Peer Revenue Impact factor: volatile

Papers

Fresh from arXiv, typeset properly. Highlighted words trade on the market.

Computer Science Robotics

cs.RO Watch Robotics

Saturday 8 August

Robotics

Action- and Language-Conditioned Video Assessment for Embodied Control

Hwanhee Kim, Jaehyun Jang, Seungmin Cha et al. (+3 more)

Vision-based embodied agents executing multi-step natural language instructions require feedback mechanisms that assess task progress over complete trajectories. Conventional approaches based on final-frame matching or continuous embedding similarity may overlook intermediate transitions that are necessary for determining whether an instruction has been completed. We propose ALVA (Action- and Language-Conditioned Video Assessment), a trajectory evaluator that conditions its assessment on visual observations, the executed action sequence, and the natural language instruction. The method uses a pre-trained vision-language model (VLM) in two stages: it first summarizes frame-to-frame visual transitions conditioned on the executed actions and then assesses the generated summary with respect to the instruction to produce a discrete trajectory-level progress score. In simulated 3D household environments, ALVA exhibits a conservative assessment pattern with near-zero false-positive rates. When used as terminal feedback for closed-loop policy optimization, it provides more effective feedback than the evaluated static image and embedding-based visual baselines and reduces the performance gap to a ground-truth oracle. These results support action- and language-conditioned video assessment as an interpretable feedback mechanism for the evaluated simulated embodied-control tasks.
Read the full abstract

All authors Hwanhee Kim, Jaehyun Jang, Seungmin Cha, Hyeonseo Yun, Donghoon Lee, Chang D. Yoo

Robotics

Spatiotemporal Context-dependent Personalized Movement Compensation in Delayed Telemanipulation

Sai Jiang and Zonghe Chua

Communication delay remains a central challenge in telerobotics, where it disrupts visuomotor coordination and reduces task precision. Motion scaling is an effective countermeasure to delay-induced overshoot, yet typical deployments rely on uniform gains that neglect individual and contextual variability. We propose a human-centered method that fits personalized delay-, direction-, and distance-specific scaling parameters for each participant. We conducted experiments with twenty participants who performed delayed reaching tasks in a virtual simulator. Scaling gains were computed to minimize mean overshoot in simulation in each combination of experimental conditions. Evaluation was done in simulation and on a telesurgical robot to evaluate assistance benefits. Performance was assessed across multiple delays, distances, and movement directions using overshoot, endpoint error, trajectory smoothness, economy of motion, and a composite error-time metric. Motion scaling consistently improved performance relative to unassisted trials, yielding up to 20-25% performance gains in key metrics. Effects were most pronounced at longer delays. Personalization demonstrated additional accuracy benefits for inward reaching at a short distance under moderate delay. The results highlight the potential of personalized scaling as a foundation for more adaptive frameworks that integrate contextual information to improve the safety and precision of teleoperated procedures.
Read the full abstract

Robotics

Multi-modal Interactive Control of Robotic Arm based on Offline Large Language Models

Hanxiao Chen

Large Language Models (LLMs) have significantly revolutionized the modern society with numerous advanced interactions between humans and AI agents, whereas the usage of most large language models including ChatGPT are not friendly open-sourced and must require the users paying a lot for such AI services continuously. Therefore, deploying open-sourced large language models on local servers can be considered as an efficient approach to design and implement creative embodied AI algorithms with lower cost and more stable free usage. Inspired by this ordinary motivation, we originally propose and implement the "Socratic Models-ChatGLM", which is a well-performed algorithm for multi-modal interactive control of robotic arm based on offline large language models via the facile PyBullet platform, even presents extraordinary potential to address complicated text-image integrated multi-step long-horizon robotic manipulation tasks.
Read the full abstract

Robotics

Event-Time Hybrid Optimal Control for Robotic Table Tennis Serves

Thomas Gossard, Till Köpff and Andreas Ziegler

Robotic table tennis serves require high ball velocity and spin while respecting the robot's kinodynamic limits. Unlike rally strokes, a valid serve must also bounce on the server's side and clear the net, yielding a hybrid system with nonlinear flight and impact dynamics. We formulate spin-controlled serve generation as an event-time \ac{OCP} that optimizes the racket impact velocity and orientation together with the bounce, net-crossing, and landing times, enabling direct enforcement at phase boundaries of table-bounce and net-clearance constraints. The racket velocity and orientation are then converted into a complete kinodynamically feasible motion through a second \ac{OCP} enforcing joint-position, velocity, and torque limits. We evaluate the method numerically and on a KUKA Agilus robot. Compared with a fixed-step formulation with root localization, the proposed event-time formulation reduces the median solve time by a factor of 4.1 while maintaining comparable landing accuracy, spin accuracy, and serve validity. Real-robot experiments demonstrate controlled placement and topspin, backspin, and sidespin serves, with a mean landing error of $13.1 \pm 7.3$~cm and spin rates up to 30~rps. These results show that event-time optimal control efficiently generates physically valid, kinodynamically feasible serves while accounting for nonlinear aerodynamic and impact effects.
Read the full abstract

Robotics · Cross-listed from cs.AI

Explore, Map, Remember, Decide: Are Embodied VLMs Ready for Safety-Critical Scenarios?

