The authors propose Reward-Guided RRT*, a sampling-based framework that directly constructs budget-satisfying task sequences in continuous state space through a series of locally joint task-selection and motion-planning stages, without requiring a precomputed task-transition graph.
The authors propose Labference, a benchmark for evaluating how robots adjust subsequent operations based on actual physical execution outcomes, separately measuring operation selection, physical execution, and task completion. Preliminary evaluation shows that the tested VLA policies still made target-selection errors even when object transfer succeeded, while GPT-driven control, though able to select correctly, encountered execution failures and higher computational cost.
The authors propose Thermal-Headroom Routing (THR), a lightweight model-selection policy that uses a closed-form RC thermal model to predict processor temperature and exclude detector configurations predicted to cross the throttling threshold.
RetroNav is a training-free lifelong navigation framework that integrates navigation, memory accumulation, and failure reflection, achieving state-of-the-art performance on five navigation benchmarks. It uses a unified memory system to link geometric and semantic scene knowledge with historical navigation experience, distinguishing objective observations from task-dependent interpretations. After each navigation attempt, an independent evaluator determines whether a retry is needed; when it is, a retrospection agent uses scene evidence to generate a retry brief. The authors say the method will be open-sourced.
The study compares cascade PID, hover LQR, BL-SMC, linear MPC with LQR fallback (LMPC-F), a preview-enabled LMPC-F, and a geometric SO(3) controller on a nonlinear plant with rotor gyroscopic coupling, motor lag, and saturation. It finds that a unified twelve-candidate tuning protocol changed the mean-RMSE ranking obtained under fixed baseline settings, moving BL-SMC from last to first; the new ranking recurred on 50 fresh realizations and remained unchanged under a 48-candidate budget.
The paper proposes that semantic failure analysis is a key step toward mature dexterous manipulation of deformable objects and supports this claim with proof-of-concept experiments. Unlike rigid-object manipulation, cables, plastic parts, cloth, and similar objects exhibit diverse and hard-to-detect failure modes such as buckling, plastic deformation, pressure changes, and tearing due to differences in materials, geometry, and behavior, requiring different corrective strategies. The authors define the components and challenges of semantic failure analysis and present experimental results indicating its potential as a research direction.
Zero-HLM proposes a real-to-sim-to-real framework that turns smartphone captures and simulated source demonstrations into scalable training data for humanoid loco-manipulation. The study generated 4,800 physically validated demonstrations and achieved zero-shot transfer on three Unitree G1 tasks; GR00T N1.7 trained in simulation reached a mean full-task success rate of 66.7%, versus 46.7% with real-data training.
PushAround proposes a physics-informed hybrid search method that combines geometric connectivity reasoning with short-horizon simulation to generate executable clearing plans for mobile robot teams. The method uses repeated W-clearance connectivity graph (WCCG) queries to identify frontier gaps that need widening and ranks them by estimated downstream cost, then generates direct or recursive pushing candidates, selecting obstacles, directions, contacts, and forces.
This study proposes using factor graphs to probabilistically model deformable linear objects (DLOs), estimating the magnitude, direction, and location of 3D forces acting on them while simultaneously reconstructing the cable's shape. Because cables are thin and unactuated, strain gauges and actuation readings are generally impractical; the method uses only one type of sensing information and performs force estimation in simulation tests that add measurement noise and have the cable move along a prescribed path. The authors state that this general method can apply to a broader range of tasks.
VISTA is a meta-agent framework that replaces post-hoc replay reflection with active self-exploration to improve the harness of embodied agents. It maintains a harness graph that records the lineage of harness versions, enabling the meta-agent to retrieve past hypotheses, reuse effective mechanisms, and avoid repeating failed directions. Each round starts from a retained version and its execution evidence, autonomously generates new rollouts, probes failure and recovery behavior, and on that basis iteratively revises the harness and re-evaluates it.
The study proposes a two-stage Region of Attraction (RoA) certification procedure: contraction-based stability analysis first determines a target set, which Taylor-model reachability analysis then verifies, thereby constructing a context-dependent safe RoA at runtime and enabling safe switching among high-assurance controller candidates.
Researchers propose JevVLA, a lightweight energy-based adaptation framework for vision-language-action (VLA) policies that, with the base policy frozen, learns a differentiable scalar energy function used both for comparing discrete candidates and for correcting or generating continuous actions.
LECO converts online trajectory optimization for dynamic target tracking into offline learning of a direct-control policy, using a differentiable kinematic model to let trajectory-level objectives directly guide control learning. Across three simulated obstacle densities, its root-mean-square linear jerk is 14.1%–31.2% lower than the best comparison method on this metric, it achieves the highest tracking success rate, and its per-cycle computation time is 4.25–5.63 ms.
This paper studies line-of-sight pointing control for a two-axis current-command gimbal under mismatched mechanical disturbances. It proposes a quantile disturbance predictor and a tube-inspired model predictive controller that reserves command headroom according to prediction uncertainty. On a two-axis embedded platform, the proposed configuration achieved a 2D pointing RMS error of 3.94 ± 0.10 mrad, with no command-saturation samples within the evaluation window; solve times and cycle overrun records are reported separately for a 100 Hz nominal implementation.
