Skip to content
Only Research#Software and compute systems
Nov 9Mon
  1. arXiv cs.RO62

    Autoregressive policies achieve real-time execution via tokenization horizon adjustment and constrained decoding

    The paper proposes that autoregressive policies can achieve real-time execution by adjusting the tokenization horizon and applying constrained decoding, thereby guaranteeing strict latency bounds and enabling multi-trajectory decoding. In simulated and real-world environments, the autoregressive policy significantly improves task completion speed over synchronous inference and consistently outperforms equivalent-level flow-matching policies. The authors note that autoregressive policies also have inherent advantages of faster convergence and better generalization in instruction following. The work was published at CoRL 2026.

  4. arXiv cs.RO79

    A2-MEAS: An Embodied Multi-Agent Framework with Code as a Scaling Interface

    The authors propose A2-MEAS, an embodied multi-agent framework that uses code as a scaling interface, combining corrective language, spatial evidence, execution history, and action proposals into correction codes; during training, MetaBrain provides supervision and runtime guidance.

    Why it matters: The framework uses code as a scaling interface, injects corrective language and execution history into the policy, and provides quantitative comparisons across benchmarks and real-robot tasks.

  5. arXiv cs.RO67

    Lodos: A lightweight simulator for contact-rich interactions with detailed meshes

    Lodos is a simulation framework designed from scratch, using new broadphase and narrowphase collision detection and contact dynamics computation to specifically handle mesh-mesh collisions between detailed meshes. The authors state that its physical fidelity is comparable to existing mature simulators, and that in mesh-mesh collision scenarios its throughput is about an order of magnitude higher; the source code has been released. The work targets contact-rich scenarios in large-scale parallel rigid-body simulation that are limited by approximating collision geometry with primitives such as spheres, capsules, cylinders, and boxes.

  6. arXiv cs.RO62

    ActPrune: Action-Grounded Visual Token Pruning for Efficient Vision-Language-Action Models

    ActPrune proposes a training-free visual token pruning method that moves the pruning decision ahead of the vision encoder and uses the action chunk predicted at the previous control step as a geometric prior. In the fixed third-person view, the method retains image patches near the projected end-effector trajectory along with semantic anchors; in the wrist view, it prunes patches that remain unchanged after compensating for action-induced camera motion.

Nov 6Fri
  1. 机器人网 · 微信公众号47

    Avnet and National-Local Joint Humanoid Robot Innovation Center Release 76-Page Embodied Intelligence Chip Selection White Paper

    Avnet, together with the National-Local Joint Humanoid Robot Innovation Center, produced a 76-page white paper (August 2026) that uses five functional domains and six core modules as its framework to break down 10 categories of industrial deployment challenges, including heterogeneous computing workloads and platform adaptation, multi-axis real-time control, and dexterous hand integration.

Nov 5Thu
  1. arXiv cs.RO62

    RegimeVGGT: A Training-Free Acceleration Method for VGGT That Preserves Spatial Structure Layer by Layer

    RegimeVGGT proposes a training-free, layer-wise compression method that achieves a 6.7x speedup over the memory-optimized VGGT* with unchanged reconstruction quality. Through spectral analysis, probing, and causal analysis, the authors find that VGGT's shallow layers lack cross-view structure, its middle layers handle cross-view alignment, and its deep layers are redundant for dense geometry, but their cross-frame attention remains critical for pose.

  2. arXiv cs.RO56

    Reflection-Robust Stereo VIO: Removing Reflective Virtual Features with Kalman-Filtered Ground-Height Estimation

    This paper proposes a lightweight reflection-aware front-end for feature-based stereo visual-inertial odometry (VIO) that estimates ground height in real time and rejects only feature candidates whose inferred 3D positions lie below the estimated ground. The method uses a ground-region mask to restrict the validation scope, constructs frame-level ground-height measurements from stereo height uncertainty and pose uncertainty, stabilizes the estimate with a scalar Kalman filter with NIS gating, and performs selective rejection with an uncertainty-aware safety margin.