Gabriele La Malfa, Nitay Alon, Emanuele La Malfa et al. (+2 more)

Theory of Space framework (ToS) assesses the spatial understanding of curiosity-driven Vision-Language Models (VLMs) under partial observability. As AI techniques are increasingly applied to safety-critical scenarios, it is crucial to understand whether VLMs possess robust spatial memory and make reliable decisions. In this paper, we assess whether VLMs' decisions are based on physical evidence or are corrupted by visual-language biases, if their memory processes align with human cognitive patterns, and how they respond to environmental hazards. We extend the ToS framework into a safety-critical, goal-driven pipeline, named Explore, Map, Remember, and Decide (EMRD). We then quantify Exploration Competence (Explore) through metrics of environmental coverage and temporal efficiency, assess Spatial Fidelity (Map), evaluate, with a suite of psychological metrics, Memory Persistence (Remember), and measure, using focal-point metrics, Cognitive Decision-Making (Decide). Our results show that in terms of decision-making capabilities, VLMs frequently select evacuation points based on pre-trained textual priors while lacking the spatial grounding to justify their choices. We also show that spatial reasoning degrades in low-light conditions, but it is not affected by texture and colour tampering. Our findings suggest that VLM memory fundamentally diverges from human cognition, creating unpredictable risks of misalignment.
Read the full abstract

All authors Gabriele La Malfa, Nitay Alon, Emanuele La Malfa, Reuth Mirsky, Stefan Sarkadi

Robotics

SurgWMBench: A Vision-Based Benchmark for World-Modeling Surgical Instrument Motion Planning

Huanrong Liu, Weiliang Huang, Bob Zhang et al. (+3 more)

Reliable surgical planning requires models that move beyond recognizing the current surgical step or imitating expert demonstrations, and instead anticipate how instrument motion reshapes subsequent operative states. Most surgical video understanding methods focus on recognizing phases, actions, or workflow states, while providing limited support for explicitly modeling instrument motion. Conversely, existing tool motion prediction methods can forecast instrument trajectories, but they generally do not capture the coupled evolution of future surgical video states. World models offer a natural framework for jointly modeling visual state transitions and instrument motion dynamics. However, existing surgical world model studies remain largely centered on visual generation quality, relying on generation-oriented metrics such as FVD and CD-FVD. These metrics are poorly aligned with instrument motion planning, as they do not directly measure whether predicted trajectories are geometrically accurate, temporally coherent, or actionable for downstream planning. This limitation is partly structural, since the field lacks public datasets and standardized evaluation protocols that provide the benchmarking infrastructure needed to assess motion-centric capabilities in surgical world models. In this paper, we introduce SurgWMBench, a vision-based benchmark for short-horizon surgical motion planning and dynamics prediction. Given intraoperative image sequences and historical instrument trajectory, SurgWMBench evaluates both near-future instrument motion prediction and stability under continuous rollout or input perturbations.
Read the full abstract

All authors Huanrong Liu, Weiliang Huang, Bob Zhang, Weichao Cai, Chunlin Tian, Qingbiao Li

Robotics

PhysX-CoT: Structured Physical Reasoning from a Single Image to Simulation-Ready 3D Assets

Jie Huang, Xiaohe Li, Jiahao Li et al. (+6 more)

Simulation-ready 3D assets are central to robotics and embodied AI. Generating them from a single image is usually framed as a vision-language model that emits a serialized asset for a decoder to turn into geometry and physical fields, leaving the image-to-3D reasoning implicit. We argue the limiting factor is this output-centric view: part placement and local shape are entangled in one global-coordinate token stream, and the intermediate physical states are never exposed for supervision, conditioning, or verification. PhysX-CoT instead casts single-image asset generation as an explicit structured physical reasoning process, an ordered and machine-parseable trajectory of part-level states covering decomposition, 2D and 3D grounding, relations, coarse geometry, and surface cues that we separately supervise, use to condition geometry, and treat as reward targets. Geometry is factorized so that 3D boxes carry placement and local codes carry shape, and CoT-aligned GRPO optimizes parse validity, grounding, geometry, placement, and physical consistency. Under a unified protocol that retrains all learned baselines on the same backbone, data, and frozen decoder, PhysX-CoT outperforms the closest full-task baseline across geometry, scale, and physical-attribute metrics. Oracle, token-matched, and state-order controls show the explicit states are functional rather than cosmetic, and in Unreal Engine~5 the generated assets parse, collide, and articulate at high validity.
Read the full abstract

All authors Jie Huang, Xiaohe Li, Jiahao Li, Fangli Mou, Chen Qian, Yuqiang Fang, Junhao Fan, Kaixin Zhang, Zide Fan

Robotics · Cross-listed from eess.SY

Parameter-Dependent LMI Synthesis for Semi-Global Differential ISS Trajectory Tracking of Nonholonomic Mobile Robots Under Multiplicative Wheel Slip

