not restricted to that type of robot.
comment: pp 309-317
☆ An approach for combining transparency and motion assistance of a lower body exoskeleton
In this paper, an approach for gait assistance with a lower body exoskeleton is described. Two concepts, transparency and motion assistance, are combined. The transparent mode, where the system is following the user's free motion with a minimum of perceived interaction forces, is realized by exploiting the gear backlash of the actuation units. During walking a superimposed assistance mode applies an additional torque guiding the legs to their estimated future position. The concept of adaptive oscillators is utilized to learn the quasi-periodic signals typical for locomotion. First experiments showed promising results.
comment: 8 pages
☆ Seeing Clearly and Deeply: An RGBD Imaging Approach with a Bio-inspired Monocentric Design
Achieving high-fidelity, compact RGBD imaging presents a dual challenge: conventional compact optics struggle with RGB sharpness across the entire depth-of-field, while software-only Monocular Depth Estimation (MDE) is an ill-posed problem reliant on unreliable semantic priors. While deep optics with elements like DOEs can encode depth, they introduce trade-offs in fabrication complexity and chromatic aberrations, compromising simplicity. To address this, we first introduce a novel bio-inspired all-spherical monocentric lens, around which we build the Bionic Monocentric Imaging (BMI) framework, a holistic co-design. This optical design naturally encodes depth into its depth-varying Point Spread Functions (PSFs) without requiring complex diffractive or freeform elements. We establish a rigorous physically-based forward model to generate a synthetic dataset by precisely simulating the optical degradation process. This simulation pipeline is co-designed with a dual-head, multi-scale reconstruction network that employs a shared encoder to jointly recover a high-fidelity All-in-Focus (AiF) image and a precise depth map from a single coded capture. Extensive experiments validate the state-of-the-art performance of the proposed framework. In depth estimation, the method attains an Abs Rel of 0.026 and an RMSE of 0.130, markedly outperforming leading software-only approaches and other deep optics systems. For image restoration, the system achieves an SSIM of 0.960 and a perceptual LPIPS score of 0.082, thereby confirming a superior balance between image fidelity and depth accuracy. This study illustrates that the integration of bio-inspired, fully spherical optics with a joint reconstruction algorithm constitutes an effective strategy for addressing the intrinsic challenges in high-performance compact RGBD imaging. Source code will be publicly available at https://github.com/ZongxiYu-ZJU/BMI.
comment: The source code will be publicly available at https://github.com/ZongxiYu-ZJU/BMI
☆ Development of Implicit-Explicit Control Based Amphibious Centipede-Type Robot and Evaluation of its Mobile Performance
Multi-legged mobile robots possess high mobility performance in rough terrain environments, stemming from their high postural stability, joint flexibility, and the redundancy provided by multiple legs. In prior research on navigating between different environments such as land and water, the primary strategy employed involves switching to a controller that generates an appropriate gait for the new environment upon entering it. However, designing appropriate gaits for each complex and diverse environment and accurately determining controller switching for each environment is challenging. Therefore, this research develops a centipede-type mobile robot that navigates both aquatic and terrestrial environments with a simple, unified control scheme, based on the implicit-explicit control philosophy and by ingeniously designing the robot's body structure. In this research, we developed the robot featuring flexible joints and left and right legs on each body segment and focused on the leg structure which has extensive contact with the environment. This paper evaluates the locomotion performance on land and water using the three developed leg structures, using the robot's leg slip rate and actuator energy consumption as evaluation metrics. The experimental results confirmed the existence of an appropriate leg structure capable of navigating both aquatic and terrestrial environments under identical control.
☆ SynHLMA:Synthesizing Hand Language Manipulation for Articulated Object with Discrete Human Object Interaction Representation
Generating hand grasps with language instructions is a widely studied topic that benefits from embodied AI and VR/AR applications. While transferring into hand articulatied object interaction (HAOI), the hand grasps synthesis requires not only object functionality but also long-term manipulation sequence along the object deformation. This paper proposes a novel HAOI sequence generation framework SynHLMA, to synthesize hand language manipulation for articulated objects. Given a complete point cloud of an articulated object, we utilize a discrete HAOI representation to model each hand object interaction frame. Along with the natural language embeddings, the representations are trained by an HAOI manipulation language model to align the grasping process with its language description in a shared representation space. A joint-aware loss is employed to ensure hand grasps follow the dynamic variations of articulated object joints. In this way, our SynHLMA achieves three typical hand manipulation tasks for articulated objects of HAOI generation, HAOI prediction and HAOI interpolation. We evaluate SynHLMA on our built HAOI-lang dataset and experimental results demonstrate the superior hand grasp sequence generation performance comparing with state-of-the-art. We also show a robotics grasp application that enables dexterous grasps execution from imitation learning using the manipulation sequence provided by our SynHLMA. Our codes and datasets will be made publicly available.
☆ Time-Optimal Transport of Loosely Placed Liquid Filled Cups along Prescribed Paths
Handling loosely placed objects with robotic manipulators is a difficult task from the point of view of trajectory planning and control. This becomes even more challenging when the object to be handled is a container filled with liquid. This paper addresses the task of transporting a liquid-filled cup placed on a tray along a prescribed path in shortest time. The objective is to minimize swapping, thus avoiding spillage of the fluid. To this end, the sloshing dynamics is incorporated into the dynamic model used within the optimal control problem formulation. The optimization problem is solved using a direct multiple shooting approach.
