Deploying a language-guided navigation policy in a new building usually means watching it fail. Public-benchmark VLAs degrade scene by scene as layout, appearance, and language drift, and the standard fix of teleoperating fresh trajectories on site does not scale. We replace human demonstration with a single brief iPhone scan. Scan2Instruction compiles that scan into a complete training environment, automatically synthesizing executable navigation episodes with grounded instructions. PRPT-KS then adapts the policy through a closed-form Pareto utility read directly off the reconstructed scene, with no learned critic and no human labels. Across five real environments, the two designs together drive our RealNav to an 85% real-robot success rate, compared with 49% for a RealNav baseline post-trained on human demonstrations, and still achieve 82% in unscanned regions of the same buildings, compared with 41% for the same baseline. Our work turns demonstration-free scene adaptation for language-guided navigation from an aspiration into a working reality.
https://arxiv.org/abs/2610.33794
We present a principled, analytical framework for segmenting multiple independently moving rigid objects during single camera motion, built directly upon the OWL perception function. OWL encodes two fundamental visual motion cues, perceived looming $L$ and perceived rotation $\omega$, into a single closed-form representation of relative 3D structure. A key consequence of this structure is that the unit direction of relative translation $\hat{t}$ is analytically recoverable per image point from $L$ and $\omega$ alone. All points belonging to a non-rotating rigid body share the same $\hat{t}$, whereas points on different rigid bodies generally yield distinct values. This furnishes a natural segmentation criterion: clustering per-point $\hat{t}$ estimates partitions the visible scene into independently moving rigid bodies. Building on the 2D complex form of OWL, we derive the framework in 3D using quaternions, for a single moving camera. The approach is parallel, camera rotation-invariant, and requires no training data, no depth information, and no semantic priors.
https://arxiv.org/abs/2610.33789
Vision-Language Models (VLMs) have demonstrated impressive performance when integrated into robot manipulation pipelines, but real-world evaluation of VLM capabilities on downstream robotics tasks is both difficult and expensive to scale. We propose that spatial and temporal ordering tasks using images from stereo and wrist cameras attached to real-world robots are particularly well-suited for evaluating visual reasoning capabilities, as they assess skills directly applicable to robots. We evaluate VLM reasoning abilities on three questions: ordering shuffled frames conditioned on a language description of the task, ordering objects depthwise based on distance from the camera, and ordering objects from left-to-right in pixel coordinates. The complexity of these tasks can also be modulated by changing the size of the set to sort. Our experiments show that several state-of-the-art vision-language models struggle on ordering tasks. To fix this, we propose a scalable autonomous training data generation pipeline, and show that both supervised finetuning and reinforcement-learning on the generated data can significantly improve VLM reasoning performance both on spatiotemporal ordering tasks as well as downstream robotics and visual reasoning tasks.
https://arxiv.org/abs/2610.33778
Simulation is widely used in autonomous-driving validation, but matching task outputs between real and synthetic images does not guarantee that a system relies on the same visual evidence. This creates a validation gap for transfer, especially in safety-critical perception functions. We present Concept-Decisive Feature Fidelity (C-DFF), a concept-level mechanism-parity diagnostic for paired real-synthetic evaluation. Starting from decisive evidence extracted from a frozen system-under-test (SUT), C-DFF projects that evidence into a shared semantic concept space using off-the-shelf segmentation and compares real and synthetic concept-importance vectors. The result is not only a scalar mechanism-parity score, but also a concept-shift diagnostic that identifies which scene concepts become more or less decisive in simulation. We evaluate C-DFF on KITTI $\leftrightarrow$ Virtual KITTI 2 with 2126 matched pairs across three frozen SUTs (steering regression, YOLOP drivable-area, YOLOP lane-line). Across these tasks, C-DFF is SUT-discriminative and yields interpretable concept-level summaries that support validation triage and simulator debugging. When reported alongside task-specific output-value context metrics, C-DFF shows that output agreement does not by itself establish mechanism parity. These results position C-DFF as a practical semantic diagnostic layer for validation and verification workflows in simulation-based AV testing.