MMGNet proposes a multimodal grasp synthesis network that uses hierarchical image-to-3D feature injection and rotational graspness modeling to integrate multi-scale image features into 3D representations and jointly predict grasp approach direction and in-plane rotation. On GraspNet-1Billion with RealSense data, it achieves a mean AP of 60.34, surpassing all compared methods; in real-world experiments with adversarial grasp objects, its grasp success rate is 8.46 percentage points higher than AnyGrasp. Ablation studies show that the two modules provide complementary gains. Code, models, and data will be released after the paper is published.
The paper proposes an uncertainty-aware control barrier function (CBF) framework that approximates cascading covariance along multi-hop localization chains using Fisher information matrices, for UAV swarm search in GNSS-denied maritime environments. The framework incorporates the covariance model and its rate of change into constraint margins for connectivity maintenance and collision avoidance, and combines a hierarchical Voronoi target strategy with formation keeping to achieve forward coverage.
The paper proposes a disturbance-separated model predictive control (MPC) scheme for constrained pointing control of airborne electro-optical gimbals under carrier vibration, input delay, and actuator constraints. The scheme uses an adaptive periodic observer combined with linear-regression frequency estimation to predict multi-harmonic vibrations, and a linear observer plus a bounded LSTM network to estimate aperiodic disturbances and their changes over the input delay; periodic predictions are embedded in the MPC model, while aperiodic compensation is applied outside the optimizer.
The paper evaluates multiple types of human behavior models using common generative model metrics in two scenarios, high-speed driving and object handover, and compares them with the downstream performance of reinforcement learning policies trained using these models. The results show that existing behavioral evaluation metrics cannot reliably predict downstream policy performance, and that evaluation results are sensitive to the trajectory representation; even when the true test behavior distribution is known, training with a different distribution may yield better downstream performance.
The authors propose ISB-NMPC, which couples spherical-bearing prediction with nonlinear quadrotor dynamics to jointly optimize target-approach motion and target visibility. Across 30 paired simulation trials, the method achieved an interception success rate of 86.7%, the highest among the compared controllers, with a mean per-trial target-visible time of 99.11%. In outdoor flights, the interceptor acquired a 30 cm target at approximately 150 m, reached 17 m/s in long-range flight, and flew at approximately 14 m/s while pursuing a maneuvering target towed by a quadrotor. Both software and hardware are open source.
SceneAgent is an agentic pipeline that converts 3D Gaussian Splatting, photogrammetry, LiDAR captures, AI-generated 3DGS scenes, and 4D dynamic videos into interactive simulation environments through simulator-based visual inspection and repair. The pipeline covers object segmentation, physical parameter estimation, part decomposition, articulation, and deformable simulation.
This 35-page review divides networked humanoid robot intelligence into four functional categories based on information exchange: communication, coordination, collaboration, and social interaction with humans. It surveys the literature along six dimensions: interaction architecture, exchanged information, physical coupling, timescale, system scale, and embodiment.
VLA-HWBench breaks down inference into seven stages for 18 open-source VLA models on NVIDIA Jetson AGX Thor and AMD Ryzen AI Max+ PRO 395, finding that action-generation strategy determines latency more than model size does, with parameter count explaining at most 15% of latency variance.
A discrete-event simulation study compared two implemented intra-aisle robotic picking configurations: without an onboard tote lift, two level-captive shuttles serve the two tested levels; with the lift, one shuttle remains on its horizontal level and its onboard mechanism moves the order tote to the adjacent upper level to complete picks.
The authors propose OSMA-Bench, a benchmark that uses spatially matched scenario sets differing only in obstacle type to test whether robot navigation adjusts clearance and speed according to obstacle semantics, and they propose the Semantic Sensitivity Index (SSI) to measure this adaptation. Evaluating VLM and vision-language-action (VLA) policies on the benchmark, they find that even with explicit safety prompts, neither consistently adjusts its motion according to obstacle safety levels.
The authors propose Dream2Assist, a model-based reinforcement learning framework that adds intent inference to a recurrent state-space world model, enabling an assistive policy to anticipate and support discrete human goals without taking over human control. The work also presents a complete deployment pipeline from world-model policy to a safe human-machine system, addressing the sim-to-real-human gap through policy down-selection and reward restructuring, and validates it in two human-subjects studies, high-speed racing (N=30) and urban driving (N=10), plus a preliminary on-vehicle deployment, where intent conditioning produced measurable improvements in torque alignment and collision rate.
The authors propose RobotAgentBench (RAB), a benchmark for evaluating robot agents in realistic open-world household environments. Using an agent-assisted, human-in-the-loop real-to-sim pipeline, it constructs 50 household layouts and includes 90 atomic tasks, 18 short-horizon tasks, 8 long-horizon tasks, and 4 open-ended scenarios.
The study conducts controlled future-video prediction loss ablations in FastWAM and OpenWAM, raising success rates by 24.38 and 72.91 percentage points, respectively, under clean conditions on seen tasks. Trajectory clustering shows higher success rates within several coarse trajectory clusters, but cluster frequencies change little, indicating the gains do not come solely from changes in the distribution of trajectory patterns.