Nov 4Wed
  1. 机器人前瞻 · 微信公众号72

    Tsinghua and Berkeley open-source embodied agent uni-robot-agent; Qwen3.8-Flash-Next scores above GPT-6 Astra on RoboDojo

    Tsinghua University and UC Berkeley jointly proposed and open-sourced uni-robot-agent, a cross-environment unified embodied agent harness for connecting tool interfaces across different robots and simulation environments. Based on this framework, Qwen3.8-Flash-Next scored 31.39 on the RoboDojo benchmark, exceeding the publicly reported 28.97 for GPT-6 Astra, at about 1/20 of Astra's cost.

Nov 2Mon
  2. arXiv cs.RO62

    Extending the Franka Emika Panda's Life: A ROS 2 Software Stack with a Reliable Position Interface

    This paper releases an open-source ROS 2 software stack that restores ROS 2 support for the Franka Emika Panda and resolves the long-standing unreliability of its external position control interface. The authors' analysis attributes vibrations and protective stops to external control loop timing and sampling jitter rather than to limitations of the robot itself, and on that basis they propose an asynchronous hardware interface that decouples real-time communication from the ROS 2 control loop, a rate-matching mechanism for slower command sources, and a position-domain reference generation strategy.

  3. arXiv cs.RO60

    Sensorium: An open-source multimodal data collection and curation library for embodied surgical AI

    The authors released Sensorium, an open-source Python library built on ROS and ROS2 middleware that records, processes, annotates, and curates multimodal surgical data from configurable robot and sensor streams, covering operator-side acquisition, configurable processing, semantic annotation, and episodic dataset generation. The library was evaluated on 1516 recordings of phantom, ex vivo, and in vivo porcine experiments involving multiple surgical tasks and sensing configurations, with preliminary usability feedback reported.

  5. arXiv cs.RO69

    cuPIQP: A Differentiable Batched QP GPU Solver for Robot Learning and Control

    cuPIQP is a GPU batched differentiable quadratic programming (QP) solver based on a proximal interior-point method, with unified support for dense, sparse, and multistage problems. In benchmarks with up to 16,384 QPs per batch, it solves up to 18.8 times faster than Moreau and 4.26 times faster than eight-core PIQP; in an end-to-end learning task, training is accelerated by 4.6 times over QPAX and 7.8 times over Moreau.

  7. arXiv cs.RO67

    Higher-Order Kinematic Derivative Methods for Aspect-Preserving Real-Time Numerical Inverse Kinematics

    The paper uses higher-order kinematic derivatives of serial robots to propose an aspect-preserving real-time numerical inverse kinematics method, yielding the QuIK4 and QuIK5 solvers with fourth- and fifth-order convergence. The method also includes an aspect-preserving step limiter that extrapolates the Jacobian determinant along the step to predict and avoid singularities, as well as a certifier that uses Hermite interpolation to verify the aspect of solutions and a certified restart algorithm. The authors claim the failure rate is reduced by up to four orders of magnitude compared with existing methods, and they release a C++ library with ROS 2 integration.

  8. arXiv cs.RO58

    GPU-Accelerated Incremental Euclidean Distance Transform for Online Motion Planning of Mobile Robots

    This paper proposes a voxel mapping system that uses parallel computing to simultaneously generate occupancy grid maps (OGMs) and Euclidean distance transforms (EDTs), and incrementally builds a global EDT to output high-frequency local distance information for online motion planning. The system uses GPU programming to compute EDT in parallel within local volumes, then continuously integrates new observations into the global EDT via a parallel wavefront algorithm while preserving historical observations.

  9. arXiv cs.RO66

    RTPolicy: Real-time multi-view visuomotor policy for edge devices, inference at 32.9 Hz

    RTPolicy is a multi-view visuomotor policy that uses view-aware feature fusion and visual compression to aggregate multi-camera information into a compact token representation, and uses a flow-matching action generator to reduce action-generation overhead. On the 50 tasks of the RoboTwin 2.0 single-task clean evaluation, it achieved an average success rate of 52.26%, higher than the strongest leaderboard baseline π0 at 46.4%.