Mohammad Sabouri

This paper presents a parameter-dependent linear matrix inequality (LMI) framework for trajectory tracking of nonholonomic mobile robots subject to severe multiplicative wheel slip on variable-terrain surfaces. The sampled convex formulation, augmented with grid-to-continuum residual certification, simultaneously establishes semi-global differential input-to-state stability, a prescribed exponential decay rate, regional pole placement, and a gain-bounded feedback proxy for actuator-limited operation. A central contribution is an explicit upper bound on the additive disturbance induced by bounded multiplicative slip in the Kanayama error coordinates, bridging the physical slip mechanism and the convex synthesis paradigm. The auxiliary gain matrix and inverse storage metric are parameterized affinely in the reference velocities, while the storage metric inherits nonlinear dependence through pointwise matrix inversion. Stability is established via a cascade analysis combining variational contraction, forward invariance, slip-induced disturbance bounds, and dissipation-based trajectory reconstruction. Numerical validation compares three controllers across six reference trajectories, six disturbance classes, and a 60-second variable-terrain test featuring six severe slip patches with bidirectional slip ratios reaching +/-50%, replicated on two geometries. Supplementary studies address Gaussian sensor noise, compound stress-testing, and embedded-platform computational feasibility. Across 100 Monte-Carlo runs the proposed controller achieves complete trajectory containment within the certified envelope. On the variable-terrain scenario, peak tracking error is reduced by 12% against the fixed-gain LMI baseline and 49% against the manual baseline, with the constant-gain baseline infeasible at the prescribed decay rate.
Read the full abstract

Robotics

Compiling and Benchmarking Task-State Horizons for Embodied Agents

Meiqi Wang and Shichao Li

Frontier agentic models are increasingly deployed as high-level planners for long-horizon embodied tasks. Existing robotic benchmarks have advanced long-horizon evaluation, but primarily characterize difficulty through action-sequence length and subtask complexity, overlooking a distinct challenge: agents must track evolving task-relevant world states induced by both their exploration and environmental dynamics. We define the span of task-relevant state transitions that an agent must track as task-state horizon (TSH). To evaluate how agent performance varies with TSH, we introduce RoboGraph, a robotic task compiler that translates state-transition dependencies into executable symbolic graphs. Specifically, RoboGraph constructs task-state horizons from spatial and temporal causal dependencies, including those induced by unexpected failures and interventions during task execution. Building on RoboGraph, we release a benchmark comprising 588 episodes across 84 scenes with varying TSHs. Experiments evaluating 15 advanced agentic models in both semantic and visual closed-loop environments show that most models struggle with demanding TSHs, revealing substantial gaps in maintaining, exploring, and updating task-relevant state over long horizon.
Read the full abstract

Robotics

DA-NBV: A Direction-Aware Next-Best-View Planner for Efficient 3D Reconstruction of Ships at Sea

Jiaming Chen, Juntao Yang, Zhentao Zou et al. (+6 more)

Accurate 3D reconstruction of ships at sea is important for maritime supervision, damage assessment, and autonomous maritime operations. Although 3D reconstruction has advanced considerably, high-quality data acquisition still largely relies on manually designed trajectories or skilled operators, resulting in high costs and limited scalability. Next-best-view (NBV) planning automates this process by selecting subsequent viewpoints based on the current state. However, existing NBV policies mainly model spatial occupancy while overlooking directional observation history. This limitation is particularly problematic for ships: their complex superstructures and severe self-occlusions require observations from multiple viewpoints, and insufficient directional coverage often yields incomplete reconstructions. These challenges are further amplified at sea, where wave-induced heave, roll, and pitch continuously alter the ship's pose and surface visibility. Meanwhile, wind disturbances and limited onboard power impose stricter requirements on scanning efficiency. To address these challenges, we propose DA-NBV, a direction-aware NBV policy that augments the conventional occupancy state with directional observation statistics. We introduce a learnable Position Advantage Field (PAF) that uses directional information to guide viewpoint selection. The policy further adopts a locally constrained action space and a nonlinear coverage-shaping reward to improve scanning efficiency. We also develop the ship-oriented SeaShip-3D dataset and a configurable sea-state simulation environment. Experiments under varying heave, roll, and pitch conditions show that DA-NBV improves reconstruction completeness by approximately 3 percentage points and reduces Chamfer distance by 43% while achieving higher path efficiency.
Read the full abstract

All authors Jiaming Chen, Juntao Yang, Zhentao Zou, Qi Ming, Yi Yu, Zhihang Zhong, Xue Yang, Xue Jiang, Yue Zhou

Robotics

4D-WAM: Infusing Spatiotemporal Awareness into World Action Models through Trajectory Fields

Lishan Yang, Wenxuan Song, Xi Wang et al. (+14 more)

Building on recent advances in world models, World Action Models (WAMs) jointly model video prediction and action generation. However, they typically represent videos in 2D pixel space, creating a representation gap with 3D space in which robotic actions are executed. Recent 3D approaches introduce 3D information, but fail to fully exploit the dynamics of 3D structures. In this work, we propose 4D-WAM, a model-agnostic training strategy that injects spatiotemporal knowledge from 3D trajectory fields into WAMs through representation alignment. To this end, we introduce two complementary objectives: 1) motion alignment, which aligns temporal feature variations across adjacent frames and encourages the model to build local 4D awareness during training, and 2) destination alignment, which guides the model to infer the final destination from the source frame by minimizing the gap between their attention-like similarity distributions. Together, these objectives provide both local motion supervision and long-horizon goal guidance, enabling WAMs to learn trajectory-level spatiotemporal representations. Extensive in-distribution and out-of-distribution experiments across different base models demonstrate the model's improvements in spatial understanding, execution precision, robustness, generalization, and versatility.
Read the full abstract

All authors Lishan Yang, Wenxuan Song, Xi Wang, Pingyue Sheng, Zheng Fang, Ziyang Zhou, Junjie He, Haodong Yan, Jiayi Chen, Nan Sun, Qiao Sun, Pengwei Wang, Lingqiao Liu, Yan Wang, Yuxiang Gao, Feras Dayoub, Haoang Li