Researchers borrow the serial-number estimation method from the WWII "German tank problem," treating a robot's chemical sensor readings as observation samples and using a running maximum and sample count to unbiasedly estimate the peak concentration of an odor plume, thereby determining the odor source location. The method provides exact confidence bounds that let a robot or UAV decide whether to keep searching or land, serving as a termination criterion for olfactory navigation. The paper derives the estimator and its confidence bounds, demonstrates its sample efficiency, and examines its robustness when its assumptions fail.
To address the problem that short-term out-of-distribution residuals can corrupt the learned disturbance model during quadrotor trajectory tracking, DGC improves the online update step size while retaining the random Fourier feature model, the offline BPTT policy, and least-squares adaptation (DAPL). It distinguishes adaptation states by comparing long- and short-term gradient memories and uses a confidence map to suppress unreliable updates.
The study proposes a multimodal framework integrating speech, vision, and touch to generate contextually appropriate robot speech and movement. The team tested it in virtual reality with 18 participants across five social interaction scenarios and compared it against a human teleoperation baseline: the autonomous system scored above the neutral midpoint on both naturalness and social appropriateness, but the teleoperation baseline received significantly higher ratings on both measures. Behavioral coding and feedback showed that conversational timing, clarification of ambiguous requests, and spatial coordination were key factors behind the gap.
The authors propose PhysLadder, a hierarchical sim-real benchmark for evaluating physical interaction capabilities in contact-rich manipulation, comprising 16 household tasks divided into three cumulative levels: contact selection and geometric alignment (L1), force regulation under sustained contact (L2), and coordinated contact reconfiguration and force regulation (L3).
TRIAGE proposes a training-free visual token acceleration method that decides whether to reuse cached KV or prune based on a token drift threshold, and fixes the recomputation budget to directly control prefill latency.
ARPG proposes separating action representation learning from action generation: first pretrain an action representation encoder on large-scale human manipulation data, then freeze that encoder and train a 130M-parameter generative backbone in its representation space to learn task-conditioned action distributions.
AuRA is an automated robot interaction and data preparation system for digital-twin reconstruction, combining execution-aware grasp selection, gripper-motion measurements derived from robot states, and registered pre-interaction geometry and appearance priors. Across 13 deformable objects, it reduces active human labor and wall-clock collection time per successfully reconstructed asset by 88.5% and 55.4%, respectively; compared with the human-collected PhysTwin baseline, it improves 7 of 9 metrics for future prediction under known gripper motion, including a 28.1% reduction in one-way L1 Chamfer distance.
This technical report proposes an acoustic surveillance framework for localizing and tracking narrowband adversarial acoustic sources that threaten offshore data centers. To address aliased TDOA measurements and an ambiguous solution space under narrowband emissions and sparse sensors, it introduces a mobile receiver that temporarily forms anchor pairs with selected fixed receivers to constrain the feasible solution space. The two-stage pipeline first estimates the source's initial position using anchor constraints, then initializes a particle filter that jointly tracks source position and velocity. In Monte Carlo simulations, the localization success rate was 91%, and 9 of 10 open-water field trials achieved tracking errors below 2 meters.
The authors propose ThinkLocoManip, a whole-body control agent framework for legged loco-manipulation that uses an interleaved reasoning-and-acting paradigm to handle long-horizon tasks in large-scale environments. The framework orchestrates perception, navigation, manipulation, and locomotion as composable tools, and it designs cross-domain coordination skills and a keyframe thinking mechanism to support error recovery. In real-robot experiments, its success rate was 77%, more than double the 37.1% of the strongest plan-then-act baseline.
The research team proposes the RynnVLA-Latent framework, which uses RynnLAM to learn continuous, 3D-grounded latent actions from visual changes without native robot control signals, converting 30.1K hours of human, robot, and simulation video into 745.52M latent action descriptors for policy pretraining.
Why it matters: This research uses 3D-grounded latent actions to turn heterogeneous interaction videos into pretraining supervision, offering a reusable path for VLA models to bypass the scarcity of action annotations.
The authors propose ScheNav, a closed-loop navigation framework that turns rough sketch guidance into exploration and stopping decisions. Its core is SchemaField, a learning module that fuses guidance with accumulated observations and predicts spatial value fields. A supporting data pipeline converts 2D floorplans into navigation training environments, producing the Furnished-RPLAN dataset, which contains over 33k furnished layouts with paired navigation episodes and noise-controllable sketches.
AdaPolicy proposes an action-driven anticipatory policy framework that shapes future-aware representations through cascaded next-chunk prediction, requiring no encoding of future observations during training and adding zero inference-time overhead. The method achieves consistent gains on MetaWorld, RoboTwin 2.0, LIBERO, and real-world deployment, with improvements of up to 14 percentage points on millimeter-precision tasks such as headphone assembly and shoelace threading.
HuGe is a family of humanoid generalist vision-language-action (VLA) models pretrained on over 2,000 hours of egocentric human and robot data. It predicts latent motion tokens to drive a pretrained whole-body controller and outputs hand-control signals.