☆ One-shot Humanoid Whole-body Motion Learning
Whole-body humanoid motion represents a cornerstone challenge in robotics, integrating balance, coordination, and adaptability to enable human-like behaviors. However, existing methods typically require multiple training samples per motion category, rendering the collection of high-quality human motion datasets both labor-intensive and costly. To address this, we propose a novel approach that trains effective humanoid motion policies using only a single non-walking target motion sample alongside readily available walking motions. The core idea lies in leveraging order-preserving optimal transport to compute distances between walking and non-walking sequences, followed by interpolation along geodesics to generate new intermediate pose skeletons, which are then optimized for collision-free configurations and retargeted to the humanoid before integration into a simulated environment for policy training via reinforcement learning. Experimental evaluations on the CMU MoCap dataset demonstrate that our method consistently outperforms baselines, achieving superior performance across metrics. Code will be released upon acceptance.
comment: 10 pages, 3 figures, 5 tables
☆ Large Language Model-assisted Autonomous Vehicle Recovery from Immobilization
Despite significant advancements in recent decades, autonomous vehicles (AVs) continue to face challenges in navigating certain traffic scenarios where human drivers excel. In such situations, AVs often become immobilized, disrupting overall traffic flow. Current recovery solutions, such as remote intervention (which is costly and inefficient) and manual takeover (which excludes non-drivers and limits AV accessibility), are inadequate. This paper introduces StuckSolver, a novel Large Language Model (LLM) driven recovery framework that enables AVs to resolve immobilization scenarios through self-reasoning and/or passenger-guided decision-making. StuckSolver is designed as a plug-in add-on module that operates on top of the AV's existing perception-planning-control stack, requiring no modification to its internal architecture. Instead, it interfaces with standard sensor data streams to detect immobilization states, interpret environmental context, and generate high-level recovery commands that can be executed by the AV's native planner. We evaluate StuckSolver on the Bench2Drive benchmark and in custom-designed uncertainty scenarios. Results show that StuckSolver achieves near-state-of-the-art performance through autonomous self-reasoning alone and exhibits further improvements when passenger guidance is incorporated.
comment: 8 pages
☆ RADRON: Cooperative Localization of Ionizing Radiation Sources by MAVs with Compton Cameras
We present a novel approach to localizing radioactive material by cooperating Micro Aerial Vehicles (MAVs). Our approach utilizes a state-of-the-art single-detector Compton camera as a highly sensitive, yet miniature detector of ionizing radiation. The detector's exceptionally low weight (40 g) opens up new possibilities of radiation detection by a team of cooperating agile MAVs. We propose a new fundamental concept of fusing the Compton camera measurements to estimate the position of the radiation source in real time even from extremely sparse measurements. The data readout and processing are performed directly onboard and the results are used in a dynamic feedback to drive the motion of the vehicles. The MAVs are stabilized in a tightly cooperating swarm to maximize the information gained by the Compton cameras, rapidly locate the radiation source, and even track a moving radiation source.
comment: 8 pages, 9 figures, submitted for review to IEEE RA-L
☆ DARTS: A Drone-Based AI-Powered Real-Time Traffic Incident Detection System
Rapid and reliable incident detection is critical for reducing crash-related fatalities, injuries, and congestion. However, conventional methods, such as closed-circuit television, dashcam footage, and sensor-based detection, separate detection from verification, suffer from limited flexibility, and require dense infrastructure or high penetration rates, restricting adaptability and scalability to shifting incident hotspots. To overcome these challenges, we developed DARTS, a drone-based, AI-powered real-time traffic incident detection system. DARTS integrates drones' high mobility and aerial perspective for adaptive surveillance, thermal imaging for better low-visibility performance and privacy protection, and a lightweight deep learning framework for real-time vehicle trajectory extraction and incident detection. The system achieved 99% detection accuracy on a self-collected dataset and supports simultaneous online visual verification, severity assessment, and incident-induced congestion propagation monitoring via a web-based interface. In a field test on Interstate 75 in Florida, DARTS detected and verified a rear-end collision 12 minutes earlier than the local transportation management center and monitored incident-induced congestion propagation, suggesting potential to support faster emergency response and enable proactive traffic control to reduce congestion and secondary crash risk. Crucially, DARTS's flexible deployment architecture reduces dependence on frequent physical patrols, indicating potential scalability and cost-effectiveness for use in remote areas and resource-constrained settings. This study presents a promising step toward a more flexible and integrated real-time traffic incident detection system, with significant implications for the operational efficiency and responsiveness of modern transportation management.
comment: Preprint version. This manuscript is currently under review at Transportation Research Part C: Emerging Technologies. The PDF corresponds to the version submitted in June 2025. The main findings of this work were recognized with the Best Intelligent Transportation Systems Paper Award at the 2025 TRB Annual Meeting
☆ A New Type of Axis-Angle Attitude Control Law for Rotational Systems: Synthesis, Analysis, and Experiments
Over the past few decades, continuous quaternion-based attitude control has been proven highly effective for driving rotational systems that can be modeled as rigid bodies, such as satellites and drones. However, methods rooted in this approach do not enforce the existence of a unique closed-loop (CL) equilibrium attitude-error quaternion (AEQ); and, for rotational errors about the attitude-error Euler axis larger than {\pi}rad, their proportional-control effect diminishes as the system state moves away from the stable equilibrium of the CL rotational dynamics. In this paper, we introduce a new type of attitude control law that more effectively leverages the attitude-error Euler axis-angle information to guarantee a unique CL equilibrium AEQ and to provide greater flexibility in the use of proportional-control efforts. Furthermore, using two different control laws as examples-through the construction of a strict Lyapunov function for the CL dynamics-we demonstrate that the resulting unique equilibrium of the CL rotational system can be enforced to be uniformly asymptotically stable. To assess and demonstrate the functionality and performance of the proposed approach, we performed numerical simulations and executed dozens of real-time tumble-recovery maneuvers using a small quadrotor. These simulations and flight tests compellingly demonstrate that the proposed axis-angle-based method achieves superior flight performance-compared with that obtained using a high-performance quaternion-based controller-in terms of stabilization time.