https://arxiv.org/abs/2610.33760
World-action models (WAM) predict future states to guide robot actions, enabling learning from both action-free video and action-labeled interaction. Most inherit pretrained video generators, leaving WAM pretraining and scaling underexplored. We introduce Genie Envisioner Act 2.0 (GE-Act 2.0), a world-action model whose trainable generative and action components are all initialized from scratch on manipulation data. It combines a control-oriented autoencoder (CoAE), a single-step visual planner (SVP), and an inverse dynamics model (IDM). CoAE retains action- and instruction-relevant information under aggressive compression, while SVP produces a complete future state in one differentiable pass, so visual planning and inverse dynamics can be pretrained separately on complementary data. The components are then jointly trained with knowledge-aligned selective optimization (KASO), which reduces mismatched supervision by selecting only predicted futures judged behaviorally compatible with the recorded action. We evaluate pretrained checkpoints directly, without per-task fine-tuning, on 100 tasks across 20 manipulation skill groups with held-out scenes, backgrounds, lighting, and object instances. Scaling co-training data from 300 to 30,000 hours raises success from 17.1% to 44.1% on G1-OP and from 13.4% to 31.1% on G2-90D; despite comprising less than 2% of the co-training data, G2-90D improves by 17.7 points, suggesting cross-embodiment transfer. Gains span 19/20 and 18/20 skill groups, and skill-specific coverage strongly correlates with zero-shot out-of-distribution (OOD) success (Pearson r=0.80; Spearman rho=0.85). Under the same protocol, the model grounds object, color, shape, and position references in at least 90% of trials and follows explicit instructions even when they conflict with an already-committed behavior or a conventional scene association.
https://arxiv.org/abs/2609.05588
Embodied perception requires positional representations that capture both physical geometry and the spatial and temporal extent of observations. We introduce POD-RoPE (Physical Observation Distribution RoPE), which represents token observation footprints as distributions over camera centers, viewing directions, and sampling times. Under independent sampling, expected rotations encode relative relationships between distributions; when each distribution collapses to a single address, the construction recovers pointwise RoPE. The same rotation statistics support Q/K matching and coherence-dependent V/O transformations, allowing observation geometry to guide both information selection and content aggregation. Per-token pre-aggregation avoids increasing sequence length or processing pairs of observation nodes. We evaluate POD-RoPE across four robot policy architectures, simulation and real-world manipulation, and complementary reconstruction, depth-estimation, and novel-view-synthesis tasks. Integrating POD-RoPE into Fast-WAM improves mean success from 91.83% to 92.88% across all 50 RoboTwin 2.0 tasks, and from 32.5% to 52.5% on two real-robot tasks on Sharpa North. Controlled reconstruction ablations support the contributions of distribution encoding and V/O transformations. In H20 model-block benchmarks, POD-RoPE achieves median forward-plus-backward speedups of $1.91\times$ over P-RoPE and $5.34\times$ over RayRoPE, with a median latency overhead of 17.4% over Standard RoPE. These results support physical observation distributions as an effective and computationally efficient basis for geometry-aware attention.
https://arxiv.org/abs/2610.33745
Floor traction influences wheel-slip behavior, dead-reckoning accuracy, and the navigation performance of wheeled mobile robots that rely on wheel odometry. As these robots become increasingly common in GNSS-denied indoor environments, including automated manufacturing and retail facilities, their onboard measurements could also be used to characterize floor conditions and improve localization. In this study, a skid-steer robot was driven over five floor surfaces with distinct traction characteristics using commanded velocity profiles designed to induce wheel slip through repeated acceleration and direction reversal. Differences between the body velocity estimated from inertial sensing and the velocity obtained from wheel odometry were evaluated using the integrated relative-velocity error, phase-derived wheel-to-body response delay, and velocity-reversal response delay. Welch power spectral density analysis was additionally applied to characterize the frequency distribution of the velocity mismatch during positive and negative acceleration. Band-integrated spectral power, spectral centroid, and spectral decay rate were extracted from each recording and evaluated using linear discriminant analysis and multiclass classification. The response-delay measurements produced consistent surface-dependent trends, with the longest delays observed on soap and the shortest delays observed on carpet. The spectral features also provided clear separation among the tested surfaces, achieving a balanced classification accuracy of 97.1% under recording-level cross-validation. These results demonstrate that both the magnitude and frequency domain characteristics of wheel-to-body velocity mismatch contain useful information about tire-surface interactions. Such information could support surface recognition and the development of slip-aware wheel-odometry weighting within robot state-estimation frameworks.
