Papers

Object-Goal Navigation (ObjectNav) is a key capability for deploying mobile robots in everyday environments such as homes, schools, and workplaces. In this task, an agent must locate an instance of a target object category in previously unseen environments using only onboard perception, requiring the integration of semantic understanding, spatial reasoning, and long-horizon planning. Reinforcement learning (RL) has become a dominant paradigm for ObjectNav, yet modern systems involve numerous design choices across perception modules, policy architectures, and inference-time strategies. The relative impact of these components, however, remains poorly understood. In this work, we present a large-scale empirical study of modular RL-based ObjectNav systems. We decompose the navigation pipeline into three key components: perception, policy, and test-time enhancement, and conduct extensive controlled experiments to analyze their individual contributions. Our results suggest that improvements in perception quality and test-time strategies often yield larger performance gains than policy improvements alone, highlighting the importance of understanding how different components interact within modular navigation systems. Motivated by these findings, we introduce a unified framework for systematically studying modular ObjectNav systems. Guided by our analysis, we build an enhanced system that achieves state-of-the-art performance on the Gibson benchmark, improving SPL by 6.6% and success rate by 2.7% over prior methods. We also introduce a human expert baseline, achieving 98% success, highlighting the significant gap between current RL agents and human-level navigation. Finally, we provide practical insights and design recommendations for each module to help guide future research. Project page: https://honwang0054.github.io/What-matters-in-RL-ObjNav-web/.

Vision-Language Models (VLMs) are increasingly applied in autonomous driving for unified perception and reasoning, but high inference latency hinders real-time deployment. Early-exit reduces latency by terminating inference at intermediate layers, yet its task-dependent nature limits generalization across diverse scenarios. We observe that this limitation aligns with autonomous driving: navigation systems can anticipate upcoming contexts (e.g., intersections, traffic lights), indicating which tasks will be required. We propose Nav-EE, a navigation-guided early-exit framework that precomputes task-specific exit layers offline and dynamically applies them online based on navigation priors. Experiments on CODA, Waymo, and BOSCH show that Nav-EE achieves accuracy comparable to full inference while reducing latency by up to 63.9%. Real-vehicle integration with Autoware Universe further demonstrates reduced inference latency (600ms to 300ms), supporting faster decision-making in complex scenarios. These results suggest that coupling navigation foresight with early-exit offers a viable path toward efficient deployment of large models in autonomous systems. Code and data are available at our anonymous repository: https://anonymous.4open.science/r/Nav-EE-BBC4

Modern automated factories increasingly run manufacturing procedures using a matrix of programmable machines, such as 3D printers, interconnected by a programmable transport system, such as a fleet of tabletop robots. To embed a manufacturing procedure into a smart factory, an operator must: (a) assign each of its processes to a machine and (b) specify how agents should transport parts between machines. The problem of embedding a manufacturing process into a smart factory is termed the Smart Factory Embedding (SFE) problem. State-of-the-art SFE solvers can only scale to factories containing a couple dozen machines. Modern smart factories, however, may contain hundreds of machines. We fill this hole by introducing the first highly scalable solution to the SFE, TS-ACES, the Traffic System based Anytime Cyclic Embedding Solver. We show that TS-ACES is complete and can scale to SFE instances based on real industrial scenarios with more than a hundred machines.

Magnetic continuum robots (MCRs) enable minimally invasive navigation through tortuous anatomical channels, yet axially magnetized designs have largely been limited to bending-only motion. To expand deformation capabilities, this paper presents a simple assembly that embeds permanent magnets radially within the catheter wall, allowing a single externally steered permanent magnet to independently induce either bending or torsion. A physics-based formulation together with finite-element analysis establishes the actuation principles, and benchtop experiments validate decoupled mode control under practical fields. Building on this, a dual-layer blockage mechanism consisting of outer grooves and inner plates leverages torsional shear to achieve on-demand drug release. Finally, an in-phantom intervention experiment demonstrates end-to-end operation: lumen following by bending for target approach, followed by twist-activated release at the site. The resulting compact, cable-free platform combines versatile deformation with precise payload delivery, indicating strong potential for next-generation, site-specific therapies.