comment: 2025 International Conference on Advanced Robotics (ICAR)
☆ Curvature-Aware Calibration of Tactile Sensors for Accurate Force Estimation on Non-Planar Surfaces
Flexible tactile sensors are increasingly used in real-world applications such as robotic grippers, prosthetic hands, wearable gloves, and assistive devices, where they need to conform to curved and irregular surfaces. However, most existing tactile sensors are calibrated only on flat substrates, and their accuracy and consistency degrade once mounted on curved geometries. This limitation restricts their reliability in practical use. To address this challenge, we develop a calibration model for a widely used resistive tactile sensor design that enables accurate force estimation on one-dimensional curved surfaces. We then train a neural network (a multilayer perceptron) to predict local curvature from baseline sensor outputs recorded under no applied load, achieving an R2 score of 0.91. The proposed approach is validated on five daily objects with varying curvatures under forces from 2 N to 8 N. Results show that the curvature-aware calibration maintains consistent force accuracy across all surfaces, while flat-surface calibration underestimates force as curvature increases. Our results demonstrate that curvature-aware modeling improves the accuracy, consistency, and reliability of flexible tactile sensors, enabling dependable performance across real-world applications.
comment: This work has been submitted to the IEEE for possible publication
☆ WaveVerif: Acoustic Side-Channel based Verification of Robotic Workflows
In this paper, we present a framework that uses acoustic side-channel analysis (ASCA) to monitor and verify whether a robot correctly executes its intended commands. We develop and evaluate a machine-learning-based workflow verification system that uses acoustic emissions generated by robotic movements. The system can determine whether real-time behavior is consistent with expected commands. The evaluation takes into account movement speed, direction, and microphone distance. The results show that individual robot movements can be validated with over 80% accuracy under baseline conditions using four different classifiers: Support Vector Machine (SVM), Deep Neural Network (DNN), Recurrent Neural Network (RNN), and Convolutional Neural Network (CNN). Additionally, workflows such as pick-and-place and packing could be identified with similarly high confidence. Our findings demonstrate that acoustic signals can support real-time, low-cost, passive verification in sensitive robotic environments without requiring hardware modifications.
comment: 11 pages, 3 figures, Corresponding Author: Prof. Shishir Nagaraja (shishir.nagaraja@newcastle.ac.uk)
☆ Risk-Aware Safety Filters with Poisson Safety Functions and Laplace Guidance Fields
Robotic systems navigating in real-world settings require a semantic understanding of their environment to properly determine safe actions. This work aims to develop the mathematical underpinnings of such a representation -- specifically, the goal is to develop safety filters that are risk-aware. To this end, we take a two step approach: encoding an understanding of the environment via Poisson's equation, and associated risk via Laplace guidance fields. That is, we first solve a Dirichlet problem for Poisson's equation to generate a safety function that encodes system safety as its 0-superlevel set. We then separately solve a Dirichlet problem for Laplace's equation to synthesize a safe \textit{guidance field} that encodes variable levels of caution around obstacles -- by enforcing a tunable flux boundary condition. The safety function and guidance fields are then combined to define a safety constraint and used to synthesize a risk-aware safety filter which, given a semantic understanding of an environment with associated risk levels of environmental features, guarantees safety while prioritizing avoidance of higher risk obstacles. We demonstrate this method in simulation and discuss how \textit{a priori} understandings of obstacle risk can be directly incorporated into the safety filter to generate safe behaviors that are risk-aware.
☆ BikeScenes: Online LiDAR Semantic Segmentation for Bicycles
The vulnerability of cyclists, exacerbated by the rising popularity of faster e-bikes, motivates adapting automotive perception technologies for bicycle safety. We use our multi-sensor 'SenseBike' research platform to develop and evaluate a 3D LiDAR segmentation approach tailored to bicycles. To bridge the automotive-to-bicycle domain gap, we introduce the novel BikeScenes-lidarseg Dataset, comprising 3021 consecutive LiDAR scans around the university campus of the TU Delft, semantically annotated for 29 dynamic and static classes. By evaluating model performance, we demonstrate that fine-tuning on our BikeScenes dataset achieves a mean Intersection-over-Union (mIoU) of 63.6%, significantly outperforming the 13.8% obtained with SemanticKITTI pre-training alone. This result underscores the necessity and effectiveness of domain-specific training. We highlight key challenges specific to bicycle-mounted, hardware-constrained perception systems and contribute the BikeScenes dataset as a resource for advancing research in cyclist-centric LiDAR segmentation.
☆ Debate2Create: Robot Co-design via Large Language Model Debates
Automating the co-design of a robot's morphology and control is a long-standing challenge due to the vast design space and the tight coupling between body and behavior. We introduce Debate2Create (D2C), a framework in which large language model (LLM) agents engage in a structured dialectical debate to jointly optimize a robot's design and its reward function. In each round, a design agent proposes targeted morphological modifications, and a control agent devises a reward function tailored to exploit the new design. A panel of pluralistic judges then evaluates the design-control pair in simulation and provides feedback that guides the next round of debate. Through iterative debates, the agents progressively refine their proposals, producing increasingly effective robot designs. Notably, D2C yields diverse and specialized morphologies despite no explicit diversity objective. On a quadruped locomotion benchmark, D2C discovers designs that travel 73% farther than the default, demonstrating that structured LLM-based debate can serve as a powerful mechanism for emergent robot co-design. Our results suggest that multi-agent debate, when coupled with physics-grounded feedback, is a promising new paradigm for automated robot design.