Robotics · Cross-listed from cs.AI

GraphThink: Graph-Enhanced LLM Thinking for Long-Horizon Embodied Task Planning

Chen Li, Sijie Cheng, Yuelin Zhang et al. (+4 more)

Embodied agents using LLM-based planners often struggle with physical hallucinations, poor generalization to long-horizon tasks, and lack of environmental awareness. We propose GraphThink, a novel framework that integrates a task graph to provide structured knowledge for robust planning and a scene graph to maintain environmental memory for event-driven replanning. Specifically, the task graph guides LLM thinking through contextual prompting and iterative refinement, effectively mitigating planning hallucinations. Furthermore, within the GRPO framework, the task graph offers delicate reward design to train the LLM planner, enhancing long-horizon planning capabilities and improving generalization. Finally, an event-driven replanning module, powered by the scene graph, enables closed-loop environment awareness and error correction. GraphThink achieves state-of-the-art performance on the ALFRED benchmark. In particular, our high-level planner surpasses leading API-based LLMs on both the validation set and held-out long-horizon tasks, underscoring its robust zero-shot and few-shot capabilities. Additional evaluations further demonstrate strong out-of-distribution generalization to novel tasks and environments.
Read the full abstract

All authors Chen Li, Sijie Cheng, Yuelin Zhang, Junxi Li, Maozhi Huang, Yang Liu, Wenbing Huang

Robotics

Auditing Instruction-Trajectory Mismatches in Multimodal Robot Demonstrations

Simon Holk, Ryosuke Takanami, Tatsuya Matsushima et al. (+4 more)

Robot demonstration datasets used to train vision-language-action policies can contain a subtle but harmful failure mode: trajectories that are behaviorally correct but paired with the wrong language instruction. We study post-hoc auditing of these Instruction-Trajectory Mismatches (ITMs). Unlike failed rollouts, ITMs often look plausible, and can corrupt the language-behavior mapping learned by the policy. We propose Multimodal Probabilistic Fusion (MMPF), a training-free auditing framework that treats each modality as an expert, estimates a task-label distribution from local neighborhood agreement and global prototype similarity, and then fuses modalities with predictive-entropy weighting in a product of experts. Across LIBERO benchmarks with injected instruction mismatches and noisy real-robot data, MMPF achieves the strongest overall ITM detection and label correction accuracy. We also show that auditing improves most downstream policy learning in settings where language is needed to disambiguate the task. We demonstrate in real robot experiments that our method can achieve improved policy performance and show the trade-off of filtering demonstrations compared to relabeling.
Read the full abstract

All authors Simon Holk, Ryosuke Takanami, Tatsuya Matsushima, Yusuke Iwasawa, Yutaka Matsuo, Yueh-Hua Wu, Kei Ota

Robotics

A Mixed-Stiffness Anthropomimetic Fingertip Broadens the Operating Range for Coin Grasping

Kaigen Go, Yinlai Jiang, Hiroshi Yokoi et al. (+1 more)

Robotic grasping of thin, flat objects such as coins on hard surfaces remains challenging because conventional methods require reorienting the object, accessing its underside, or adding a dedicated nail mechanism. We previously showed that a rigid nail arrests soft-pad deformation and thereby forms a geometric constraint that improves precision grasping. Here we asked whether an additional constraint-forming boundary, created within the pad by material choice rather than by anatomy, could extend the conditions under which that constraint holds. We fabricated anthropomimetic fingertips with Shore E10 silicone at the center and Shore A60 at the sides, and compared them with uniformly soft E10 fingertips. An automated apparatus performed an oblique rotational tip pinch in which the pad engaged the coin's lateral surface, lifting it from flush contact with no gap beneath it. Over variations in horizontal approach distances, vertical finger displacements, and index-finger rotation, the mixed-stiffness pair maintained high success rates across more tested settings than the uniform pair during both geometric-constraint formation and the transition to a stable grasp. The nail-free pair failed in all 36 conditions of Experiment 1-1. However, the uniform pair performed better when coin position along the finger axis was varied, a condition-dependent trade-off. After tuning for coin size, both fingertip types grasped all six Japanese denominations. These results suggest that the operating range for thin-object grasping depends not only on pad softness but also on where stiffness is placed within a nail-supported pad, making boundary placement a candidate fingertip design variable.
Read the full abstract

All authors Kaigen Go, Yinlai Jiang, Hiroshi Yokoi, Shunta Togo

Robotics · Cross-listed from cs.LG

V-Simba: Unleashing the Architectural Potential of RL in Visual Continuous Control

Donghu Kim, Youngdo Lee, Hojoon Lee et al. (+6 more)

Improving sample efficiency remains a core challenge in reinforcement learning (RL), especially in real-world settings like robotics, where data collection is costly. This challenge is pronounced in visual RL, where high-dimensional inputs often obscure learning signals. While prior work in visual RL has focused on algorithmic solutions, such as better dynamics models or exploration strategies, recent advances in state-based RL show that architectural design alone can lead to significant gains in sample efficiency. This raises an important question: Can these architectural principles transfer to visual RL? In response, we introduce V-Simba, a simple yet effective visual RL architecture inspired by the Simba architecture from state-based RL. Built on top of Soft Actor-Critic (SAC) with data augmentation, V-Simba modifies the architecture by adding normalization layers to stabilize training and using pointwise convolutions to reduce computation. Despite its simplicity, V-Simba matches or outperforms the state-of-the-art methods across the DMC, Adroit, and Meta-World benchmarks, while being more computationally efficient than DrQ-v2. We make our code publicly available at https://github.com/DAVIAN-Robotics/V-Simba.
Read the full abstract

