Nachrichten & Forschung

arXiv:2610.41540v1 Announce Type: cross Abstract: Imitation learning from observations must recover behavior from state-only demonstrations without known temporal correspondence. Differentiable simulators provide pathwise policy gradients, but differentiating through long, contact-rich trajectories can make optimization unstable. We introduce FOCUS, which combines local optimal-transport alignment with short-horizon differentiable actor-critic learning. FOCUS matches agent rollout windows to multiple sampled expert segments using a soft best-of-$K$ objective that emphasizes locally compatible matches. The resulting transport-based costs provide differentiable imitation rewards, while a learned critic bootstraps value beyond the rollout horizon. FOCUS requires neither temporally aligned trajectories nor expert-aligned initialization or teacher forcing. Across contact-rich locomotion tasks in DFlex, including a musculoskeletal humanoid controlled by 152 muscle activations, FOCUS improves

arXiv:2610.41408v1 Announce Type: cross Abstract: Vision-Language-Action (VLA) models enable capable robot behavior, but their large memory footprint limits deployment on resource-constrained robotic platforms. Weight quantization is a natural remedy, yet post-training quantization (PTQ) recipes inherited from large language models degrade when applied to VLAs at ultra-low precision. Existing low-bit PTQ methods often fail to explicitly account for the structure of downstream operations and allocate precision without considering its effect on the policy actions. We introduce SACQ, an ultra- low-bit weight-only vector-codebook quantization framework for VLAs. SACQ combines (1) structure-aware codebook fitting with role-specific quadratic metrics that reflect input statistics and downstream sensitivity; (2) action-aware mixed-bit allocation based on the increase in final-action error as individual layers are compressed; and (3) two-stage post-quantization refinement that reassigns codewo

arXiv:2610.41127v1 Announce Type: cross Abstract: Computer vision models learn hand action only from its visual consequences, without access to the underlying embodied motor signals. We address this absence with MotorSense, the first large-scale paired egocentric video and surface electromyography (EMG) dataset. MotorSense pairs bimanual EMG bands with egocentric video and yields synchronized video, bimanual EMG and IMU, and audio across 100 hours of unscripted manipulation. EMG offers a powerful scalable alternative to manual annotation: it captures finger movement, hand configuration, and the forces exerted between hand and environment as a dense signal free of occlusion and labeler bias, whereas manual annotation is costly, subjective, and unreliable under occlusion. We use MotorSense to learn sensorimotor alignment, producing an organization of action that is shared across the visual and motoric spaces, through cross-modal pre-training between video and EMG. The resultant vision mo

arXiv:2610.41044v1 Announce Type: cross Abstract: The use of large language models (LLMs) to coordinate multi-agent teams raises a fundamental architectural question: whether task allocation should be centralized in a single LLM controller or distributed across autonomous LLM agents, particularly when communication is unreliable. In a deliberately simplified perimeter-defense simulation against LLM-controlled intruders, we compared a central commander, an auction-based consensus team and independent agents under five network conditions. Breach rates were 15% for the consensus team, 32% for the commander and 29% for independent agents, and of the different network conditions, only limited radio range changed a team's breach rate, raising the commander's. In consensus team, giving a low-cost LLM more authority over the bid computation raised breaches by 38 percentage points. With birds and civilian drones in airspace, we added a photograph for each to the input text state. Only the near-

arXiv:2610.41021v1 Announce Type: cross Abstract: Quadratic programs (QPs) are fundamental building blocks of many robotic tasks such as inverse kinematics, whole-body control, and model predictive control. Extending these methods to large-scale robot learning and control requires solving many QPs concurrently and, for end-to-end learning, differentiating through their solutions. Efficient batched optimization is therefore essential to bringing established model-based methods into GPU-based robotics pipelines. We present \textsc{cuPIQP}, a batched and differentiable QP solver on GPUs based on a proximal interior-point method, with unified support for dense, sparse, and multistage problems. The solver exploits parallelism across independent QPs and their internal structure while preserving per-problem numerical updates and termination decisions. Benchmarks with up to 16384 QPs per batch demonstrate solve-time speedups of up to $18.8\times$ over \textsc{Moreau} and $4.26\times$ over eigh