☆ Hybrid Vision Servoing with Depp Alignment and GRU-Based Occlusion Recovery
Vision-based control systems, such as image-based visual servoing (IBVS), have been extensively explored for precise robot manipulation. A persistent challenge, however, is maintaining robust target tracking under partial or full occlusions. Classical methods like Lucas-Kanade (LK) offer lightweight tracking but are fragile to occlusion and drift, while deep learning-based approaches often require continuous visibility and intensive computation. To address these gaps, we propose a hybrid visual tracking framework that bridges advanced perception with real-time servo control. First, a fast global template matcher constrains the pose search region; next, a deep-feature Lucas-Kanade module operating on early VGG layers refines alignment to sub-pixel accuracy (<2px); then, a lightweight residual regressor corrects local misalignments caused by texture degradation or partial occlusion. When visual confidence falls below a threshold, a GRU-based predictor seamlessly extrapolates pose updates from recent motion history. Crucially, the pipeline's final outputs-translation, rotation, and scale deltas-are packaged as direct control signals for 30Hz image-based servo loops. Evaluated on handheld video sequences with up to 90% occlusion, our system sustains under 2px tracking error, demonstrating the robustness and low-latency precision essential for reliable real-world robot vision applications.
☆ RoadSens-4M: A Multimodal Smartphone & Camera Dataset for Holistic Road-way Analysis
It's important to monitor road issues such as bumps and potholes to enhance safety and improve road conditions. Smartphones are equipped with various built-in sensors that offer a cost-effective and straightforward way to assess road quality. However, progress in this area has been slow due to the lack of high-quality, standardized datasets. This paper discusses a new dataset created by a mobile app that collects sensor data from devices like GPS, accelerometers, gyroscopes, magnetometers, gravity sensors, and orientation sensors. This dataset is one of the few that integrates Geographic Information System (GIS) data with weather information and video footage of road conditions, providing a comprehensive understanding of road issues with geographic context. The dataset allows for a clearer analysis of road conditions by compiling essential data, including vehicle speed, acceleration, rotation rates, and magnetic field intensity, along with the visual and spatial context provided by GIS, weather, and video data. Its goal is to provide funding for initiatives that enhance traffic management, infrastructure development, road safety, and urban planning. Additionally, the dataset will be publicly accessible to promote further research and innovation in smart transportation systems.
☆ SoraNav: Adaptive UAV Task-Centric Navigation via Zeroshot VLM Reasoning
Interpreting visual observations and natural language instructions for complex task execution remains a key challenge in robotics and AI. Despite recent advances, language-driven navigation is still difficult, particularly for UAVs in small-scale 3D environments. Existing Vision-Language Navigation (VLN) approaches are mostly designed for ground robots and struggle to generalize to aerial tasks that require full 3D spatial reasoning. The emergence of large Vision-Language Models (VLMs), such as GPT and Claude, enables zero-shot semantic reasoning from visual and textual inputs. However, these models lack spatial grounding and are not directly applicable to navigation. To address these limitations, SoraNav is introduced, an adaptive UAV navigation framework that integrates zero-shot VLM reasoning with geometry-aware decision-making. Geometric priors are incorporated into image annotations to constrain the VLM action space and improve decision quality. A hybrid switching strategy leverages navigation history to alternate between VLM reasoning and geometry-based exploration, mitigating dead-ends and redundant revisits. A PX4-based hardware-software platform, comprising both a digital twin and a physical micro-UAV, enables reproducible evaluation. Experimental results show that in 2.5D scenarios, our method improves Success Rate (SR) by 25.7% and Success weighted by Path Length (SPL) by 17%. In 3D scenarios, it improves SR by 29.5% and SPL by 18.5% relative to the baseline.
☆ Learning Spatial-Aware Manipulation Ordering NeurIPS 2025
Manipulation in cluttered environments is challenging due to spatial dependencies among objects, where an improper manipulation order can cause collisions or blocked access. Existing approaches often overlook these spatial relationships, limiting their flexibility and scalability. To address these limitations, we propose OrderMind, a unified spatial-aware manipulation ordering framework that directly learns object manipulation priorities based on spatial context. Our architecture integrates a spatial context encoder with a temporal priority structuring module. We construct a spatial graph using k-Nearest Neighbors to aggregate geometric information from the local layout and encode both object-object and object-manipulator interactions to support accurate manipulation ordering in real-time. To generate physically and semantically plausible supervision signals, we introduce a spatial prior labeling method that guides a vision-language model to produce reasonable manipulation orders for distillation. We evaluate OrderMind on our Manipulation Ordering Benchmark, comprising 163,222 samples of varying difficulty. Extensive experiments in both simulation and real-world environments demonstrate that our method significantly outperforms prior approaches in effectiveness and efficiency, enabling robust manipulation in cluttered scenes.
comment: Accepted to NeurIPS 2025
☆ NanoVLA: Routing Decoupled Vision-Language Understanding for Nano-sized Generalist Robotic Policies
Vision-language-action (VLA) models have significantly advanced robotic manipulation by integrating vision-language models (VLMs), and action decoders into a unified architecture. However, their deployment on resource-constrained edge devices, such as mobile robots or embedded systems (e.g., Jetson Orin Nano), remains challenging due to high computational demands, especially in real-world scenarios where power, latency, and computational resources are critical. To close this gap, we introduce Nano-scale Vision-Language Action (NanoVLA), a family of lightweight VLA architectures that achieve high performance with minimal resources. Our core innovations include: (1) vision-language decoupling that moves conventional early vision and language inputs fusion in VLM to late stage, achieving better performance while enabling caching and reduce inference overhead and latency; (2) long-short action chunking to ensure smooth, coherent multi-step planning without sacrificing real-time responsiveness; (3) dynamic routing that adaptively assigns lightweight or heavy backbones based on task complexity, further optimizing inference efficiency. Experimental results on several benchmarks, as well as real-world deployments, demonstrate that NanoVLA achieves up to 52x faster inference on edge devices compared to previous state-of-the-art VLA models, with 98% less parameters while maintaining or surpassing their task accuracy and generalization. Ablation studies confirm that our decoupling strategy preserves cross-task transferability, and the routing module enhances cost-performance trade-offs, enabling practical, high-precision robotic manipulation on resource-constrained hardware.