Vision-Language-Action (VLA) models have shown strong capabilities in robot manipulation by leveraging rich representations from pre-trained Vision-Language Models (VLMs). However, their representations arguably remain suboptimal, lacking sensitivity to robotic signals such as control actions and proprioceptive information. To address the issue, we introduce Robot State-aware Contrastive Loss (RS-CL), a simple and effective representation regularization for VLA models, designed to bridge the gap between VLM representations and robotic signals. In particular, RS-CL aligns the representations more closely with the robot's proprioceptive states by using relative distances between the states as soft supervision. Complementing the original action prediction objective, RS-CL enhances control-relevant representation learning, while being lightweight and fully compatible with standard VLA training pipelines. Our empirical results demonstrate that RS-CL substantially improves the performance of state-of-the-art VLA models; it pushes the prior art to 69.7% achieving the state-of-the-art performance on the RoboCasa-Kitchen benchmark, and boosts success rates from 45.0% to 58.3% on challenging real-robot manipulation tasks.

Humanoid whole-body control (WBC) policies trained in simulation often suffer from the sim-to-real gap, which fundamentally arises from simulator inductive bias, the inherent assumptions and limitations of any single simulator. These biases lead to nontrivial discrepancies both across simulators and between simulation and the real world. To mitigate the effect of simulator inductive bias, the key idea is to train policies jointly across multiple simulators, encouraging the learned controller to capture dynamics that generalize beyond any single simulator's assumptions. We thus introduce PolySim, a WBC training platform that integrates multiple heterogeneous simulators. PolySim can launch parallel environments from different engines simultaneously within a single training run, thereby realizing dynamics-level domain randomization. Theoretically, we show that PolySim yields a tighter upper bound on simulator inductive bias than single-simulator training. In experiments, PolySim substantially reduces motion-tracking error in sim-to-sim evaluations; for example, on MuJoCo, it improves execution success by 52.8 over an IsaacSim baseline. PolySim further enables zero-shot deployment on a real Unitree G1 without additional fine-tuning, showing effective transfer from simulation to the real world. We will release the PolySim code upon acceptance of this work.

This work presents a geometric backstepping controller for a variable-tilt omnidirectional multirotor that explicitly accounts for both servo and rotor dynamics. Considering actuator dynamics is essential for more effective and reliable operation, particularly during aggressive flight maneuvers or recovery from sudden disturbances. While prior studies have investigated actuator-aware control for conventional and fixed-tilt multirotors, these approaches rely on linear relationships between actuator input and wrench, which cannot capture the nonlinearities induced by variable tilt angles. In this work, we exploit the cascade structure between the rigid-body dynamics of the multirotor and its nonlinear actuator dynamics to design the proposed backstepping controller and establish exponential stability of the overall system. Furthermore, we reveal parametric uncertainty in the actuator model through experiments, and we demonstrate that the proposed controller remains robust against such uncertainty. The controller was compared against a baseline that does not account for actuator dynamics across three experimental scenarios: fast translational tracking, rapid rotational tracking, and recovery from sudden disturbance. The proposed method consistently achieved better tracking performance, and notably, while the baseline diverged and crashed during the fastest translational trajectory tracking and the recovery experiment, the proposed controller maintained stability and successfully completed the tasks, thereby demonstrating its effectiveness.

Non-rigid structure-from-motion (NRSfM), a promising technique for addressing the mapping challenges in monocular visual deformable simultaneous localization and mapping (SLAM), has attracted growing attention. We introduce a novel method, called Con-NRSfM, for NRSfM under conformal deformations, encompassing isometric deformations as a subset. Our approach performs point-wise reconstruction using 2D selected image warps optimized through a graph-based framework. Unlike existing methods that rely on strict assumptions, such as locally planar surfaces or locally linear deformations, and fail to recover the conformal scale, our method eliminates these constraints and accurately computes the local conformal scale. Additionally, our framework decouples constraints on depth and conformal scale, which are inseparable in other approaches, enabling more precise depth estimation. To address the sensitivity of the formulated problem, we employ a parallel separable iterative optimization strategy. Furthermore, a self-supervised learning framework, utilizing an encoder-decoder network, is incorporated to generate dense 3D point clouds with texture. Simulation and experimental results using both synthetic and real datasets demonstrate that our method surpasses existing approaches in terms of reconstruction accuracy and robustness. The code for the proposed method will be made publicly available on the project website: https://sites.google.com/view/con-nrsfm.