  10. arXiv cs.RO58

    Physically Constrained Kinematic Calibration for High-Precision Robotic Needle Guidance

    The paper proposes a bound-constrained kinematic calibration method to improve the positioning accuracy of a 6-DoF robotic needle-guidance manipulator. The method constrains a 19-parameter geometric identification model with physically plausible parameter ranges, treating the bounds as active optimization constraints to mitigate parameter coupling and overfitting. In simulation, end-effector RMSE dropped from 3.422 mm to 0.690 mm (a 79.8% improvement); in benchtop experiments, trajectory validation RMSE was 0.865 mm and coordinate grid validation RMSE was 0.610 mm.

  12. arXiv cs.RO60

    Veloxity: Memory-Safe UAV Embedded Firmware Written in Rust

    The authors present Veloxity, an open-source embedded UAV autopilot written entirely in the memory-safe Rust language, offering a minimal, modular foundation for autonomous drones. They compared it with equivalent C++ autopilots in simulation, on hardware testbenches, and in outdoor waypoint-tracking flight tests; the results showed functional parity with improved performance, indicating that memory-safe languages can be used for reliable, high-performance autonomous aerospace firmware.

Oct 30Fri
  1. 机器之心 · 微信公众号88

    Tsinghua, AgiBot and Infinigence AI propose Collective Experience Learning (CEL), validated on 96 real robots for scalable gains

    Tsinghua University, AgiBot and Infinigence AI jointly proposed the Collective Experience Learning (CEL) paradigm and completed prototype validation on 96 robots. In a phone-packaging task, the trained policy completed 120 tests on two robots of the same model that did not participate in training, raising the success rate from 79.2% to 98.3% and reducing failures from 25 to 2.

    Why it matters: The 96 real robots validated the scalable gains of collective experience sharing, giving readers a way to understand one path toward robots autonomously accumulating experience instead of relying on human demonstrations.

  2. Xbotics具身智能实验室 · 微信公众号88

    Infinigence AI, Tsinghua and AgiBot propose collective experience learning (CEL) paradigm, validated on 96 robots

    Infinigence AI, together with Tsinghua University and AgiBot, proposed the collective experience learning (CEL) paradigm and completed prototype validation on 96 robots. In a phone packaging task, a policy trained with CEL completed 120 tests on two robots of the same model that did not participate in training, with success rate rising from 79.2% to 98.3%; when the number of robots participating in learning increased from 16 to 96, the test success rate rose from 86.7% to 98.3%.

    Why it matters: Readers can learn about a training paradigm that lets robots share successes and failures and reduces reliance on human demonstration, as well as its validation results on 96 real robots.

  3. 3D视觉工坊 · 微信公众号88

    Infinigence AI, Tsinghua and AgiBot propose Collective Experience Learning (CEL), validated on 96 robots with 98.3% success rate

    Infinigence AI, together with Tsinghua University and AgiBot, proposed the Collective Experience Learning (CEL) paradigm and completed prototype validation on 96 robots. In a phone packaging task, policies trained with CEL completed 120 tests on two robots of the same model that did not participate in training, with the success rate rising from 79.2% to 98.3%; when the number of robots participating in learning increased from 16 to 96, the test success rate rose from 86.7% to 98.3%.

    Why it matters: Fault-tolerant asynchronous training across a 96-robot real-machine cluster and failure-guided experience retrieval provide a reproducible system and scale-benefit data for Collective Experience Learning.

  4. 量子位72

    Tsinghua University, Infinigence AI, and AgiBot propose collective experience learning CEL, validated on a 96-robot cluster

    Tsinghua University, Infinigence AI, and AgiBot jointly proposed the collective experience learning (CEL) paradigm, which lets robots share the successful experience of their peers rather than relying only on human demonstrations, and validated it on a cluster of 96 real robots. Two robots that did not participate in training were tested 600 times in total across five tasks—flower arranging, clothes folding, phone charging, stool assembly, and phone packaging—with 15 failures, for an overall success rate of 97.5%, all exceeding the average level of 6 human operators; the success rate on the phone packaging task rose from 79.2% to 98.3%.