https://arxiv.org/abs/2610.33735
Many robot manipulation videos do not indicate whether the task was successfully completed. RoboJev-27B is an open-weight vision-language model that takes a task instruction, 10 frames of a manipulation attempt, and (if available) final frames from auxiliary cameras and predicts whether the robot accomplished the task. RoboJev-27B is a variant of JEV-27B-VL, trained with a rank-32 LoRA adapter on the language model only (keeping the vision encoder frozen) using 52,598 episode records. It also uses a source-specific KL penalty that applies only to research data, anchoring predictions closer to the original model on those sources. It uses a calibrated two-token readout that outputs success/failure probabilities in a single forward pass. On 268 held-out episodes of the same robot model (SO-101), RoboJev-27B attains 95.2% balanced accuracy and 0.993 AUROC. It also generalises to episodes collected by unseen labs/robots (AUROC = 0.890 over 459 episodes) as well as to clearly labelled community-collected episodes (accuracy = 93.3% over 135 episodes), where the base model attains an AUROC = 0.900 on the unseen-lab/robot subset with the same inputs. It takes 0.26 seconds per episode (median, serial) to infer on a single H100, which is about \$0.19 per 1,000 success predictions at \$2.69 per GPU-hour (not including media processing and serving costs). RoboJev-27B is released together with its inference code, along with a transparent evaluation: 20,628 predictions organized into 29 model/protocol entries are made available to facilitate browsing and to enable reproducible benchmarking of success detection. We also compare against strong API-based frontier VLMs, which take in fewer input frames (4 images from the primary camera only); note that this comparison is not strictly like-for-like because RoboJev-27B is given more frames at inference.
https://arxiv.org/abs/2610.33718
The integration of quantum computing into robotics requires mathematical representations that preserve the geometric structure and configuration dependence of robot motion. This paper introduces a structure-preserving quantum representation of rigid-body transformations and serial-manipulator kinematics, establishing a correspondence between classical kinematic models and parameterized quantum circuits. Building on the Bloch-sphere representation, spatial directions are encoded into single-qubit states, while quantum rotations reproduce the ordered composition of reference-frame transformations. Translational contributions are combined through quantum-state aggregation, with their metric scales retained classically, enabling the reconstruction of end-effector position and orientation from Pauli expectation values. The formulation preserves the original joint variables and accommodates different kinematic conventions and joint types. Beyond forward kinematics, we establish that the manipulator geometric Jacobian can be reconstructed from quantum-observable derivatives using the parameter-shift rule, with analytical equivalence to its classical counterpart under ideal readout. To demonstrate the applicability of this representation to robot-level algorithms, we introduce Quantum Closed Loop Inverse Kinematics, a hybrid closed-loop method that exploits quantum-derived pose and Jacobian information. The proposed framework provides a geometrically interpretable foundation for developing and investigating quantum and hybrid computational methods for robotic kinematics and motion.
https://arxiv.org/abs/2610.33715
A robot in a home it has never seen must open doors, drawers, lids and taps whose physics a picture does not show. Instructions such as "open the drawer halfway, gently" depend on that physics: how far a part travels, how it resists, and what locks it. We present Open-World Law-to-Action (L2A). A vision-language model (VLM) proposes, for each movable part, its motions, the terms of its law of motion, and what may lock it. Each proposal becomes a prior with a floor, so that touch can always overturn it. The robot identifies each part's law online by Bayesian linear regression with closed-form evidence, for parts that move in one way or in two, and keeps a hypothesis that an unseen part or motion locks the target. We prove that touch alone cannot rule this hypothesis out, derive when another look is worth its cost, and show that a constant load and friction cannot be told apart while a part moves one way. All experiments are in simulation, developed on some scenes and run once on held-out ones. In new kitchens the robot completes 80% of tasks from the VLM's proposals, against 98% when every part is known. Proposed physics raises success on objects that need new terms from 353 to 378-383 of 400 episodes, and a second look raises success in mixed scenes from 284 to 329 of 366 tasks. A mobile robot that takes nothing from the simulator completes 70 of 100 requests in new scenes, against 62 when the simulator says which part each box is (after one change made between two held-out tests).