All authors Donghu Kim, Youngdo Lee, Hojoon Lee, Johan Obando-Ceron, Byungkun Lee, Aaron Courville, Pablo Samuel Castro, Jaegul Choo, Clare Lyle

Friday 7 August

Robotics

StepTrigger: Contact-State-Triggered Backdoor Attacks on VLM-Powered Legged Robots

Jiageng Zhang, Doniyorkhon Obidov and Kaichen Yang

Large language models and vision-language models are increasingly used as high-level planners in robotic systems, using task goals and sensor summaries to select navigation or manipulation actions. This creates a new backdoor surface: a compromised planner can behave normally in most runs, yet change its target selection when a hidden trigger is present. Prior attacks on LLM-powered or embodied agents mainly rely on triggers that appear in language, camera-visible objects, scene semantics, or specific sequences of past actions. This paper presents StepTrigger, a contact-state-triggered backdoor attack for VLM-powered legged robots. The trigger is not a prompt token or a visible marker. It is produced by pressure and foot-ground contact patterns that arise when a Unitree Go1 quadruped walks across a dense terrain patch. Unlike conventional visual or textual triggers, contact signals are inherently noisy and may also arise during benign locomotion. To avoid treating every pressure anomaly as a trigger, StepTrigger learns a selective backdoor policy from multimodal robot state, using incidental pressure events as benign examples and dense-patch contacts as poisoned examples. In a stratified offline evaluation, the trained planner achieved 98.75% clean behavior preservation, 92.50% false-trigger rejection, 76.25% true-trigger activation, and 89.17% overall parsed behavior accuracy. These results reveal a backdoor surface in proprioceptive and contact channels that is not captured by defenses focused only on language, vision, or action history.
Read the full abstract

Robotics

Design and Implementation of an Ultra-Low-Cost Wall-Climbing Robot for Infrastructure Crack Detection

Mrinmoy Modak, Supreyo Chakravorty Pretom, Shourv Tarafder et al. (+1 more)

Crack detection is a crucial process to ensure the safety and longevity of buildings and other infrastructure. In this paper, we developed a low-cost, automated crack detection robot that leverages CNN, EfficientNet-B0, and YOLOv8 for efficient identification of cracks in concrete surfaces with a curated crack image dataset introduced to support training and evaluation. YOLOv8's real-time object detection enhances crack localization, while CNN and EfficientNet-B0 provide binary classification, ensuring high precision and recall. The system consists of two stages. In the first stage, YOLOv8 detects and localizes wall regions from the video frame, and the bounding boxes are cropped. The second stage performs crack detection using one of three models by analyzing the cropped regions. Cost-effective approaches are also taken for robot design. The robot features a fan-based negative pressure adhesion system, a 4-wheeled skid-steering drive, and an ESP32-CAM for real-time image capture. Its lightweight 3D-printed chassis ensures stability, allowing it to navigate both walls and ceilings while capturing images for crack analysis. Unlike conventional wall-climbing robot designs, this robot incorporates a funnel-shaped body that enhances negative pressure generation and achieves a 44% reduction in duty cycle, significantly lowering power consumption. By combining low-cost hardware with a deep learning pipeline, our system provides a scalable, efficient, and accessible solution for real-time infrastructure inspection at an approximate total cost of $25, with a lightweight web application enabling smartphone-based control. This affordability makes the system more suitable for the developing world, where infrastructure inspection is often limited by budget constraints, labor intensity, and safety risks.
Read the full abstract

All authors Mrinmoy Modak, Supreyo Chakravorty Pretom, Shourv Tarafder, Daniel S. Drew

Robotics

VLN on the Fly: An Onboard Vision-Language Navigation Stack for Aerial Robots

Marco S. Tayar, Felipe Tommaselli, Gianluca Capezutto et al. (+6 more)

Running vision-language navigation fully onboard an aerial robot is hard, since grounding, planning, and control must share limited compute and a single-stage error is difficult to isolate in flight. End-to-end aerial policies fuse these stages into one network, giving up the observability and safety checks a modular stack keeps available. We propose VLN on the Fly, an onboard stack that keeps grounding, planning, and control as separate, inspectable stages. A quantized VLM grounds an instruction to a coarse image cell, depth lifts it to a 3D goal, a fast B-spline planner returns a feasible trajectory, and a pretrained reinforcement learning policy tracks it to motor commands across quadrotors. Across 15 onboard flights over three everyday referents in a controlled indoor volume, the stack reaches the target in 13 of 15 trials with 5.72 cm mean goal error and 39.3% average GPU utilization. In 6 additional cluttered-environment trials, the stack tracks collision-free trajectories under onboard perception gating.
Read the full abstract

All authors Marco S. Tayar, Felipe Tommaselli, Gianluca Capezutto, Pedro Antonio Rabelo Saraiva, Pedro H. V. de Freitas, Lucas Kido, Guilherme Sonego, Ricardo V. Godoy, Marcelo Becker

Robotics

Selective Cross-View Consistency for World Action Models: Held-Out Viewpoint Robustness Without Test-Time Camera Information

