Skip to content
#Data and simulation
Nov 5Thu
  1. 雷科技60

    Cainiao opens global warehouse network to provide real-world training clusters for AI robots

    On November 5, Cainiao held an AI and embodied intelligence industry cooperation launch event at its Hong Kong Smart Port, announcing that it will open its global warehouse network to industry partners and build a logistics real-world training cluster covering dozens of industries and hundreds of real-world scenarios, including warehouse picking, intelligent navigation and delivery, high-altitude inspection, and auto parts management.

  2. HuggingFace Daily Papers58

    DRAP: Dynamics-Risk Adaptive Plasticity for Continual Learning in Physical AI

    The study proposes DRAP, a replay-free continual learning method that uses locally trajectory-amplified, risk-weighted action gradients to estimate the historical importance of model tensors and scales each tensor's effective AdamW learning rate accordingly. The method is instantiated on OpenVLA's KappaTune subnetwork and sequentially adapted on the LIBERO Spatial, Object, and Goal suites; compared with standard AdamW fine-tuning and EWC, it markedly improves retention of old behaviors while maintaining the ability to learn new suites.

  4. Guyuehome47

    Deep reinforcement learning-based quadrotor target tracking control: experiments with 64×64 visual input in Isaac Lab

    A quadrotor tracks a yellow target ball using only a 64×64 camera image. The official pretrained policy was loaded to complete tracking experiments in Isaac Lab, with the target visible about 40% of the time and the quadrotor autonomously hovering to reacquire the target after losing it. The article also discusses known gaps when migrating the policy to the DJI Tello.

  5. Guyuehome60

    RoboSimStudio tested: multi-material soft-body physics simulation parameter calibration and reproduction

    In testing of the RoboSimStudio multi-material physics workbench, the three examples of rope knotting, a towel, and a canvas bag were bit-for-bit identical across two runs, and the material parameters matched the repository calibration. After line-by-line transcription of the elastica mathematics in the repository, zero-deviation reproduction was achieved.

  7. Robotics & Automation News78

    AGIBOT releases embodied foundation model GO-3 Preview, tops three embodied AI benchmarks

    AGIBOT has released GO-3 Preview, an embodied foundation model built on its self-developed WRAM (World Reasoning Action Model) architecture that unifies task understanding, future state prediction, and action generation. It ranks first on three embodied AI benchmarks: RoboDojo, RoboTwin, and RoboColiseum, and is planned for full open source release by the end of 2026.

    Why it matters: AGIBOT has disclosed the architecture and three-stage training design of GO-3 Preview, which can be compared against its technical roadmap evolution from GO-1 to GO-3.

  8. arXiv cs.RO71

    RoboTryOn: Synthesizing Robot Dexterous Manipulation Demonstrations from Human Videos via Virtual Try-On

    RoboTryOn proposes treating human-to-robot translation as virtual try-on: it preserves the demonstrated motion, objects, and background, and only replaces the human with the target robot specified by a reference image. The method autoregressively converts keyframes and uses them to guide video synthesis, with motion conditioned on the human hand skeleton; it then uses an inverse dynamics model (IDM) with depth features to infer action labels from the synthesized videos for policy learning. Experiments show that policies trained only on synthetic data improve average success rate from 2% to 60% on new tasks, including dual-arm manipulation, and reach the level of policies taught by pure teleoperation at about one-third of the collection cost.

  9. arXiv cs.RO60

    Study: How inertia mismatch affects robot policy learning when robots are heavier than objects

    This paper investigates how inertia mismatch between a robotic arm and the objects it manipulates affects robot policy learning. Through controlled simulation experiments that vary robot and object masses, it finds that reinforcement learning succeeds across a broad range of mass combinations, while behavior cloning learns more reliable policies when a lighter robot manipulates a heavier object.