https://arxiv.org/abs/2610.33705
Vision-Language-Action (VLA) models are bottlenecked by the scarcity of robot data. Egocentric human manipulation videos offer an abundant, diverse, and semantically-rich alternative that inherently captures manipulation intent. However, the embodiment gap makes it difficult to extract action-relevant representations from these videos. From the perspective of action execution, human hands and robots share analogous manipulation primitives---reaching, grasping, placing---yet exhibit distinct kinematics tied to their embodiments. This might motivate our key approach: egocentric video pretraining can yield shared action representations capturing transferable manipulation skills, while human-specific motion patterns should be disentangled to avoid interfering with robot policy learning. To this end, we propose EXE, an Egocentric Cross(X)-Embodiment pretraining framework that learns a factorized latent action representation from egocentric human videos, separating shared manipulation priors from embodiment-specific signals. By leveraging the shared action space, the robot policy acquires generalizable manipulation semantics from diverse human demonstrations without being biased by human-specific kinematics. This shared knowledge transfers across embodiments with different kinematic structures, enabling the same pretraining to benefit both parallel grippers and dexterous hands. EXE yields consistent gains over baseline without pretraining on RoboTwin 2.0, RoboCasa GR1, and real-world manipulation tasks.
https://arxiv.org/abs/2610.33685
Fault diagnosis during regulated motion requires informative signals even when feedback attenuates tracking deviations. This paper develops an energy-based diagnostic framework that constructs mechanical port-power residuals without a dedicated dynamic observer. Spectral Hamiltonian Algebraic Power Evaluation (SHAPE) combines port-Hamiltonian variables and differential flatness to generate nominal power references from planned motion for fully actuated mechanical systems with collocated actuation. Subtracting these references from commanded effort multiplied by measured flow yields residuals whose final evaluation requires neither recursive observer updates nor measured-state dynamics calculations. Continuous wavelet transform (CWT) scalograms and a shared-encoder convolutional neural network are used to distinguish nominal operation, actuator loss of effectiveness, anomalous friction, and vibration from complete trajectories. Evaluation uses a simulated six degree of freedom Kinova manipulator under model predictive control and fixed plant-model mismatch, with single faults at a specified joint. Across five training seeds and 3000 held-out test runs, SHAPE-CWT achieves a mean accuracy of 84.69 percent, versus 81.67 percent for torque residuals and 97.17 percent for the generalized momentum observer under common CWT processing. The respective sample standard deviations are 0.70, 0.70, and 0.37 percentage points. These results demonstrate useful fault-type discrimination from algebraic power residuals, while severity-excitation profiles and offline velocity perturbations reveal limits to their diagnostic capability.
https://arxiv.org/abs/2610.33679
Continuous aerial tracking often requires multiple drones to maintain target coverage over long periods. During handoff, the receiving UAV may lack prior target observations and must actively acquire a usable view without disrupting the sender's ongoing coverage. This work introduces HARP, a handoff-aware recovery planner that registers bounded sender predictions in the receiver's local map, distinguishes field-of-view failure from physical occlusion, and generates failure-specific yaw, translation, and parallax viewpoints. Candidate motions are checked for receiver clearance, sender-fiducial observability, protected sender--target visibility, temporal validity, and deadline compliance. When a complete trajectory cannot be admitted, HARP may execute a sampled admissible nonzero prefix and replan from updated evidence. In simulation, HARP acquired the target in all 49 standard trials and in 15 of 16 stress trials. Field trials validated HARP's recovery geometry under field-of-view limitation and environmental occlusions.