Bingqi Huang, Bingchuan Wei, Yingkai Cai et al. (+1 more)

World action models (WAMs) jointly denoise future video frames and robot actions, and the video prior is expected to generalize their control. Camera viewpoint change remains one of their hardest perturbation axes. We study a question specific to this model class: when training with same-state cross-view image pairs, on which output coordinates should a consistency loss be imposed? The WAM denoising target mixes view-covariant coordinates, namely the predicted future scene, with view-invariant coordinates, namely the action chunk, future proprioception, and value. We show that consistency applied to the covariant block is provably harmful, shrinking legitimate view-specific content to a fraction $1/(1+4\lambda)$ of its true value, and we verify this shrinkage law in controlled experiments. Selective cross-view consistency (SCVC) therefore constrains only the invariant block, requires no camera labels, extrinsics, depth, or view synthesis at training or test time, and leaves the deployment interface unchanged. We introduce a carve-and-hold-out evaluation protocol on the LIBERO-Plus camera track that separates a distribution-matched ceiling from genuine interpolation and extrapolation to held-out viewpoints, with a matched pair-trained control isolating the effect of the consistency term from pair exposure. On held-out orbital viewpoints beyond the training envelope, SCVC improves closed-loop success over the matched control by 12.2 points (95% CI [7.4, 17.0]; +15.5, CI [11.7, 19.4], under an independent second seed) – an effect two further camera axes replicate – while interpolation within the envelope shows no gain in either seed (-1.2 and -4.3 points) and in-distribution competence is preserved (-0.6, -0.2). We also report a cross-backbone audit showing that published camera-robustness numbers are confounded by wrist-camera pose stability.
Read the full abstract

All authors Bingqi Huang, Bingchuan Wei, Yingkai Cai, Zhaokui Wang

Robotics

Drone-Assisted UAV-UGV Collaboration for Autonomous Navigation in Snow-Covered Terrain

Shreyam Gupta, P. Agrawal, Priyam Gupta et al. (+1 more)

This paper presents a collaborative UAV-UGV navigation framework for high-altitude, snow-covered terrain, where reduced visibility and unstable ground render conventional methods ineffective. We introduce a custom efficient U-Net architecture that falls under the computational constraints for real-time road segmentation, utilizing a novel synthetic snow data augmentation technique to achieve 96.5% segmentation accuracy. For UAV localization, we implement an Extended Kalman Filter (EKF) fusing onboard GPS and IMU data, achieving a maximum observed positional error of +-0.5 meters. The UGV position is determined via a visual tracking pipeline using YOLOv5 and depth data from the UAV's RGB-D camera. A dynamic path planning algorithm utilizes this segmentation to adjust for snow drifts, enabling successful navigation in obscured test environment with minimal deviation.
Read the full abstract

All authors Shreyam Gupta, P. Agrawal, Priyam Gupta, R. Gautam

Robotics

CoCoNav: Conformal Control for Safe Robot Navigation in Crowds

Cheng Guo, Mingzhe Ni, Zheng Liang et al. (+5 more)

Safe and efficient robot navigation in crowds requires anticipating pedestrian motion despite uncertain and potentially shifting prediction errors. Existing reactive methods can produce oscillatory behavior, while predictive planners often treat forecasts as exact or rely on restrictive error models. Incorporating conservative uncertainty sets as hard constraints can also render model predictive control (MPC) infeasible. We propose CoCoNav, a crowd-navigation framework that combines online conformal calibration with runtime-certified planning. A horizon-specific conformal proportional–integral controller adapts trajectory-error bounds to regulate long-run empirical coverage, enabling the framework to respond to changing prediction errors. A relax-then-verify planner preserves solver feasibility by generating nominal trajectories with soft-constrained MPC and separately certifying them, together with contingency maneuvers, against the calibrated bounds before execution. Simulations and quadruped experiments show that CoCoNav achieves a favorable balance among collision avoidance, task success, and navigation efficiency relative to the evaluated baselines.
Read the full abstract

All authors Cheng Guo, Mingzhe Ni, Zheng Liang, Yihu Ling, Yuan Hu, Michele Caprio, Daniele Pucci, Wei Pan

Robotics · Cross-listed from cs.LG

LUCID: Latent-Skill Unified Control via Imagined Dynamics for Long-Horizon Humanoid Loco-Manipulation

Cheng Guo, Mingzhe Ni, Angelo Cangelosi et al. (+1 more)

Long-horizon humanoid loco-manipulation requires composing versatile whole-body skills and reliable high-level decision making. Existing methods often coordinate pretrained skills with scripted planners, finite-state machines or task-specific model-free policies, restricting their ability to handle complex task sequences. To address this limitation, we propose LUCID, a hierarchical model-based reinforcement learning framework that plans over reusable skills through imagined rollouts of a learned dynamics model. LUCID first trains a structured latent-conditioned low-level policy via adversarial imitation and then freezes it while jointly learning a high-level policy and macro-dynamics world model. The world model predicts the temporally extended state transitions induced by latent decisions, enabling high-level policy optimization through imagined rollouts. We evaluate our framework across various simulated multi-object rearrangement scenarios. Experimental results show that LUCID improves the full-task success and partial-completion rates compared to prior baseline methods, demonstrating its effectiveness in complex sequential loco-manipulation tasks.
Read the full abstract