  10. arXiv cs.RO67

    Tactus: A tactile and pose-sensing glove for natural dexterous manipulation data collection

    The authors propose Tactus, a textile-integrated sensing glove that provides both capacitive tactile force sensing and IMU-fused pose tracking with a 1.5 mm fingertip stackup thickness, for collecting natural dexterous manipulation demonstration data. At 95% confidence, the glove's minimum detectable force is 0.59 N, 1.6–2.7 times lower than two commercial tactile glove baselines, with cross-subject consistency of 0.87; under the most severe manipulation occlusion, the relative error of fingertip keypoints is 11.0 mm.

  11. arXiv cs.RO58

    PriMoRetarget: Humanoid Motion Transfer via Prioritized Motion Retargeting

    The authors propose PriMoRetarget, a motion retargeting framework that replaces relative weight weighting with an explicit priority hierarchy, so that lower-priority updates are constrained to preserve higher-priority objectives. The method is evaluated on four datasets covering single-person motion, person-object manipulation, and person-person interaction; the generated motions better preserve task-relevant interactions and provide reference trajectories that are easier to track than those of commonly used retargeting methods.

  12. arXiv cs.RO67

    Reverse to Advance: Learning Hard Manipulation Policies with Reversed Easy-Task Data to Cut Teleoperation Costs

    This paper proposes a teleoperation-cost-effective framework that uses temporally reversed trajectories from easy tasks as supervision signals for hard manipulation tasks, reducing manual demonstration collection. The framework has three parts: a closed-loop data collection pipeline, hierarchical data refinement based on kinematic priors and critic-guided advantage filtering, and iterative policy learning. In two simulation benchmarks and real-robot experiments, the method improved success rates on hard tasks compared with reversal-based and reinforcement-learning baselines, with higher data efficiency and more stable training.

  13. arXiv cs.RO65

    InDex: Adapting VLA models to dexterous manipulation via intent-conditioned fine-tuning

    The study proposes the InDex framework, which uses a normalized aperture signal derived from retargeted demonstrations to supervise a scalar grasp intent, adapting parallel-gripper VLA models to multi-fingered dexterous hands in two stages. Across four simulated tasks, InDex with π0.5 achieved 85.8% mean task success, a 35.5 percentage point improvement over the native policy, and reached 69.0% on new objects with similar geometry without fine-tuning. It was also effective across three VLA backbones and on a separately trained real-world policy, whose mean success rate rose from 71.0% to 83.0%.

  14. arXiv cs.RO65

    Learning Soft Robot Dynamics: An Uncertainty-Aware Active Exploration Framework

    The authors propose an uncertainty-aware active exploration framework that can autonomously learn task-agnostic, generalizable dynamics models for soft robots. The method was evaluated on three simulation platforms—a continuum arm, an articulated fish in fluid, and a musculoskeletal leg—as well as on a real pneumatically actuated continuum arm. Compared with random exploration and task-specific baselines, it achieves higher modeling accuracy and yields higher returns in zero-shot control on unseen tasks, while remaining robust under sensing noise, actuation delays, and material nonlinearities.

  15. arXiv cs.RO58

    Fluidic Body: Controlling swarm robots as a fluid extension of the hand

    The authors propose the Fluidic Body system, which lets an operator change the properties of a virtual fluid between the hand and a swarm robot collective through hand movements, thereby adjusting swarm behavior: opening the hand makes the swarm soft and deformable so it can flow around obstacles, while clenching the fist makes the swarm hard and cohesive so it can aggregate into a single body to push objects.

  16. arXiv cs.RO84

    HALO: Humanoid Autonomy System Trained on Synthetic Simulation Data Achieves Zero-Shot Real-World Transfer

    The research team proposes HALO, a humanoid robot autonomy system that synthesizes vision-language-action data across roughly one thousand procedurally generated home scenes, eliminating reliance on teleoperation data collection. HALO learns skills including sitting on chairs, navigating to specified objects, stepping sideways through half-open doors, crouching and crawling under tables, and pushing boxes, and can combine them to complete long-horizon tasks. Without ever seeing real home scenes during training, HALO transfers zero-shot to the physical world, achieving a 91% success rate across 360 trials in six unmapped real-world environments.