arXiv:2610.40921v1 Announce Type: cross Abstract: Autonomous delivery robots increasingly share residential spaces with companion animals, yet navigation behavior is rarely designed with animal welfare in mind. This paper presents a pilot study evaluating a curve-out maneuver, a laterally arcing pass-by trajectory triggered at the dog's alert distance, proposed as a candidate welfare-informed navigation intervention grounded in canine social communication research and legible motion theory. Following a research through design approach, we developed and evaluated a first design iteration of the artifact in a naturalistic home environment with a companion dog. Contrary to our hypothesis, the curve-out produced a more acute stress response than the straight-line baseline. We argue that discontinuous execution limitation, the robot stopping and reorienting multiple times to approximate the arc, undermined legibility, which in itself became a stressor. The pilot surfaces the legibility of r

arXiv:2610.40835v1 Announce Type: cross Abstract: Teams that cannot communicate seem stuck with what each member sees. We show that their opponent can tell them the rest. Silent hunters, jammed robots and traders who never meet all face an adversary that reacts to the group as a whole; every member can read that reaction, a tell, and learn what no member knows alone. One tell can do the work of a message from every member. The adversary would rather hide, but hiding interferes with its own task, and a team that commits in advance to how it will respond decides how much hiding is worth. A committed team never loses when its opponent cares more about its own goals, and does as well as a fully informed team exactly when the opponent's stake outweighs any gain from lying. Below that point the team learns more by trusting less: in a solved Gaussian model it obtains the truth by promising not to use all of it and keeps the cost the opponent would have paid to conceal; an exact test decides w

arXiv:2610.42141v1 Announce Type: new Abstract: Recent work shows that a frozen robot policy performs far better inside an agent harness, where control code decides when to query the policy, consult vision tools, or command the arm directly. Yet today these harnesses are built by hand and redesigned for each new robot or benchmark. We show that harness design can be automated, offering a strong alternative to finetuning on a novel setup. In RoboArchitect, a coding model observes a robot's rollouts, diagnoses failures, and iteratively revises the harness around the frozen policy. The harness evolves on a set of episodes for its optimization and is evaluated on held-out scenes and tasks. Without updating any model weights, RoboArchitect raises average success from 9.7% to 43.3% on unseen scenes (4.5x) and from 12.7% to 40.2% on unseen tasks (3.2x), across multiple benchmarks and policies. The gains are largest where the benchmark is far from the policy's training data: wrapping a general

arXiv:2610.42097v1 Announce Type: new Abstract: A fallen humanoid beneath an overhead obstacle faces a coupled recovery problem. It must create clearance while organizing support for a stable rise, since attempting to stand before making room can further increase obstruction. We introduce a reinforcement learning method with contact-supported geometric escape guidance, which rewards incremental reduction of robot-obstacle overlap without prescribing an escape path. To achieve this, a learned motion prior, together with posture-dependent support and joint-stall costs, is proposed to shape a single policy that learns when to clear the obstacle, how to establish support, and when to complete the rise, all using proprioception alone. In simulation, it recovers in 88.5% of trials beneath plates that can only be lifted, compared with 18.9% without geometric guidance, and in 98.6% of trials beneath free plates with randomized size, mass and placement. On a Unitree G1, the deployed policy reco

arXiv:2610.42049v1 Announce Type: new Abstract: Autonomous wheelchairs can improve the quality of life for individuals with mobility impairments by providing greater independence. However, reliable navigation especially in human environments is a major challenge due to factors such as dynamic obstacles, unpredictable pedestrian behavior, and the users' need to complete multiple stops along the way before reaching their final destination. This paper presents a hybrid path planning framework for multi-goal navigation in dynamic environments. The framework integrates the Traveling Salesman Problem (TSP) for global task ordering with the Potential Field (PF) method and heuristic local planners such as Rapidly-exploring Random Tree (RRT) and A-star (A*) to achieve collision-free navigation. Within this hybrid framework, TSP determines the visiting order of intermediate stations, PF transfers global path information, and local planners avoid obstacles while guiding the robot to the stations.