All authors Cheng Guo, Mingzhe Ni, Angelo Cangelosi, Arash Ajoudani

Robotics

Enhancing Autonomous Vehicle Navigation with a Clothoid-Based Lateral Controller

Aashish Shaju, Steve Southward and Mehdi Ahmadian

This study introduces an advanced lateral control strategy for autonomous vehicles using a clothoid-based approach integrated with an adaptive lookahead mechanism. The primary focus is on enhancing lateral stability and path-tracking accuracy through the application of Euler spirals for smooth curvature transitions, thereby reducing passenger discomfort and the risk of vehicle rollover. An innovative aspect of our work is the adaptive adjustment of lookahead distance based on real-time vehicle dynamics and road geometry, which ensures optimal path following under varying conditions. A quasi-feedback control algorithm constructs optimal clothoids at each time step, generating the appropriate steering input. A lead filter compensates for the vehicle's lateral dynamics lag, improving control responsiveness and stability. The effectiveness of the proposed controller is validated through a comprehensive co-simulation using TruckSim and Simulink, demonstrating significant improvements in lateral control performance across diverse driving scenarios. Future directions include scaling the controller for higher-speed applications and further optimization to minimize off-track errors, particularly for articulated vehicles.
Read the full abstract

Robotics

Complete, Scalable, and Robust Prioritized Planning for Multi-Robot Ordered Storage and Retrieval at Maximum Capacity

William Zhang, Tzvika Geft, Jingjin Yu et al. (+1 more)

Automated warehouses face a fundamental trade-off between maximizing storage density and achieving high retrieval throughput. While puzzle-based storage (PBS) architectures increase capacity by eliminating aisles, coordinating multiple robots in these high-density spaces is computationally challenging. This paper formalizes the challenge through a novel multi-robot problem formulation for ordered storage and retrieval: We consider rectangular 2D grids, where uniform-sized loads are first stored, up to full capacity, and subsequently retrieved according to prescribed arrival and departure sequences. The main contribution of this work is an online prioritized multi-agent path planning algorithm for this problem. The algorithm builds on prior work that constructs arrangements supporting sequential storage and retrieval, i.e., of one load at a time, without relocating loads. By exploiting the structural invariants of such arrangements, we achieve the scalability of decoupled planning while guaranteeing complete, deadlock-free parallel execution even at full storage density. Experiments demonstrate that the algorithm achieves near-linear improvement in makespan with respect to the number of robots, up to $C$ robots, where $C$ is the width of the grid's open side. Furthermore, the algorithm supports robust storage arrangements that accommodate bounded uncertainty in the departure sequence, with negligible impact on execution makespan.
Read the full abstract

All authors William Zhang, Tzvika Geft, Jingjin Yu, Kostas Bekris

Robotics

Hölder Signed Distance: A Differentiable, Signed, Parallelizable Metric for Robotics

Felipe Bartelt, Ali Umut Kaypak, Anthony Tzes et al. (+3 more)

Computing distances between sets is essential in robotic motion planning and control, where differentiable gradients enable real-time optimization. The Euclidean Signed Distance Function (SDF), however, is not differentiable everywhere, and existing alternatives often sacrifice differentiability, sign information, or computational efficiency. In this letter, we introduce a novel differentiable signed distance between convex polyhedra. To this end, we first propose differentiable versions of the minimum and maximum operators, termed the Hölder minimum and Hölder maximum. We then replace the original min-max operators in the classical SDF formulation, yielding the Hölder signed distance. Unlike prior differentiable distance formulations that rely on iterative algorithms, our approach is computed in closed form, eliminating convergence issues while remaining naturally amenable to GPU parallelization. We validate the practical advantages and computational performance of the proposed distance through runtime comparisons with existing approaches. We also present a robotic manipulator experiment, demonstrating its suitability for applications in control.
Read the full abstract

All authors Felipe Bartelt, Ali Umut Kaypak, Anthony Tzes, Farshad Khorrami, Luciano C. A. Pimenta, Vinicius M. Gonçalves

Robotics

A Reconfigurable Tracked Robot for Enhanced Obstacle Traversal Through Movable Articulation Point and Internal Mass Relocation

Yuki Uda, Yasutaka Nakashima, Motoji Yamamoto et al. (+1 more)

Tracked robots are widely used in unstructured environments; however, their obstacle traversal capability is fundamentally limited by a tradeoff between front-end reachability and locomotion stability. This study presents TRASER (Tracked Robot with Articulated Spine for Extended Reach), a reconfigurable tracked robot capable of relocating both its articulation point and internal mass. TRASER employs a tape-spring mechanism that localizes compliance to the bending region while maintaining high stiffness in the remaining body, thereby improving both front-end reachability and center-of-mass (CoM) shifting capability. Geometric and static models are developed to analyze the effects of articulation point and CoM position on step and ditch traversal performances. Experiments demonstrate step traversal, suspended-platform traversal, and ditch traversal of 74%, 66%, and 59% of the robot body length, respectively. To the best of our knowledge, these results represent the highest reported obstacle traversal capabilities among tracked mobile robots.
Read the full abstract

All authors Yuki Uda, Yasutaka Nakashima, Motoji Yamamoto, Ayato Kanada

Robotics · Cross-listed from cs.AI

CMU-Drive and V2V-VLA: Cooperative Multi-agent Unified Driving with Reasoning Benchmark and Vehicle-to-Vehicle Vision-Language-Action Models

