Agility Robotics has announced Digit 5, the latest generation of its humanoid robot aimed at warehouse and manufacturing work. The company says the new model is designed to operate alongside human workers without the protective barriers or cages typically required for industrial robots. Further technical details and availability were not specified in the report.
A new arXiv paper proposes conflict-predictive variable horizons for distributed model predictive control used in multi-drone collision avoidance. Fixed horizons force a trade-off between computational cost and early reaction to nearby drones, while the proposed method adjusts the horizon based on predicted conflicts. The approach aims to keep control cheap while still responding in time to approaching neighbors.
A new arXiv paper proposes a metrological inspection framework for high-mix, low-volume manufacturing, where parts, specs, and work orders change frequently. The approach steers a vision-language-action model to measure parts and gates its outputs against deterministic evidence rather than relying on fixed sensing routines. It aims to remove the need for repeated task-specific programming in inspection automation.
Researchers released TOPO-Bench, an open-source evaluation framework for topological mapping aimed at standardizing how navigation systems are compared. It introduces metrics, datasets and protocols, including a way to quantify perceptual aliasing, a common failure mode in place recognition. The work addresses the lack of shared benchmarks that has made results across topological mapping systems hard to compare.
A new arXiv preprint proposes ReWeight, a technique for post-training vision-language-action models when in-domain robot demonstrations are scarce. The approach retrieves relevant egocentric human demonstrations and applies sample weighting to make better use of that human data, since collecting robot-specific data is expensive. The paper targets adapting VLA models to particular robots and tasks.
This paper investigates how extra, non-deployable information about a physical system's state influences how quickly and reliably a reinforcement learning agent learns effective control policies when trained on real hardware instead of simulation. The experiments use a cylinder on a tabletop water channel as the test setup. It is a revised preprint posted to arXiv's artificial intelligence and machine learning sections.
A new arXiv paper examines how multi-agent reinforcement learning can be applied to task assignment and navigation for robot fleets in industrial settings. The authors focus on the gap between simulation training and deployment on physical multi-robot systems, a step that remains difficult in practice. The work reports on transferring learned policies from simulated environments to real factory robots.
A new arXiv preprint examines how legal norms can be encoded so that robots using world-model-based planning can follow them. The work extends the isomorphism problem of matching legal source texts to their formal encodings and identifies two main challenges for normative control of robots. It is a conceptual and measurement-focused contribution rather than a system release.
A new arXiv paper proposes a transformer-based meta-controller for adaptive control of robotic manipulators that must cope with friction memory they cannot directly observe. Existing attention-based controllers fix the number of attention heads before training and rely on expensive offline tuning; this work instead adjusts capacity incrementally at runtime. The authors position the method as a reinforcement-learning approach to adaptive control in a cross-listed machine learning submission.
A new arXiv paper proposes using language guidance to learn material representations that transfer across different vision-based tactile sensors. Because each sensor produces distinct readings of the same material, the approach aims to close that gap by grounding learned features in shared language descriptions. The work targets robotic touch perception, where properties such as softness and texture are difficult to recover from vision alone.
The paper introduces a pretraining method that uses proprioceptive signals as an anchor to align different sensory modalities, aiming to help robots handle contact-rich assembly tasks. Such tasks demand submillimeter precision and accurate force interpretation during sustained contact, which makes sim-to-real transfer difficult. The approach targets zero-shot deployment of policies trained in simulation onto real hardware. This is a replacement listing on arXiv cs.AI.
A new arXiv paper proposes TIDAL, a control scheme that alternates between diffusion-based planning and action execution to keep vision-language-action models running at high frequency. The authors argue that current VLA systems rely on a low-frequency batch-and-execute approach, and that the resulting mismatch between model inference speed and robot control rate creates gaps in which the agent cannot react. TIDAL aims to close that blind spot while retaining the semantic generalization of large VLA models.
Researchers released vla-eval, an evaluation harness designed to simplify how vision-language-action models are tested across multiple simulation benchmarks. The tool addresses the friction of conflicting dependencies and inconsistent evaluation protocols that arise when benchmarks are combined in a single pipeline. It aims to make VLA evaluation more reproducible and easier to extend with new benchmarks.