arXiv:2610.42021v1 Announce Type: new Abstract: Large-scale egocentric human videos provide rich and scalable manipulation experience for robot policy learning, yet substantial embodiment gaps in visual appearance, kinematic structure, and action spaces hinder effective knowledge transfer. Existing methods often decouple cross-embodiment knowledge alignment from robot policy learning, making it difficult for human manipulation knowledge to directly and fully contribute to robot action learning. To address this issue, we propose Nexus-VLA, a hybrid policy learning framework that enables cross-embodiment knowledge transfer through end-to-end human-robot joint training, allowing human manipulation experience to directly contribute to robot policy learning. We first introduce a hierarchical cross-embodiment semantic alignment paradigm that progressively establishes shared manipulation semantics from task intent to fine-grained interactions, enabling more complete use of the manipulation kn

arXiv:2610.41982v1 Announce Type: new Abstract: Robotic Manipulation Surfaces (RMS) reposition objects by deforming the surface on which they rest. A recent class of RMS suspends a continuous fabric sheet from a sparse grid of vertical linear actuators lowering the actuator density of such surfaces but how to design such surfaces so that objects can be controlled everywhere on it, and how to extend it beyond a single module so that objects can pass between modules and achieve parallel, remain open. This paper presents a separable catenary model of the suspended sheet that is fast to compute and provides both the surface shape for given actuator heights and the actuator heights for a desired local slope. The closed form inverse lets a simple PID controller position objects on the sheet, and the model relates fabric length, module span and actuator stroke to the reachable workspace, yielding a minimum fabric length bound and design charts for these surfaces. The approach is then extended

arXiv:2610.41981v1 Announce Type: new Abstract: Data-driven controllers have enabled lower-limb robotic exoskeletons to adapt assistance to the user across many tasks. A leading example is biological torque control, which estimates the user's own joint moment from onboard sensors and assists in proportion to that estimate. However, its estimators are trained on data recorded with the target device, so deploying the controller on new hardware requires collecting new data. Existing open-source efforts release hardware, software, or controllers separately, so this data collection must still be repeated on the assembled device. Here, we present an open-source knee exoskeleton, built from off-the-shelf components and widely available fabrication tools, with a biological torque controller derived entirely from simulation. A muscle-driven imitation policy reproduces squatting motions from an open biomechanics dataset, and the resulting knee moments are distilled into an estimator that uses on

arXiv:2610.41892v1 Announce Type: new Abstract: How to reuse a pretrained locomotion policy for precise loco-manipulation is still an open problem. We propose an online method to learn the closed-loop dynamics under a frozen locomotion policy. The learned model predicts floating-base motion and improves loco-manipulation accuracy. However, unobserved latent conditions of the policy introduce heterogeneous responses, meaning that the same observed state will lead to different motions. Direct learning without accounting for this heterogeneity introduces bias and fluctuations into dynamics estimates, degrading floating-base prediction and end-effector tracking. To address this problem, a contaminated Gaussian model is proposed to approximate the effects of latent conditions. The model uses two Gaussian components to approximate predictable and unpredictable responses under the same observed state. Then, the contaminated Gaussian model is combined with a Gaussian process to construct an on

arXiv:2610.41877v1 Announce Type: new Abstract: Scaling humanoid motion foundation models is constrained by the scarcity of diverse, robot-executable data and by the difficulty of jointly learning from heterogeneous multimodal inputs while supporting real-time closed-loop control. We present Movens-5M, constructed from raw Internet videos through clip generation, quality filtering, semantic annotation, motion recovery, and physical refinement with humanoid retargeting and execution verification. It contains 15 million human-motion sequences and 5 million physically executable humanoid-motion sequences spanning over 10,000 hours, aligned with text, RGB video, speech, and music. Building on this data, we introduce Movens-Zero, a scalable understanding-generation-execution architecture that supports joint learning across modalities and real-time closed-loop execution. Task-mixed flow-matching training jointly optimizes a single action expert across seven generation tasks, enabling direct