Oct 28Wed
  1. arXiv cs.RO60

    From Language to Navigation Goals: A Vision-Language Approach for Semantic Navigation of Mobile Robots Using RGB-D Perception

    The study proposes a language-driven navigation framework that enables mobile robots to understand natural-language requests such as "go to the mailbox," identify the referenced object, estimate its position using RGB-D data, and generate a navigation goal that is then executed by the ROS 2 Nav2 navigation stack.

  2. arXiv cs.RO62

    RAPID: A Robotic Agent Platform for EV Battery Pack Disassembly

    The authors propose RAPID, a robotic agent disassembly platform that integrates a gantry-mounted industrial robotic arm, RGB-D perception, and an automated nut-running tool to remove fasteners on a full-scale EV battery pack. Open-vocabulary object detection achieves 0.9757 mAP50; among three one-shot fastener removal strategies, taught-in poses achieved a 97% success rate in 24 minutes, one-shot vision execution achieved 57% in 29 minutes, and visual servoing achieved 83% in 36 minutes (n=204).

  3. arXiv cs.RO42

    ASPEN: Adaptive Semantic Reasoning via Command Persistence Prediction to Improve Navigation VLA Efficiency

    ASPEN is an event-triggered scheduler based on command persistence prediction, designed to reduce repeated semantic reasoning calls in navigation VLA policies. On frozen NaVILA in the cross-dataset RxR-CE val-unseen evaluation, ASPEN reduces VLA calls by 9.5% and estimated inference time by 8.1% (including scheduling overhead), while preserving action sequences identical to the baseline in 92.1% of episodes.

  4. arXiv cs.RO67

    cuArms: A GPU-Accelerated Conflict-Based Planner for Multi-Arm Manipulation

    cuArms is a conflict-based planner co-designed with GPU parallelism. It decomposes the search space using reusable multi-modal roadmaps and resolves conflicts and deadlocks through lazy search. It integrates with a high-level scheduler to jointly optimize task assignment, scheduling, motion, and manipulation skills. Experiments report real-time initial solutions and anytime improvements in motion quality and scheduling, without domain training for learning-based methods. The authors say it provides demo videos on both simulated and real industrial tasks.

Oct 27Tue
  1. CVer · 微信公众号72

    FastHarness: A Robot Execution Framework That Schedules LLM Review Using Task Experience

    FastHarness is an experience-driven robot execution framework that uses π0.5 to generate actions and GPT-6 Astra to review and correct them, adjusting the review interval according to task stage and past review results. Across four simulated manipulation tasks in RoboDojo, it achieved an average success rate of 97.19%, higher than the 80% reported for π0.5 + GPT-6 Astra in the GPT-as-Policy report.

  2. CVer · 微信公众号77

    Tsinghua AIR proposes OxyGen: unified KV Cache management lets robots talk while acting

    The Institute for AI Industry Research (AIR) at Tsinghua University, together with Tsinghua's Department of Electronic Engineering and the University of Science and Technology of China, has proposed OxyGen, which uses unified KV Cache management to let an embodied model's action and language reasoning share prefix computation. The paper has been accepted to NeurIPS 2026.

    Why it matters: The paper presents a unified KV Cache management scheme and measured data across multiple platforms, offering a reference for on-device embodied inference deployment.

  3. arXiv cs.RO58

    ActECC: An Action Chunk Temporal Scheduling System for Edge-Cloud Collaborative VLA Inference

    The authors propose ActECC, which addresses the difficulty of determining the valid range of action chunks and scheduling replanning requests in edge-cloud collaborative VLA inference through kinematics-guided usable range assessment and delay-aware replanning scheduling. Experiments show it reduces per-action latency by up to 27.4%, cuts cloud calls by up to 45.1%, and lowers average execution error.