    Why it matters: The paper presents a complete path from synthesizing vision-language-action data using procedurally generated home scenes to zero-shot transfer to real robots; its approach to data scaling is worth referencing.

  17. arXiv cs.RO65

    VABAR: A Variational Automation Benchmark for Agentic Robotics released

    The authors propose VABAR, a Variational Automation Benchmark for agentic robotics, built on open-source tools and the Newton physics engine for simulated manipulation evaluation. The benchmark covers six manipulation task classes: tool sorting, tool hanging, gear meshing, syringe packing, cable insertion, and cable routing. It includes standardized task variants, difficulty levels, success metrics, 33 generalizable skills, and human-written programmatic reference solutions. The paper also presents an initial baseline evaluation based on Graph-as-Policy (GaP) and Isaac Lab Arena.

  18. arXiv cs.RO62

    Uncertainty-Triggered Aerial Refinement for Off-Road Navigation

    The paper proposes a planner-conditioned air-ground assistance method that evaluates how map uncertainty affects the UGV's current MPPI rollouts. When navigation remains degraded, it re-scores the same rollouts under resolved map hypotheses, selects the map layer whose request yields the largest improvement, requests the corresponding aerial measurement, and fuses it to replan.

  19. arXiv cs.RO72

    soccerbot: Learning Humanoid Soccer Skills with Latent Motion Priors

    The authors propose soccerbot, a hierarchical reinforcement learning framework that distills a structured latent motion prior from large-scale human soccer motion capture data, constraining policy learning within a dynamically feasible motion manifold. A single skill-conditioned policy thereby learns moving, dribbling, trapping, and kicking in a unified way, while maintaining human-like motion and fluid transitions during skill switching. After robustness-oriented simulation design, the framework transfers zero-shot to the Unitree G1 humanoid robot, completing the full soccer skill set in real-world environments.

  20. arXiv cs.RO62

    Failure-Boundary Profiles: A New Method for Closed-Loop Evaluation of Driving and Robot Navigation Agents

    The study proposes a failure-boundary profile evaluation method in which each scenario family varies only one measurable factor (such as corridor width or the time gap in oncoming traffic), and each variant is labeled from measurement as one where the agent should proceed or one where it should stop. Along this axis, two curves are estimated for each agent: commitment (the probability of attempting the passage) and competence (the probability of passing without contact). The method locates where each curve crosses 0.5 and gives intervals that account for the simulator's run-to-run variance, while also using certified labels to derive a decision accuracy metric.

  21. arXiv cs.RO64

    TwinBridge: Using digital twins to transfer policies and rewards for real-world reinforcement learning in laboratory robotics

    TwinBridge proposes a digital twin-driven framework that provides policy initialization and dense visual rewards for real-world online reinforcement learning. The method builds digital twins by combining 3D Gaussian splatting with physics simulation; simulator states provide task events and geometric progress, which are used for pretraining with simulated demonstrations and for reinforcement learning, then distilled into a visual reward model. Combined with execution history, this computes dense rewards and supports human-in-the-loop policy optimization.

  22. arXiv cs.RO56

    Learning manipulation actions from demonstrations: online adaptation to latent physical properties

    The authors propose a learning-from-demonstrations method that jointly learns a dynamics model and object-specific latent representations of physical properties, and conditions an imitation policy on online estimates of these representations, enabling the robot to adjust manipulation actions such as regrasping and contact locations based on physical-property information acquired through interaction. During execution, the dynamics model and policy weights remain fixed, and only the latent representation is updated based on the discrepancy between predicted and actual observations. In real-world and simulation experiments with objects having different centers of mass, online adaptation improved task success rates, but in simulation the success rate remained about 30% at center-of-mass positions not covered by the demonstrations, indicating limited adaptation to unseen physical properties.