☆ Mean-Shift Theory and Its Applications in Swarm Robotics: A New Way to Enhance the Efficiency of Multi-Robot Collaboration
Swarms evolving from collective behaviors among multiple individuals are commonly seen in nature, which enables biological systems to exhibit more efficient and robust collaboration. Creating similar swarm intelligence in engineered robots poses challenges to the design of collaborative algorithms that can be programmed at large scales. The assignment-based method has played an eminent role for a very long time in solving collaboration problems of robot swarms. However, it faces fundamental limitations in terms of efficiency and robustness due to its unscalability to swarm variants. This article presents a tutorial review on recent advances in assignment-free collaboration of robot swarms, focusing on the problem of shape formation. A key theoretical component is the recently developed \emph{mean-shift exploration} strategy, which improves the collaboration efficiency of large-scale swarms by dozens of times. Further, the efficiency improvement is more significant as the swarm scale increases. Finally, this article discusses three important applications of the mean-shift exploration strategy, including precise shape formation, area coverage formation, and maneuvering formation, as well as their corresponding industrial scenarios in smart warehousing, area exploration, and cargo transportation.
☆ Enhancing Underwater Object Detection through Spatio-Temporal Analysis and Spatial Attention Networks
This study examines the effectiveness of spatio-temporal modeling and the integration of spatial attention mechanisms in deep learning models for underwater object detection. Specifically, in the first phase, the performance of temporal-enhanced YOLOv5 variant T-YOLOv5 is evaluated, in comparison with the standard YOLOv5. For the second phase, an augmented version of T-YOLOv5 is developed, through the addition of a Convolutional Block Attention Module (CBAM). By examining the effectiveness of the already pre-existing YOLOv5 and T-YOLOv5 models and of the newly developed T-YOLOv5 with CBAM. With CBAM, the research highlights how temporal modeling improves detection accuracy in dynamic marine environments, particularly under conditions of sudden movements, partial occlusions, and gradual motion. The testing results showed that YOLOv5 achieved a mAP@50-95 of 0.563, while T-YOLOv5 and T-YOLOv5 with CBAM outperformed with mAP@50-95 scores of 0.813 and 0.811, respectively, highlighting their superior accuracy and generalization in detecting complex objects. The findings demonstrate that T-YOLOv5 significantly enhances detection reliability compared to the standard model, while T-YOLOv5 with CBAM further improves performance in challenging scenarios, although there is a loss of accuracy when it comes to simpler scenarios.
☆ Non-Invasive Calibration Of A Stewart Platform By Photogrammetry
Accurate calibration of a Stewart platform is important for their precise and efficient operation. However, the calibration of these platforms using forward kinematics is a challenge for researchers because forward kinematics normally generates multiple feasible and unfeasible solutions for any pose of the moving platform. The complex kinematic relations among the six actuator paths connecting the fixed base to the moving platform further compound the difficulty in establishing a straightforward and efficient calibration method. The authors developed a new forward kinematics-based calibration method using Denavit-Hartenberg convention and used the Stewart platform Tiger 66.1 developed in their lab for experimenting with the photogrammetry-based calibration strategies described in this paper. This system became operational upon completion of construction, marking its inaugural use. The authors used their calibration model for estimating the errors in the system and adopted three compensation options or strategies as per Least Square method to improve the accuracy of the system. These strategies leveraged a high-resolution digital camera and off-the-shelf software to capture the poses of the moving platform's center. This process is non-invasive and does not need any additional equipment to be attached to the hexapod or any alteration of the hexapod hardware. This photogrammetry-based calibration process involves multiple high-resolution images from different angles to measure the position and orientation of the platform center in the three-dimensional space. The Target poses and Actual poses are then compared, and the error compensations are estimated using the Least-Squared methods to calculate the Predicted poses. Results from each of the three compensation approaches demonstrated noticeable enhancements in platform pose accuracies, suggesting room for further improvements.
comment: The International Journal of Advanced Manufacturing Technology, 2024
☆ Scalable predictive processing framework for multitask caregiving robots
The rapid aging of societies is intensifying demand for autonomous care robots; however, most existing systems are task-specific and rely on handcrafted preprocessing, limiting their ability to generalize across diverse scenarios. A prevailing theory in cognitive neuroscience proposes that the human brain operates through hierarchical predictive processing, which underlies flexible cognition and behavior by integrating multimodal sensory signals. Inspired by this principle, we introduce a hierarchical multimodal recurrent neural network grounded in predictive processing under the free-energy principle, capable of directly integrating over 30,000-dimensional visuo-proprioceptive inputs without dimensionality reduction. The model was able to learn two representative caregiving tasks, rigid-body repositioning and flexible-towel wiping, without task-specific feature engineering. We demonstrate three key properties: (i) self-organization of hierarchical latent dynamics that regulate task transitions, capture variability in uncertainty, and infer occluded states; (ii) robustness to degraded vision through visuo-proprioceptive integration; and (iii) asymmetric interference in multitask learning, where the more variable wiping task had little influence on repositioning, whereas learning the repositioning task led to a modest reduction in wiping performance, while the model maintained overall robustness. Although the evaluation was limited to simulation, these results establish predictive processing as a universal and scalable computational principle, pointing toward robust, flexible, and autonomous caregiving robots while offering theoretical insight into the human brain's ability to achieve flexible adaptation in uncertain real-world environments.