https://arxiv.org/abs/2610.33640
Cooperative transport with heterogeneous legged robots requires coordinating different motion capabilities while supporting an unfixed shared payload. We present HA-DiPO (Heterogeneous-Agent Diffusion Policy Optimization), a hierarchical multi-agent reinforcement learning framework under centralized training with decentralized execution (CTDE). The framework decouples team-level coordination from embodiment-specific motion control: at the high level, a centralized critic uses team-value estimates to jointly train independent coordination policies, aligning robot-specific action responses with a shared transport objective; at the low level, embodiment-specific robust motion policies track the coordination commands. The high-level diffusion policies offer more flexible action distributions than single-Gaussian parameterizations, allowing potentially multimodal coordination responses to be represented. The coordination policies learn from scratch through reward-driven simulator interaction without demonstrations. At execution, each robot generates coordination commands using only local robot and payload observations, without partner-state inputs or inter-robot messages. Simulation comparisons show that HA-DiPO achieves higher final mean returns and velocity-tracking success rates, as well as lower velocity errors, than the evaluated baselines. Real-world experiments further demonstrate quadrupedal-wheeled-legged and human-robot cooperative transport.
https://arxiv.org/abs/2610.33584
Simulation provides a low-cost sandbox where embodied agents can attempt tasks, observe outcomes, and learn from experience. Running many diverse simulation environments concurrently can consume substantial CPU and GPU memory because independently constructed scenes duplicate resources that remain unchanged across instances. We present Patchwork, a resource-sharing system for domain-randomized simulation, built around the representation $E=T+\Delta$: a shared template contains common resources, and a patch records the mutations needed to construct a target environment. Patchwork separates randomization sampling from replay and selects reusable templates together with patches that preserve each target environment's intended configuration. Its layered runtime shares compiled model structure and immutable assets while keeping mutable parameters and simulation state private to each instance. We implement Patchwork on robosuite and MuJoCo and evaluate it on LIBERO-plus and RoboCasa365 workloads. At 32 replicas, shared-template execution reduces CPU RSS by 88-92% relative to process-per-environment execution. Controlled experiments show how sharing shifts memory growth toward template diversity and unique asset content rather than instance count, within the supported sharing boundaries.
https://arxiv.org/abs/2610.33582
Running a language model and a dense 3D map on one device means the two share one memory, and exceeding its limit is an out-of-memory failure. Map systems that keep a bounded working set bound it by a count of keyframes, but the memory one keyframe occupies varies widely with what it carries, so a count does not control what the working set occupies. The unit of admission is as poorly matched, since a radius around the pose, a fixed cell, or the graph neighbors of the last matched keyframe all move parts of the map that are not needed. We therefore propose two decisions. First, the memory the resident map uses is bounded directly, and the cost of every admission is known before any data is read, from per-keyframe resident sizes computed offline and carried with the map. Second, keyframes are admitted and released at a semantic spatial unit, the rooms and corridors of the building, so that the robot holds the zone it occupies together with the zone it is about to enter while the zones it has passed are released. Every admission measures the resident total, makes room, and only then reads, so the budget is never overshot. On two NVIDIA Jetson boards the budget holds in every run and the model gives up at most 0.8 percent of its throughput, whereas without it the operating system kills the model or the map. Under that budget, and over the environments evaluated, admission by semantic zone localizes 6-9 points more of the query frames than a geometric partition of matched zone count and about twice as much as the retrieval of RTAB-Map, at a lower ATE. Bounding memory lets the map and the model coexist, and moving the map in semantic zones makes the bounded map usable.
https://arxiv.org/abs/2610.33574
Robot policy evaluation informs deployment decisions and policy improvement, but physical testing and simulator preparation require substantial effort. World models offer a complementary route through closed-loop interaction in generated environments. Recent benchmarks have demonstrated their potential as policy evaluators through outcome-based evaluation (i.e., assessing task success or failure). However, such evaluation does not establish whether world models faithfully simulate how policies fail. Misrepresenting failures can lead to incorrect diagnoses of policy weaknesses and make policies appear ready for deployment even when small scene changes cause them to fail. We therefore ask: do world models faithfully simulate how and at which stages policies fail, and under what conditions they transition from success to failure? To answer this question, we introduce WorldFailBench with three evaluation tasks. Failure reason tests whether models reproduce the observable errors policies make. Failure stage assesses whether failures concentrate at the same stages of full-task execution. Failure boundary examines whether models reproduce the conditions under which policies transition from success to failure under controlled scene changes in simulation. The benchmark covers manipulation tasks in the LIBERO and RoboTwin simulation environments and real-world manipulation tasks performed by the G1 humanoid robot. Experiments across seven world models reveal substantial gaps in their understanding of policy failures, providing guidance for developing more reliable world models for policy evaluation.