Multi-step manipulation in dynamic environments remains challenging. Imitation learning (IL) is reactive but lacks compositional generalization, since monolithic policies do not decide which skill to reuse when scenes change. Classical task-and-motion planning (TAMP) offers compositionality, but its high planning latency prevents real-time failure recovery. We introduce SymSkill, a unified framework that jointly learns predicates, operators, and skills from unlabeled, unsegmented demonstrations, combining compositional generalization with real-time recovery. Offline, SymSkill learns symbolic abstractions and goal-oriented skills directly from demonstrations. Online, given a conjunction of learned predicates, it uses a symbolic planner to compose and reorder skills to achieve symbolic goals while recovering from failures at both the motion and symbolic levels in real time. Coupled with a compliant controller, SymSkill supports safe execution under human and environmental disturbances. In RoboCasa simulation, SymSkill executes 12 single-step tasks with 85% success and composes them into multi-step plans without additional data. On a real Franka robot, it learns from 5 minutes of play data and performs 12-step tasks from goal specifications. Code and additional analysis are available at https://symskill.github.io/ .

A fundamental challenge in robust visual-inertial odometry (VIO) is to dynamically assess the reliability of sensor measurements. This assessment is crucial for properly weighting the contribution of each measurement to the state estimate. Conventional methods often simplify this by assuming a static, uniform uncertainty for all measurements. This heuristic, however, may be limited in its ability to capture the dynamic error characteristics inherent in real-world data. To improve this limitation, we present a statistical framework that learns measurement reliability assessment online, directly from sensor data and optimization results. Our approach leverages multi-view geometric consistency as a form of self-supervision. This enables the system to infer landmark uncertainty and adaptively weight visual measurements during optimization. We evaluated our method on the public EuRoC dataset, demonstrating improvements in tracking accuracy with average reductions of approximately 24% in translation error and 42% in rotation error compared to baseline methods with fixed uncertainty parameters. The resulting framework operates in real time while showing enhanced accuracy and robustness. To facilitate reproducibility and encourage further research, the source code will be made publicly available.

Recent advances in robotic manipulation have integrated low-level robotic control into Vision-Language Models (VLMs), extending them into Vision-Language-Action (VLA) models. Although state-of-the-art VLAs achieve strong performance in downstream robotic applications, supported by large-scale crowd-sourced robot training data, they still inevitably encounter failures during execution. Enabling robots to reason and recover from unpredictable and abrupt failures remains a critical challenge. Existing robotic manipulation datasets, collected in either simulation or the real world, primarily provide only ground-truth trajectories, leaving robots unable to recover once failures occur. Moreover, the few datasets that address failure detection typically offer only textual explanations, which are difficult to utilize directly in VLA models. To address this gap, we introduce FailSafe, a novel failure generation and recovery system that automatically produces diverse failure cases paired with executable recovery actions. FailSafe can be easily adapted to a wide range of manipulation tasks in simulators with motion planning support, enabling scalable creation of failure-action data. To demonstrate its effectiveness, we fine-tune LLaVA-OneVision-7B (LLaVA-OV-7B) to build FailSafe-VLM. Experimental results show that FailSafe-VLM successfully helps robotic arms detect and recover from potential failures, improving the performance of three state-of-the-art VLA models (Pi-0-FAST, OpenVLA, OpenVLA-OFT) by up to 22.6% on average across several tasks in ManiSkill. Furthermore, FailSafe-VLM could generalize across different spatial configurations, camera viewpoints, object and robotic embodiments.