  23. arXiv cs.RO65

    Learning Object-Centric Motion Priors from Human Data for Robotic Dexterous Manipulation

    The paper proposes learning to holistically predict future hand-object states from human-object interaction (HOI) datasets, using the predicted future object states as a general-purpose reward term for reinforcement learning, reducing reliance on task-specific reward engineering. The authors experiment on three manipulation tasks in simulation and the real world, reporting better success rates and generalization to novel objects than existing SOTA methods, and validate cross-embodiment compatibility by deploying skills to different robot hands. The work has been accepted as an oral paper at AAAI 2026.

  24. arXiv cs.RO58

    Diversity-Aware Experience Replay via Statistical Leverage Scores

    The paper proposes scoring each transition in the replay buffer with a statistical leverage score and drawing mini-batches from a mixture of this score and uniform sampling, to alleviate the under-sampling of rare events in robot offline experience. When actions are stored as one-hot encodings, the method instead uses the feature column with the largest kurtosis, and it drops the importance-weight correction of prioritized replay, directly optimizing a rebalanced objective; for rare cases related to the target, it applies the same tail rule to short trajectory windows.

  25. arXiv cs.RO58

    Can RRT* work on non-flat terrain triangle meshes?

    The author examines the applicability of RRT* to triangle-mesh terrain, noting that the original RRT* cannot run directly on a mesh because the surface has zero volume, so uniform sampling never hits it. The proposed Mesh RRT* solved every run on terrain with slopes up to 75 degrees, but its cost function is biased: after 30,000 iterations the reported cost is on average up to 2.7% below the exact geodesic, and the returned path is on average 0.1-0.5% longer.

  27. arXiv cs.RO57

    Multi-Robot Closed-Loop Adaptive Sampling Framework Based on a Generative Ocean Model

    The authors propose a multi-robot closed-loop adaptive sampling framework in which a single learned generative model performs forecasting, uncertainty estimation, planning, and measurement assimilation simultaneously. The model generates ensemble forecasts in latent space and decodes them into physical ocean states, uses ensemble statistics to construct A- or D-optimal sampling objectives, selects sampling locations within the robot-reachable range constrained by predicted currents, and feeds measurements back into the latent state through observation-conditioned diffusion sampling.

  28. arXiv cs.RO62

    OdorSpaces: A Visual-Olfactory Embodied Navigation Simulation Platform and Benchmark for 3D Environments

    The authors propose OdorSpaces, a physics-simulation-based platform and benchmark for visual-olfactory embodied navigation. It integrates 3D odor transport and dispersion simulated by Fire Dynamics Simulator (FDS) into Habitat-Sim, enabling agents to navigate realistic indoor environments using both visual and olfactory observations.

  29. arXiv cs.RO62

    Contact-Implicit Whole-Body Trajectory Optimization for Humanoid Robots on Parameterizable Uneven Terrain

    This paper proposes a contact-implicit whole-body optimization framework for humanoid robots to generate dynamic motion trajectories on parameterizable complex terrains such as steps and ramps, without requiring a predefined contact sequence. The method approximates ground geometry with smooth surfaces and extends rigid-body interaction from planar surfaces to arbitrary smooth surfaces through new contact constraints, using nonlinear programming to plan motions across various terrain topologies. The authors conducted preliminary feasibility tests on non-coplanar terrains in simulation and on the real robot ErgoCub.

  32. Guyuehome58

    SkillMimic uses multiplicative rewards and hierarchical skill control to imitate basketball skills

    SkillMimic uses multiplicative Gaussian kernel rewards to turn pose, interaction, and contact into multiplicative joint conditions, learning 19 types of interactive skills from basketball demonstrations. Its hierarchical version uses discrete skill selection to drive a frozen low-level policy and reproduces four types of tasks on a new-generation simulation platform.

  33. HuggingFace Daily Papers78

    PyRUA-DataFarm: Using LLM agents instead of teleoperation to generate real-ready robot data

    PyRUA-DataFarm proposes having a coding agent rebuild tasks in simulation, with a video model rendering them into real-ready data, replacing teleoperation for collecting robot demonstrations. On five real-robot tasks, policies trained only on this data achieved 93/100 successes, versus 84/100 for policies trained on teleoperated data; under an unseen tablecloth and colored lighting, the former kept 87 successes while the latter fell to 46.

