Anqing Duan

Open-H-Embodiment Consortium, Nigel Nelson, Juo-Tung Chen et al.

Autonomous medical robots hold promise to improve patient outcomes, reduce provider workload, democratize access to care, and enable superhuman precision. However, autonomous medical robotics has been limited by a fundamental data problem: existing medical robotic datasets are small, single-embodiment, and rarely shared openly, restricting the development of foundation models that the field needs to advance. We introduce Open-H-Embodiment, the largest open dataset of medical robotic video with synchronized kinematics to date, spanning more than 49 institutions and multiple robotic platforms including the CMR Versius, Intuitive Surgical's da Vinci, da Vinci Research Kit (dVRK), Rob Surgical BiTrack, Virtual Incision's MIRA, Moon Surgical Maestro, and a variety of custom systems, spanning surgical manipulation, robotic ultrasound, and endoscopy procedures. We demonstrate the research enabled by this dataset through two foundation models. GR00T-H is the first open foundation vision-language-action model for medical robotics, which is the only evaluated model to achieve full end-to-end task completion on a structured suturing benchmark (25% of trials vs. 0% for all others) and achieves 64% average success across a 29-step ex vivo suturing sequence. We also train Cosmos-H-Surgical-Simulator, the first action-conditioned world model to enable multi-embodiment surgical simulation from a single checkpoint, spanning nine robotic platforms and supporting in silico policy evaluation and synthetic data generation for the medical domain. These results suggest that open, large-scale medical robot data collection can serve as critical infrastructure for the research community, enabling advances in robot learning, world modeling, and beyond.

An Dinh Vuong, Tuan Van Vo, Abdullah Sohail et al.

World Models (WMs) have emerged as a promising approach for post-training Vision-Language-Action (VLA) policies to improve robustness and generalization under environmental changes. However, most WM-based post-training methods rely on pixel-space supervision, making policies sensitive to pixel-level artifacts and hallucination from imperfect WM rollouts. We introduce World2Act, a post-training framework that aligns VLA actions directly with WM video-dynamics latents using a contrastive matching objective, reducing dependence on pixels. Post-training performance is tied to rollout quality, yet current WMs struggle with arbitrary-length video generation as they are mostly trained on fixed-length clips while robotic execution durations vary widely. To address this, we propose an automatic LLM-based skill-decomposition pipeline that segments high-level instructions into low-level prompts. Our pipeline produces RoboCasa-Skill and LIBERO-Skill, supporting skill-compositional WMs that remain temporally consistent across diverse task horizons. Empirically, applying World2Act to VLAs like GR00T-N1.6 and Cosmos Policy achieves state-of-the-art results on RoboCasa and LIBERO, and improves real-world performance by 6.7%, enhancing embodied agent generalization.

Fangyuan Wang, Shipeng Lyu, Peng Zhou et al.

Enabling humanoid robots to perform long-horizon mobile manipulation planning in real-world environments based on embodied perception and comprehension abilities has been a longstanding challenge. With the recent rise of large language models (LLMs), there has been a notable increase in the development of LLM-based planners. These approaches either utilize human-provided textual representations of the real world or heavily depend on prompt engineering to extract such representations, lacking the capability to quantitatively understand the environment, such as determining the feasibility of manipulating objects. To address these limitations, we present the Instruction-Augmented Long-Horizon Planning (IALP) system, a novel framework that employs LLMs to generate feasible and optimal actions based on real-time sensor feedback, including grounded knowledge of the environment, in a closed-loop interaction. Distinct from prior works, our approach augments user instructions into PDDL problems by leveraging both the abstract reasoning capabilities of LLMs and grounding mechanisms. By conducting various real-world long-horizon tasks, each consisting of seven distinct manipulatory skills, our results demonstrate that the IALP system can efficiently solve these tasks with an average success rate exceeding 80%. Our proposed method can operate as a high-level planner, equipping robots with substantial autonomy in unstructured environments through the utilization of multi-modal sensor inputs.

Haoran Ding, Liang Ma, Yaxun Yang et al.