☆ Force Characterization of Insect-Scale Aquatic Propulsion Based on Fluid-Structure Interaction
We present force characterizations of two newly developed insect-scale propulsors--one single-tailed and one double-tailed--for microrobotic swimmers that leverage fluid-structure interaction (FSI) to generate thrust. The designs of these two devices were inspired by anguilliform swimming and are driven by soft tails excited by high-work-density (HWD) actuators powered by shape-memory alloy (SMA) wires. While these propulsors have been demonstrated to be suitable for microrobotic aquatic locomotion and controllable with simple architectures for trajectory tracking in the two-dimensional (2D) space, the characteristics and magnitudes of the associated forces have not been studied systematically. In the research presented here, we adopted a theoretical framework based on the notion of reactive forces and obtained experimental data for characterization using a custom-built micro-N-resolution force sensor. We measured maximum and cycle-averaged force values with multi-test means of respectively 0.45 mN and 2.97 micro-N, for the tested single-tail propulsor. For the dual-tail propulsor, we measured maximum and cycle-averaged force values with multi-test means of 0.61 mN and 22.6 micro-N, respectively. These results represent the first measurements of the instantaneous thrust generated by insect-scale propulsors of this type and provide insights into FSI for efficient microrobotic propulsion.
comment: To be presented at ICAR 2025 in San Juan, Argentina
♻ ☆ Taxonomy and Trends in Reinforcement Learning for Robotics and Control Systems: A Structured Review
Reinforcement learning (RL) has become a foundational approach for enabling intelligent robotic behavior in dynamic and uncertain environments. This work presents an in-depth review of RL principles, advanced deep reinforcement learning (DRL) algorithms, and their integration into robotic and control systems. Beginning with the formalism of Markov Decision Processes (MDPs), the study outlines essential elements of the agent-environment interaction and explores core algorithmic strategies including actor-critic methods, value-based learning, and policy gradients. Emphasis is placed on modern DRL techniques such as DDPG, TD3, PPO, and SAC, which have shown promise in solving high-dimensional, continuous control tasks. A structured taxonomy is introduced to categorize RL applications across domains such as locomotion, manipulation, multi-agent coordination, and human-robot interaction, along with training methodologies and deployment readiness levels. The review synthesizes recent research efforts, highlighting technical trends, design patterns, and the growing maturity of RL in real-world robotics. Overall, this work aims to bridge theoretical advances with practical implementations, providing a consolidated perspective on the evolving role of RL in autonomous robotic systems.
♻ ☆ RoboOmni: Proactive Robot Manipulation in Omni-modal Context
Recent advances in Multimodal Large Language Models (MLLMs) have driven rapid progress in Vision-Language-Action (VLA) models for robotic manipulation. Although effective in many scenarios, current approaches largely rely on explicit instructions, whereas in real-world interactions, humans rarely issue instructions directly. Effective collaboration requires robots to infer user intentions proactively. In this work, we introduce cross-modal contextual instructions, a new setting where intent is derived from spoken dialogue, environmental sounds, and visual cues rather than explicit commands. To address this new setting, we present RoboOmni, a Perceiver-Thinker-Talker-Executor framework based on end-to-end omni-modal LLMs that unifies intention recognition, interaction confirmation, and action execution. RoboOmni fuses auditory and visual signals spatiotemporally for robust intention recognition, while supporting direct speech interaction. To address the absence of training data for proactive intention recognition in robotic manipulation, we build OmniAction, comprising 140k episodes, 5k+ speakers, 2.4k event sounds, 640 backgrounds, and six contextual instruction types. Experiments in simulation and real-world settings show that RoboOmni surpasses text- and ASR-based baselines in success rate, inference speed, intention recognition, and proactive assistance.
♻ ☆ Optimal Kinematic Synthesis and Prototype Development of Knee Exoskeleton
The range of rotation (RoR) in a knee exoskeleton is a critical factor in rehabilitation, as it directly influences joint mobility, muscle activation, and recovery outcomes. A well-designed RoR ensures that patients achieve near-natural knee kinematics, which is essential for restoring gait patterns and preventing compensatory movements. This paper presents optimal design of one degree of freedom knee exoskeleton. In kinematic analysis, the existing design being represented by nonlinear and nonconvex mathematical functions. To obtain feasible and optimum measurement of the links of knee exoskeleton, an optimization problem is formulated based on the kinematic analysis and average human's leg measurement. The optimized solution increases the range of motion of knee exoskeleton during sit to stand motion by $24 \%$ as compared with inspired design. Furthermore, misalignment study is conducted by comparing the trajectory of human's knee and exoskeleton's knee during sit to stand motion. The joint movement is calculated using marker and camera system. Finally, a prototype of the knee joint exoskeleton is being developed based on optimal dimensions which validate the maximum range of motion achieved during simulation.
♻ ☆ Dual-Regularized Riccati Recursions for Interior-Point Optimal Control
We derive closed-form extensions of Riccati's recursions (both sequential and parallel) for solving dual-regularized LQR problems. We show how these methods can be used to solve general constrained, non-convex, discrete-time optimal control problems via a regularized interior point method, while guaranteeing that each step is a descent direction of an Augmented Barrier-Lagrangian merit function. We provide MIT-licensed implementations of our methods in C++ and JAX.