    Why it matters: The paper presents a complete pipeline and cost comparison for using a coding agent to replace teleoperation in collecting demonstrations, which can be used to assess the room for replacing this data-collection route.

  34. The Humanoid Hub62

    Perceptron launches Egocentric 1.0 self-serve API to turn first-person video into training data

    Perceptron has released the Egocentric 1.0 self-serve API, which converts first-person human and robot video into structured training data, outputting a timestamped subtask timeline (with sub-second-level action precision), independent activity timelines for each hand or gripper, and 3D hand reconstruction without camera calibration. The API claims to match GPT-6 Astra on subtask quality at about 1/8 the cost ($9 per video hour versus about $76), with a full-stack price of $29.

  35. 简智机器人 · X20

    GenRobot's DAS 2.0 to debut at CoRL 2026

    🚀 5 days to go until CoRL 2026! Meet GenRobot's next-generation data collection hardware, DAS 2.0, featuring Dex 2.0 and Ego 2.0. From precise hand motion and tactile sensing to first-person multimodal collection, learn how we are scaling human data for embodied intelligence. 📍 Booth 200 | JW Marriott Austin, 2nd floor 📅 November 9–11, 2026 See you in Austin! #CoRL2026 #GenRobot #EmbodiedAI #HumanData #PhysicalAI

  36. 机器人全球资讯 · 微信公众号65

    UBTech releases Thinker-Sim embodied intelligence simulation platform

    UBTech has released Thinker-Sim, an embodied intelligence simulation platform that builds a one-stop system covering digitization of real scenarios, large-scale synthetic simulation data, and robot capability evaluation. The platform uses the self-developed Thinker-GS high-fidelity scene reconstruction algorithm and an end-to-end SimReady asset automated generation technology, enabling 3D reconstruction based on real factory production line environments and flexible loading of foreground assets such as robots, production equipment, and personnel.

  38. 机器人前瞻 · 微信公众号72

    LiberAI releases world action model Liber-0, pretrained on 2.2 million hours of human data

    Physical world model startup LiberAI has released the world action model Liber-0, pretrained on over 2.2 million hours of human experience data, converting human manipulation experience into robot manipulation capability. In 60 tasks randomly generated by GPT-6 and 600 real-robot tests, Liber-0 achieved a zero-shot overall success rate of 36.0%, reaching 63.5% on pick-and-place tasks; the GPT-6 Astra control achieved an overall success rate of only 5%.

  39. 未来机器人网 · 微信公众号77

    D-Robotics and University of Tokyo release RawVLA: frozen VLA, only imaging front end trained, low-light dual-arm task success rate rises from 0 to 66.5%

    D-Robotics, together with the University of Tokyo and other institutions, released the RawVLA technical approach, which adds a learnable neural ISP module before the VLA policy while keeping the VLA model frozen, using action error to train the imaging front end in reverse. On real robots, four dual-arm tasks were tested with π0.5; the average success rate in low light rose from 0% with the default ISP to 66.5%, versus 75.0% in normal light, where the default ISP achieved 76.5%.

    Why it matters: The approach brings the imaging front end into the learnable pipeline and provides comparative data under low light and normal light, offering a reference for vision pipeline design.

  40. 未来机器人网 · 微信公众号40

    Wang Xingxing at Wuzhen: 2027 likely the tipping point for embodied intelligence; humans' core value is setting goals and scoring for AI

    At the 2026 World Internet Conference Wuzhen Summit, Unitree founder Wang Xingxing judged that next year is likely the tipping point for the robotics industry or robot models, based on the rapid capability gains of foundation models such as GPT-6. He said the biggest current bottleneck is aligning models with real machines in the "last few millimeters" of precision assembly and flexible manipulation. Unitree is advancing a physical AI robot self-evolution system: AI agents autonomously search papers and open-source solutions to generate control code, which is validated in simulation and then deployed for real-machine testing, followed by closed-loop iteration combining automated evaluation and human scoring.