Visuomotor policies learned from demonstrations often overfit to nuisance visual factors in raw RGB observations, resulting in brittle behavior under appearance shifts such as background changes and object recoloring. We propose a task-aware observation interface that canonicalizes visual input into a shared representation, improving robustness to out-of-distribution (OOD) appearance changes without modifying or fine-tuning the policy. Given an RGB image and an open-vocabulary specification of task-relevant entities, we use SAM3 to segment the target object and robot/gripper. We construct an L0 observation by repainting segmented entities with predefined semantic colors on a constant background. For tasks requiring stronger geometric cues, we further inject monocular depth from Depth Anything 3 into the segmented regions via depth-guided overwrite, yielding a unified semantic--geometric observation (L1) that remains a standard 3-channel, image-like input. We evaluate on RoboMimic (Lift), ManiSkill YCB grasping under clutter, four RLBench tasks under controlled appearance shifts, and two real-world Franka tasks (ReachX and CloseCabinet). Across benchmarks and policy backbones (Flow Matching Policy and SmolVLA), our interface preserves in-distribution performance while substantially improving robustness under OOD visual shifts.

Shengzeng Huo, Anqing Duan, Chengxi Li et al.

This paper addresses the problem of contact-based manipulation of deformable linear objects (DLOs) towards desired shapes with a dual-arm robotic system. To alleviate the burden of high-dimensional continuous state-action spaces, we model the DLO as a kinematic multibody system via our proposed keypoint detection network. This new perception network is trained on a synthetic labeled image dataset and transferred to real manipulation scenarios without conducting any manual annotations. Our goal-conditioned policy can efficiently learn to rearrange the configuration of the DLO based on the detected keypoints. The proposed hierarchical action framework tackles the manipulation problem in a coarse-to-fine manner (with high-level task planning and low-level motion control) by leveraging on two action primitives. The identification of deformation properties is avoided since the algorithm replans its motion after each bimanual execution. The conducted experimental results reveal that our method achieves high performance in state representation of the DLO, and is robust to uncertain environmental constraints.

Peng Zhou, Omar Zahra, Anqing Duan et al.

Cloth folding is a widespread domestic task that is seemingly performed by humans but which is highly challenging for autonomous robots to execute due to the highly deformable nature of textiles; It is hard to engineer and learn manipulation pipelines to efficiently execute it. In this paper, we propose a new solution for robotic cloth folding (using a standard folding board) via learning from demonstrations. Our demonstration video encoding is based on a high-level abstraction, namely, a refined optical flow-based spatiotemporal graph, as opposed to a low-level encoding such as image pixels. By constructing a new spatiotemporal graph with an advanced visual corresponding descriptor, the policy learning can focus on key points and relations with a 3D spatial configuration, which allows to quickly generalize across different environments. To further boost the policy searching, we combine optical flow and static motion saliency maps to discriminate the dominant motions for better handling the system dynamics in real-time, which aligns with the attentional motion mechanism that dominates the human imitation process. To validate the proposed approach, we analyze the manual folding procedure and developed a custom-made end-effector to efficiently interact with the folding board. Multiple experiments on a real robotic platform were conducted to validate the effectiveness and robustness of the proposed method.

Hoi-Yin Lee, Peng Zhou, Anqing Duan et al.

This paper presents the development of a multi-sensor user interface to facilitate the instruction of arc welding tasks. Traditional methods to acquire hand-eye coordination skills are typically conducted through one-to-one instruction where trainees must wear protective helmets and conduct several tests. This approach is inefficient as the harmful light emitted from the electric arc impedes the close monitoring of the process; Practitioners can only observe a small bright spot. To tackle these problems, recent training approaches have leveraged virtual reality to safely simulate the process and visualize the geometry of the workpieces. However, the synthetic nature of these types of simulation platforms reduces their effectiveness as they fail to comprise actual welding interactions with the environment, which hinders the trainees' learning process. To provide users with a real welding experience, we have developed a new multi-sensor extended reality platform for arc welding training. Our system is composed of: (1) An HDR camera, monitoring the real welding spot in real-time; (2) A depth sensor, capturing the 3D geometry of the scene; and (3) A head-mounted VR display, visualizing the process safely. Our innovative platform provides users with a "bot trainer", virtual cues of the seam geometry, automatic spot tracking, and performance scores. To validate the platform's feasibility, we conduct extensive experiments with several welding training tasks. We show that compared with the traditional training practice and recent virtual reality approaches, our automated multi-sensor method achieves better performances in terms of accuracy, learning curve, and effectiveness.