♻ ☆ Classification of Driver Behaviour Using External Observation Techniques for Autonomous Vehicles
Road traffic accidents remain a significant global concern, with human error, particularly distracted and impaired driving, among the leading causes. This study introduces a novel driver behaviour classification system that uses external observation techniques to detect indicators of distraction and impairment. The proposed framework employs advanced computer vision methodologies, including real-time object tracking, lateral displacement analysis, and lane position monitoring. The system identifies unsafe driving behaviours such as excessive lateral movement and erratic trajectory patterns by implementing the YOLO object detection model and custom lane estimation algorithms. Unlike systems reliant on inter-vehicular communication, this vision-based approach enables behavioural analysis of non-connected vehicles. Experimental evaluations on diverse video datasets demonstrate the framework's reliability and adaptability across varying road and environmental conditions.
♻ ☆ SNN-Based Online Learning of Concepts and Action Laws in an Open World
We present the architecture of a fully autonomous, bio-inspired cognitive agent built around a spiking neural network (SNN) implementing the agent's semantic memory. This agent explores its universe and learns concepts of objects/situations and of its own actions in a one-shot manner. While object/situation concepts are unary, action concepts are triples made up of an initial situation, a motor activity, and an outcome. They embody the agent's knowledge of its universe's action laws. Both kinds of concepts have different degrees of generality. To make decisions the agent queries its semantic memory for the expected outcomes of envisaged actions and chooses the action to take on the basis of these predictions. Our experiments show that the agent handles new situations by appealing to previously learned general concepts and rapidly modifies its concepts to adapt to environment changes.
♻ ☆ Redistributing Rewards Across Time and Agents for Multi-Agent Reinforcement Learning
Credit assignmen, disentangling each agent's contribution to a shared reward, is a critical challenge in cooperative multi-agent reinforcement learning (MARL). To be effective, credit assignment methods must preserve the environment's optimal policy. Some recent approaches attempt this by enforcing return equivalence, where the sum of distributed rewards must equal the team reward. However, their guarantees are conditional on a learned model's regression accuracy, making them unreliable in practice. We introduce Temporal-Agent Reward Redistribution (TAR$^2$), an approach that decouples credit modeling from this constraint. A neural network learns unnormalized contribution scores, while a separate, deterministic normalization step enforces return equivalence by construction. We demonstrate that this method is equivalent to a valid Potential-Based Reward Shaping (PBRS), which guarantees the optimal policy is preserved regardless of model accuracy. Empirically, on challenging SMACLite and Google Research Football (GRF) benchmarks, TAR$^2$ accelerates learning and achieves higher final performance than strong baselines. These results establish our method as an effective solution for the agent-temporal credit assignment problem.
comment: 16 pages, 4 figures, 4 tables
♻ ☆ A Constrained Saddle Search Approach for Constructing Singular and Flexible Bar Frameworks
Singularity analysis is essential in robot kinematics, as singular configurations cause loss of control and kinematic indeterminacy. This paper models singularities in bar frameworks as saddle points on constrained manifolds. Given an under-constrained, non-singular bar framework, by allowing one edge to vary its length while fixing lengths of others, we define the squared length of the free edge as an energy functional and show that its local saddle points correspond to singular and flexible frameworks. Using our constrained saddle search approach, we identify previously unknown singular and flexible bar frameworks, providing new insights into singular robotics design and analysis.
comment: 9 pages, 3 figures
♻ ☆ STATE-NAV: Stability-Aware Traversability Estimation for Bipedal Navigation on Rough Terrain
Bipedal robots have advantages in maneuvering human-centered environments, but face greater failure risk compared to other stable mobile plarforms such as wheeled or quadrupedal robots. While learning-based traversability has been widely studied for these platforms, bipedal traversability has instead relied on manually designed rules with limited consideration of locomotion stability on rough terrain. In this work, we present the first learning-based traversability estimation and risk-sensitive navigation framework for bipedal robots operating in diverse, uneven environments. TravFormer, a transformer-based neural network, is trained to predict bipedal instability with uncertainty, enabling risk-aware and adaptive planning. Based on the network, we define traversability as stability-aware command velocity-the fastest command velocity that keeps instability below a user-defined limit. This velocity-based traversability is integrated into a hierarchical planner that combines traversability-informed Rapid Random Tree Star (TravRRT*) for time-efficient planning and Model Predictive Control (MPC) for safe execution. We validate our method in MuJoCo simulation and the real world, demonstrating improved navigation performance, with enhanced robustness and time efficiency across varying terrains compared to existing methods.
♻ ☆ SARM: Stage-Aware Reward Modeling for Long Horizon Robot Manipulation
Large-scale robot learning has recently shown promise for enabling robots to perform complex tasks by integrating perception, control, and language understanding. Yet, it struggles with long-horizon, contact-rich manipulation such as deformable object handling, where demonstration quality is inconsistent. Reward modeling offers a natural solution: by providing grounded progress signals, it transforms noisy demonstrations into stable supervision that generalizes across diverse trajectories. We introduce a stage-aware, video-based reward modeling framework that jointly predicts high-level task stages and fine-grained progress. Reward labels are automatically derived from natural language subtask annotations, ensuring consistent progress estimation across variable-length demonstrations. This design overcomes frame-index labeling, which fails in variable-duration tasks like folding a T-shirt. Our reward model demonstrates robustness to variability, generalization to out-of-distribution settings, and strong utility for policy training. Building on it, we propose Reward-Aligned Behavior Cloning (RA-BC), which filters high-quality data and reweights samples by reward. Experiments show the reward model alone outperforms baselines on validation and real robot rollouts. Integrated into RA-BC, our approach achieves 83% success on folding T-shirts from the flattened state and 67% from the crumpled state -- far surpassing vanilla behavior cloning, which attains only 8% and 0% success. Overall, our results highlight reward modeling as a key enabler for scalable, annotation-efficient, and robust imitation learning in long-horizon manipulation.