https://arxiv.org/abs/2610.33565
Recent vision-and-language navigation (VLN) policies act on their visual history, and an incorrect action can take them off the instructed route. Paired interventions on four VLN policies show that clearing the history doubles pooled success at the onset of a deviation but lowers it on route, while later in a deviation, removing only the off-route frames helps much less. After a deviation, the retained history keeps the policy from turning where a turn is needed, which we call behavioral inertia. We propose MOON, a framework that equips a frozen VLN policy with two lightweight modules, adding less than 1% of its parameters. A deviation judge decides when to forget, raising an alarm that clears the history when it predicts that the agent is off route. A recovery policy, trained with GRPO from alarm states, learns how to move on, turning more of the successes found only by sampling into a single greedy rollout. Experiments on R2R-CE and RxR-CE show that MOON reaches the highest R2R-CE success rate among the compared methods and raises the success rate of its frozen base by 5.2 and 6.0 points, and an ablation with a random trigger indicates that this gain depends on when the history is cleared. Without retraining, MOON also improves zero-shot object-goal navigation on HM3D-OVON. Together, these results suggest that deviation recovery depends on knowing when to forget, not only how to act. Code is available at this https URL.
https://arxiv.org/abs/2610.33546
Humanoid motion trackers reproduce diverse behaviors, but sharing a policy across bodies requires accounting for differences in structure, morphology and dynamics. We present MimicAll, a reinforcement-learning framework for shared whole-body motion tracking across heterogeneous humanoids. A structured embodiment descriptor is encoded into global and joint-aligned conditions, which are combined with observation history and body-matched motion references to predict joint-target offsets. Physics-consistent embodiment randomization, body-calibrated retargeting and heterogeneous simulation support training across 4,096 variants from 16 humanoid families. We introduce a benchmark of 464 bodies across five distribution strata to evaluate motion generalization, within-family embodiment shifts and unseen-family transfer. Controlled ablations show that embodiment information improves motion completion with simple concatenation, while self-attention yields the lowest global position errors. With equal training transitions, shared training achieves 99.4% 30-second success on randomized G1 bodies in 100STYLE, versus 42.9% for a single-body specialist. With Flat-MLP encoding and concatenation, continued training on a motion subset excluding obstacle categories achieves 98.3% and 93.5% 30-second success on 100STYLE for sampled training bodies and out-of-distribution variants within seen families, respectively. On an eight-motion LAFAN1 subset excluding ground and get-up sequences across 464 bodies, mismatched body conditions reduce self-attention's 30-second success from 61.7% to 47.0%, demonstrating a closed-loop benefit from embodiment information.
https://arxiv.org/abs/2610.33508
Robots can use historical observations to answer spatiotemporal queries in familiar environments, but the presence of relevant evidence does not imply that existing tools support the operations a task requires. We introduce ReSight, a task-driven self-improving memory system for embodied agents that expands the capabilities for processing and accessing historical observations as task demands arise. ReSight retains keyframes and their spatiotemporal associations, separating observation evidence from extensible derived fields so that stored observations can be reprocessed for new needs. An execution agent and a coding agent share this memory through different interfaces: the former calls registered tools within a bounded online budget, while the latter uses a programmable workspace and a larger exploration budget to inspect evidence, write programs, and acquire models as needed. To realize capability expansion in ReSight, we develop Explore-Distill-Mount (EDM), which identifies capability gaps from execution and exploration records without task reference answers and converts exploration-derived solutions into reusable memory builders, query tools, and skills, extending memory representations and access capabilities as needed. Experiments on three public benchmarks demonstrate effective base memory access and capability expansion. With the execution model and online budget held fixed within each benchmark, EDM reduces mean position error on OC-NaVQA from 33.10 m to 24.72 m and improves high-level success on FindingDory from 38.3% to 63.8%. With the tools and skill frozen, success still improves by 28.3 percentage points on FindingDory episodes excluded from exploration and distillation. Experiments on real robot patrol memories further demonstrate reuse of these capabilities for subsequent queries and new memories, with fewer online tool calls on average.
https://arxiv.org/abs/2610.33487