Hsu-kuang Chiu and Stephen F. Smith

Vision-Language-Action (VLA) models have recently achieved impressive performance for end-to-end autonomous driving, yet existing approaches are primarily designed for an individual single autonomous driving agent with limited support for cooperative perception, reasoning, and planning. We present Cooperative Multi-agent Unified Driving with Reasoning (CMU-Drive), a closed-loop end-to-end benchmark for evaluating cooperative autonomous driving with multiple connected autonomous vehicles (CAVs) operating in safety-critical driving scenarios with background traffic participants. We further propose Vehicle-to-Vehicle Vision-Language-Action (V2V-VLA), a cooperative VLA model that integrates cooperative driving into a single forward pass by jointly generating driving actions, future waypoints, language reasoning, and communication policies. Experiments on CMU-Drive establish the first benchmark and baseline for cooperative VLA driving and provide a foundation for future research on multi-agent, closed-loop, end-to-end cooperative autonomous driving. Our code, benchmark, and model checkpoint will be publicly released to facilitate open-source research.
Read the full abstract

Robotics

GWM-VLA: Geometry-Aware Latent World Modeling for Vision-Language-Action Learning

Yanping Zhao, Hang Yu, Yiwei Wang et al. (+8 more)

Vision-Language-Action (VLA) models achieve strong robotic manipulation performance but often degrade under visual and environmental shifts. Latent world modeling offers a promising approach to improving robustness, yet existing methods commonly encode camera views independently and predict holistic scene dynamics without explicitly modeling their geometric relationships. We propose GWM-VLA, a geometry-aware latent world modeling framework for VLA learning. GWM-VLA combines geometry-aware multi-view state encoding, global context-conditioned target-view prediction, and shared latent-action representations grounded by robot-action supervision. Specifically, VGGT-$\Omega$ jointly aggregates multi-view observations at each timestep to construct geometry-aware multi-view states. The latent world model predicts the next-step patch tokens of a selected target view using patch and register tokens obtained after multi-view aggregation, thereby retaining multi-view geometric information without predicting the complete multi-view state. We use the wrist view as the target in our experiments, placing greater emphasis on end-effector motion and local gripper-object interactions. Finally, the shared latent-action representations condition both the latent world model and the flow-matching action head, allowing latent-prediction supervision and ground-truth robot-action supervision to jointly shape the same latent-action representations. Experiments across both simulation and real-world environments demonstrate the effectiveness and robustness of GWM-VLA.
Read the full abstract

All authors Yanping Zhao, Hang Yu, Yiwei Wang, Chen Ye, Siyu Tian, Di Zhang, Qingjun Wang, Qian Chen, Junqiao Zhao, Chen Ye, Guang Chen

Robotics

Depth-Wise Probing and Pruning of the Planning Token in a Driving Vision-Language-Action Model

Harisankar Babu, Benjamin Coors, Christopher Lang et al. (+3 more)

Vision-language-action (VLA) models route driving decisions through a deep language model, but it is unclear how much of that depth the action itself requires. We study a representative driving VLA whose entire plan is carried by a single planning token that a generative planner decodes into a trajectory. Borrowing the planner as a trajectory-space logit lens, we decode the planning token from every one of the 32 decoder layers and measure two signals: the linear decodability of the navigation command and trajectory compatibility with the frozen native planner. Our diagnostic shows that semantic intent is linearly decodable early: command-probe accuracy reaches 97.7% after the first decoder layer, compared with 16.7% chance. In contrast, compatibility with the frozen native planner improves gradually across depth, with open-loop Avg-L2 reaching its minimum of 2.11\,m only at the final layer. Learned readouts from the first layer recover much of this gap, indicating that planning information is already present early but is not yet represented in the format expected by the deployed planner. Ranking decoder layers by the angular deviation they induce in the planning token permits removal of 8 of 32 layers within an approximately 5% relative open-loop error increase and yields a measured 1.33$\times$ decoder speedup. At the evaluated sample size, no family-specific degradation is statistically resolved. These findings are limited to the evaluated ORION checkpoint and Bench2Drive setup.
Read the full abstract

All authors Harisankar Babu, Benjamin Coors, Christopher Lang, Hendrik Berkemeyer, Tamim Asfour, Simon Foell

Robotics

Learning Fault-Tolerant Locomotion with Adaptive Gait Timing

Giovanbattista Gravina, Luca Rossini, Carlo Rizzardo et al. (+2 more)

Hardware failures require legged robots to rapidly reorganize coordination and gait timing to maintain stability and mobility. This is particularly challenging for larger quadrupeds, where increased mass and tighter actuation limits reduce the feasibility of aggressive, high-frequency compensation strategies often observed on smaller platforms. In this work, we propose a deep reinforcement learning approach for fault-tolerant locomotion under actuator power loss. The method employs an asymmetric actor-critic architecture in which the critic has access to privileged information during training, while the actor learns to reconstruct a corresponding latent representation from proprioceptive observations. We introduce a latent-alignment loss that encourages consistency between actor and critic representations. Additionally, we augment the action space with a learnable gait frequency parameter, enabling adaptive gait timing in response to terrain variations and actuator degradation without predefined faulty-leg strategies. The approach is validated in high-fidelity simulation on uneven terrain and real-world experiments on flat ground using a 68 kg quadruped robot.
Read the full abstract

All authors Giovanbattista Gravina, Luca Rossini, Carlo Rizzardo, Arturo Laurenzi, Nikos Tsagarakis