♻ ☆ Multi-robot Motion Planning based on Nets-within-Nets Modeling and Simulation
This paper focuses on designing motion plans for a heterogeneous team of robots that must cooperate to fulfill a global mission. Robots move in an environment that contains some regions of interest, while the specification for the entire team can include avoidance, visits, or sequencing of these regions of interest. The mission is expressed in terms of a Petri net corresponding to an automaton, while each robot is also modeled by a state machine Petri net. The current work brings about the following contributions with respect to existing solutions for related problems. First, we propose a novel model, denoted High-Level robot team Petri Net (HLrtPN) system, to incorporate the specification and robot models into the Nets-within-Nets paradigm. A guard function, named Global Enabling Function, is designed to synchronize the firing of transitions so that robot motions do not violate the specification. Then, the solution is found by simulating the HLrtPN system in a specific software tool that accommodates Nets-within-Nets. Illustrative examples based on Linear Temporal Logic missions support the computational feasibility of the proposed framework.
comment: [Note for readers] This paper has been extended from a previous submission to 62nd IEEE Conference on Decision and Control, Dec. 13-15, 2023. This work has been submitted to the IEEE for possible publication
♻ ☆ Control Modes of Teleoperated Surgical Robotic System's Tools in Ophthalmic Surgery
The introduction of a teleoperated surgical robotic system designed for minimally invasive procedures enables the emulation of two distinct control modes through a dedicated input device of the surgical console: (1) Inside Control Mode, which emulates tool manipulation near the distal end as if the surgeon was holding the tip of the instrument inside the patient's body; (2) Outside Control Mode, which emulates manipulation near the proximal end as if the surgeon was holding the tool externally. The aim of this research is to compare the surgeon's performance on these two modes of operation along with various scaling factors in a simulated vitreoretinal surgical setting. The console of Intraocular Robotic Interventional Surgical System (IRISS) was utilized but the surgical robot itself and the human eye anatomy was simulated by a virtual environment projected microscope view of an intraocular setup to a VR headset. Five experienced vitreoretinal surgeons and five subjects with no surgical experience used the system to perform four fundamental tool/tissue tasks common to vitreoretinal surgery: touch and reset; grasp and drop; inject; circular tracking. Results indicate that Inside Control outperforms Outside Control across multiple tasks and metrics. Higher scaling factors generally performed better, particularly for reducing trajectory errors and tissue damage. This improvement suggests that larger scaling factors enable more precise control, making them the preferred option for fine manipulation. However, completion time was not consistently reduced across all conditions, indicating that surgeons need to balance speed and accuracy based on surgical requirements. By optimizing control dynamics and user interface, robotic teleoperation has the potential to reduce complications, enhance dexterity, and expand the accessibility of high precision procedures to a broader range of practitioners.
comment: 10 pages, 11 figures
♻ ☆ RoboCerebra: A Large-scale Benchmark for Long-horizon Robotic Manipulation Evaluation NeurIPS 2025
Recent advances in vision-language models (VLMs) have enabled instruction-conditioned robotic systems with improved generalization. However, most existing work focuses on reactive System 1 policies, underutilizing VLMs' strengths in semantic reasoning and long-horizon planning. These System 2 capabilities-characterized by deliberative, goal-directed thinking-remain under explored due to the limited temporal scale and structural complexity of current benchmarks. To address this gap, we introduce RoboCerebra, a benchmark for evaluating high-level reasoning in long-horizon robotic manipulation. RoboCerebra includes: (1) a large-scale simulation dataset with extended task horizons and diverse subtask sequences in household environments; (2) a hierarchical framework combining a high-level VLM planner with a low-level vision-language-action (VLA) controller; and (3) an evaluation protocol targeting planning, reflection, and memory through structured System 1-System 2 interaction. The dataset is constructed via a top-down pipeline, where GPT generates task instructions and decomposes them into subtask sequences. Human operators execute the subtasks in simulation, yielding high-quality trajectories with dynamic object variations. Compared to prior benchmarks, RoboCerebra features significantly longer action sequences and denser annotations. We further benchmark state-of-the-art VLMs as System 2 modules and analyze their performance across key cognitive dimensions, advancing the development of more capable and generalizable robotic planners.
comment: 25 pages, 18 figures, Accepted by NeurIPS 2025
♻ ☆ ES-HPC-MPC: Exponentially Stable Hybrid Perception Constrained MPC for Quadrotor with Suspended Payloads
Aerial transportation using quadrotors with cable-suspended payloads holds great potential for applications in disaster response, logistics, and infrastructure maintenance. However, their hybrid and underactuated dynamics pose significant control and perception challenges. Traditional approaches often assume a taut cable condition, limiting their effectiveness in real-world applications where slack-to-taut transitions occur due to disturbances. We introduce ES-HPC-MPC, a model predictive control framework that enforces exponential stability and perception-constrained control under hybrid dynamics. Our method leverages Exponentially Stabilizing Control Lyapunov Functions (ES-CLFs) to enforce stability during the tasks and Control Barrier Functions (CBFs) to maintain the payload within the onboard camera's field of view (FoV). We validate our method through both simulation and real-world experiments, demonstrating stable trajectory tracking and reliable payload perception. We validate that our method maintains stability and satisfies perception constraints while tracking dynamically infeasible trajectories and when the system is subjected to hybrid mode transitions caused by unexpected disturbances.
comment: Accepted to IEEE Robotics and Automation Letters