A new arXiv preprint presents a method for designing passive counterweights for serial robotic manipulators based on the distribution of tasks the robot will actually perform, rather than optimizing for a single pose. The work combines counterweight synthesis with constrained co-design, aiming to better compensate gravity across realistic operating configurations. It falls within robotics and machine learning research rather than a commercial product release.
A new arXiv paper proposes using diffusion models to synthesize musculoskeletal gait trajectories tailored to individual patient parameters, a step toward better reference signals for wearable assistive devices. The authors frame this as a challenge for lower-limb exoskeletons and rehabilitation robotics, where control depends on trajectories matched to each user. The abstract is preliminary and reports no experimental results yet.
A new arXiv paper introduces IMPACT-VLA, a method for tracing how much each input modality — camera images, proprioceptive state, and language instructions — contributes to a vision-language-action policy's decisions at different points during task execution. The approach relies on counterfactual trajectories to isolate the effect of individual inputs, addressing the difficulty of interpreting these multimodal robot policies. The abstract excerpt does not detail experimental results or benchmarks.
Researchers introduce Bench2Dex, a benchmark for evaluating bimanual dexterous manipulation that combines visual and tactile sensing. The work addresses the lack of standardized tactile hardware designs across dexterous hands, which vary in finger structure, contact surfaces, and sensor layout. It aims to make performance comparisons across different hand platforms more consistent.
A new arXiv paper argues that the action-generation backbones in Vision-Language-Action models are far larger than the task requires, since robot actions carry much less information than image pixels. The authors introduce a freeze-share-shrink strategy to cut parameters in diffusion and flow-matching policies while preserving performance. The work targets more efficient manipulation models for robotics.
Researchers describe an end-to-end pipeline for recognizing touch gestures on a plush companion robot, whose soft body and spread-out tactile sensors complicate reliable detection. The work covers the design of small machine-learning models along with embedded validation to confirm they run effectively on the device itself. The abstract is truncated, so full details of the methods and results are not yet available.
An arXiv paper proposes Agent as Policy (AGP), a method in which a general-purpose agent handles both planning and execution while directly operating a physical robot. The authors report that this requires no training tailored to a specific task or environment. The work sits in the cs.CL category and is a revised submission.
Researchers introduce SlipSense, a multimodal tactile framework for detecting object slip during robotic dexterous manipulation. The work emphasizes measuring detection latency precisely and testing how well the approach transfers across different hardware platforms, areas the authors say prior systems handle poorly. It is presented as an arXiv preprint in the cs.AI and cs.LG categories.
A new arXiv paper argues that overall accuracy scores hide which stage of affordance prediction vision-language models actually fail at. The authors break the task into locating the relevant object part and knowing what action to apply, then test models on each separately. Their results indicate the bottleneck lies in part grounding rather than action knowledge.
A new arXiv paper compares near-infrared and standard color imaging for daytime farm-machinery traversability using a site-disjoint evaluation that removes spatial data leakage. The authors find NIR does not consistently beat color once frames from the same locations are kept out of both training and test splits, and note that earlier benchmarks reporting an NIR edge used sequence-level splits. The work suggests sensor choices for agricultural perception should be revisited on properly separated data.
A new arXiv paper presents Zonal RL-RRT, a path-planning method that merges reinforcement learning with rapidly-exploring random trees. The approach divides the environment into zones to guide tree growth, aiming to cut planning time while keeping success rates and path costs reasonable. It targets navigation in cluttered, complex spaces where conventional planners struggle.
A new arXiv paper introduces ActSafeGuard, a differentiable approach for enforcing hard physical constraints in vision-language-action and world-action models used for robotic manipulation. The method aligns constraint enforcement with training so that generated actions remain feasible and safe. The authors argue that current models can produce actions that violate physical limits, making them unsafe for deployment.
A research team has developed a method that lets a humanoid robot acquire badminton skills resembling human play. The work addresses the difficulty of combining fast, explosive movement with precise racket control, which differs from ordinary walking or stationary manipulation tasks. The paper is posted on arXiv as a cross-listing replacement in the cs.AI and cs.LG categories.