  4. arXiv cs.RO47

    Robot Motion as a Keyed Communication Channel: Transmitting Messages via Motion Steganography

    Researchers propose a "motion communication" method that encodes arbitrary short messages (such as an agent's current intent) as noise into the actions of any pre-trained policy, making robot motion detectable via remote vision or motion capture without degrading policy performance. This physical channel does not rely on wireless communication and requires no additional hardware or direct link to the robot. On real robots running at 50 Hz, four robots jointly recovered an 8-bit message at a total rate of 0.8 bits/s, with validation completed in both simulated environments and real-robot deployment.

  5. arXiv cs.RO60

    Plug, Play and Comply: A Plugin-Based Compliant Control Framework in ROS 2

    The paper proposes a robot-agnostic compliant control framework that separates ROS controller boilerplate from control-law implementation through a plugin-based execution architecture, extending the ROS control ecosystem. The framework uses Pinocchio to compute robot kinematics and dynamics from URDF models, and its standardized joint and Cartesian command interfaces support motion references, stiffness and damping gains, nullspace targets, desired interaction wrenches, and independent feedforward inputs.

  6. arXiv cs.RO65

    DWPP: Dynamic Window Pure Pursuit Considering Velocity and Acceleration Constraints

    The authors propose dynamic window pure pursuit (DWPP), which moves command velocity computation into the v-ω velocity space and selects, within the dynamic window, the point closest to the line ω=κv, thereby explicitly incorporating velocity and acceleration constraints. Experiments show that DWPP avoids constraint-violating commands and achieves better path-tracking accuracy than conventional pure pursuit methods. The method has been integrated into the official Nav2 repository and made publicly available, and the paper has been accepted by Robotics and Autonomous Systems.

  7. arXiv cs.RO66

    OSWAM: A World Action Model with One-Step Action Generation

    OSWAM proposes a world action model that generates actions with just one diffusion step, for autonomous driving and embodied control. The authors initialize the diffusion process with imagination tokens to provide physical priors and action-conditioned representations, and introduce privileged information distillation with future conditioning, making action quality comparable to or better than direct-policy WAM and Joint-WAM, which require tens of steps. It is about 3 times faster than the direct-policy WAM baseline, with compiled inference latency as low as 40ms on an NVIDIA RTX 5090 GPU.

  8. arXiv cs.RO34

    PARS: Post-Quantum IoT Blockchain Security for Mission-Critical Physical AI Agents

    Addressing the threat quantum computers pose to public-key authentication and key agreement, the study proposes the "rotational secrecy" architectural objective and the PARS reference mechanism, which coordinates policy-triggered key establishment, rotation, and reconciliation among authorized Physical AI agents. The optional PARS-ESL permissioned chain, maintained by authorized validators or edge gateways, records only authentication state such as key epochs, group membership, revocation, and algorithm-suite policies, not high-frequency business transactions, allowing resource-constrained agents to join as lightweight clients.

  9. arXiv cs.RO58

    Patchwork: A Resource-Sharing System for Domain-Randomized Embodied Simulation

    The authors propose Patchwork, a resource-sharing system for domain-randomized simulation built on the E=T+Δ representation, which uses shared templates plus patches to record environment changes, implemented on robosuite and MuJoCo. On LIBERO-plus and RoboCasa365 workloads, at 32 replicas, shared-template execution reduces CPU RSS by 88-92% compared with process-per-environment execution. Experiments show that memory growth depends more on template diversity and unique asset content than on the number of instances.

  10. arXiv cs.RO62

    Specola: An STL Specification-Driven Robot Simulation Evaluation Tool

    Specola is a tool for evaluating robot behavior in simulation against Signal Temporal Logic (STL) specifications. It provides a Python API and command-line interface that can be integrated into existing policy execution loops. It offers an adapter interface and already provides adapters for nine simulation environments and benchmarks, including LIBERO, RoboCasa, Isaac Lab, and RoboDojo, to supplement success rate with information such as degree of satisfaction, parameter sensitivity, and violated requirements.