A new arXiv preprint describes a method for combining individually trained controllers into composed behaviors for legged and humanoid robots. The authors note that specialized skills are fast to train and converge reliably because each addresses a narrow problem. The work is categorized under cross-listings on arXiv's cs.AI section.
A new arXiv preprint introduces a method for adjusting pretrained generative robot policies at deployment time using lexicographic preferences. This lets operators express prioritized requirements that were not captured during training, instead of retraining or fine-tuning the policy. The approach targets situations where deployed robots face new constraints or user preferences.
An updated arXiv paper presents a representation for spatial information called a stochastic map, along with methods for constructing it, extracting information from it, and revising it step by step as new data arrives. The work targets robotics settings where position estimates are uncertain and must be maintained incrementally rather than recomputed from scratch.
A new arXiv paper surveys how diffusion-based learning methods can be applied to robot learning, framing the problem through ergodic control and controlled diffusion. It is written as a combined review and tutorial, intended to give researchers a structured entry point to the underlying statistical machinery. The work is cross-listed in machine learning and focuses on deriving complex distributions from data for control tasks.
A new arXiv paper introduces FiberTune, a fine-tuning approach for vision-language-action (VLA) robot policies. The authors note that conventional action-supervised fine-tuning constrains only the directions that alter predicted actions, leaving other visual structure unregulated. FiberTune addresses this by maintaining visual residual structure that remains consistent across action-equivalent states.
A newly posted arXiv preprint examines whether large language model agents can autonomously carry out long-horizon physical tasks, which require continuous observation of the environment and consequential actions. The authors frame the question around self-adaptive physical AI, where agents are expected to operate with limited or no human oversight. The abstract indicates a research analysis or position piece rather than a released system.
A research paper describes three robotic art installations that explore adaptive behavior as an aesthetic theme. The works combine embodied machine learning with digital evolution to create an artificial ecosystem where open-ended novelty emerges through trial and error. Viewers are drawn into observing how the robots change and evolve over time.
Researchers introduce GzDRL, a reinforcement learning framework that runs Gazebo robotics simulations in a single process rather than relying on conventional middleware-based bridges. The authors say this design targets long-standing obstacles to reproducible and scalable robotics experimentation. The work is posted as an arXiv preprint (2609.13243).
A new arXiv paper introduces ShieldVLA, a method for safety alignment in vision-language-action models used in robotic manipulation and navigation. The authors argue that existing fine-tuning approaches, which largely depend on Lagrangian optimization, offer only limited safety guarantees. The work instead frames safety as a feasibility-aware alignment problem.
A new arXiv preprint introduces VGFM, a technique that guides flow-matching generative models with dense value signals to produce more expressive robot control policies. The approach targets robot learning from large offline datasets, where multimodal action representations are needed to capture varied behaviors. It aims to improve policy expressiveness within this offline learning paradigm.
A new survey paper catalogues methods that score robot behaviors, ranging from success detectors and reward models to runtime monitors. It examines how these verification approaches serve two roles: evaluating vision-language-action policies and providing training signals for them. The work appears on arXiv, cross-listed between the AI and machine learning categories.
A new arXiv paper addresses the tendency of open-source large language models to reason inconsistently and act inefficiently over long tasks when used inside agentic robotics stacks. The authors propose pairing a MetaTool component with the Robot Operating System so that LLM-driven agents can sustain coherent planning and execute actions more reliably. The work focuses on human-robot interaction settings where such instability has limited practical deployment.
A new arXiv preprint proposes a diffusion-based multiple-shooting method for indirect optimal control, aimed at generating fuel-efficient spacecraft trajectories. The authors argue that prior diffusion-based control work has largely neglected optimality guarantees, which their approach seeks to address. The work sits at the intersection of generative modeling, robotics-style control, and aerospace trajectory design.
A new arXiv paper examines how deploying the SmolVLA vision-language-action model in different runtime formats affects both inference speed and closed-loop task performance. The authors benchmark HuggingFaceVLA/smolvla_libero on a 6 GB RTX 2060 across the LIBERO Spatial and Object suites using MuJoCo and LeRobot with a fixed seed. The results indicate that cutting latency through optimized deployment can shift task behavior, so faster inference does not automatically mean better outcomes.