Vision-Language-Action (VLA) models aim to unify perception, language understanding, and action generation, offering strong cross-task and cross-scene generalization with broad impact on embodied AI. However, current VLA models often lack explicit step-by-step reasoning, instead emitting final actions without considering affordance constraints or geometric relations. Their post-training pipelines also rarely reinforce reasoning quality, relying primarily on supervised fine-tuning with weak reward design. To address these challenges, we present VLA-R1, a reasoning-enhanced VLA that integrates Reinforcement Learning from Verifiable Rewards (RLVR) with Group Relative Policy Optimization (GRPO) to systematically optimize both reasoning and execution. Specifically, we design an RLVR-based post-training strategy with verifiable rewards for region alignment, trajectory consistency, and output formatting, thereby strengthening reasoning robustness and execution accuracy. Moreover, we develop VLA-CoT-13K, a high-quality dataset that provides chain-of-thought supervision explicitly aligned with affordance and trajectory annotations. Furthermore, extensive evaluations on in-domain, out-of-domain, simulation, and real-robot platforms demonstrate that VLA-R1 achieves superior generalization and real-world performance compared to prior VLA methods. We plan to release the model, code, and dataset following the publication of this work. Code: https://github.com/GigaAI-research/VLA-R1. Website: https://gigaai-research.github.io/VLA-R1.

We present ActiveUMI, a framework for a data collection system that transfers in-the-wild human demonstrations to robots capable of complex bimanual manipulation. ActiveUMI couples a portable VR teleoperation kit with sensorized controllers that mirror the robot's end-effectors, bridging human-robot kinematics via precise pose alignment. To ensure mobility and data quality, we introduce several key techniques, including immersive 3D model rendering, a self-contained wearable computer, and efficient calibration methods. ActiveUMI's defining feature is its capture of active, egocentric perception. By recording an operator's deliberate head movements via a head-mounted display, our system learns the crucial link between visual attention and manipulation. We evaluate ActiveUMI on six challenging bimanual tasks. Policies trained exclusively on ActiveUMI data achieve an average success rate of 70% on in-distribution tasks and demonstrate strong generalization, retaining a 56% success rate when tested on novel objects and in new environments. Our results demonstrate that portable data collection systems, when coupled with learned active perception, provide an effective and scalable pathway toward creating generalizable and highly capable real-world robot policies.

Bimanual robot manipulators can achieve impressive dexterity, but typically rely on two full six- or seven- degree-of-freedom arms so that paired grippers can coordinate effectively. This traditional framework increases system complexity while only exploiting a fraction of the overall workspace for dexterous interaction. We introduce the MiniBEE (Miniature Bimanual End-effector), a compact system in which two reduced-mobility arms (3+ DOF each) are coupled into a kinematic chain that preserves full relative positioning between grippers. To guide our design, we formulate a kinematic dexterity metric that enlarges the dexterous workspace while keeping the mechanism lightweight and wearable. The resulting system supports two complementary modes: (i) wearable kinesthetic data collection with self-tracked gripper poses, and (ii) deployment on a standard robot arm, extending dexterity across its entire workspace. We present kinematic analysis and design optimization methods for maximizing dexterous range, and demonstrate an end-to-end pipeline in which wearable demonstrations train imitation learning policies that perform robust, real-world bimanual manipulation.

This paper proposes a real-time multi-plane segmentation method based on GPU-accelerated high-resolution 3D voxel mapping for legged robot locomotion. Existing online planar mapping approaches struggle to balance accuracy and computational efficiency: direct depth image segmentation from specific sensors suffers from poor temporal integration, height map-based methods cannot represent complex 3D structures like overhangs, and voxel-based plane segmentation remains unexplored for real-time applications. To address these limitations, we develop a novel framework that integrates vertex-based connected component labeling with random sample consensus based plane detection and convex hull, leveraging GPU parallel computing to rapidly extract planar regions from point clouds accumulated in high-resolution 3D voxel maps. Experimental results demonstrate that the proposed method achieves fast and accurate 3D multi-plane segmentation at over 30 Hz update rate even at a resolution of 0.01 m, enabling the detected planes to be utilized in real time for locomotion tasks. Furthermore, we validate the effectiveness of our approach through experiments in both simulated environments and physical legged robot platforms, confirming robust locomotion performance when considering 3D planar structures.

Learning from human involvement aims to incorporate the human subject to monitor and correct agent behavior errors. Although most interactive imitation learning methods focus on correcting the agent's action at the current state, they do not adjust its actions in future states, which may be potentially more hazardous. To address this, we introduce Predictive Preference Learning from Human Interventions (PPL), which leverages the implicit preference signals contained in human interventions to inform predictions of future rollouts. The key idea of PPL is to bootstrap each human intervention into L future time steps, called the preference horizon, with the assumption that the agent follows the same action and the human makes the same intervention in the preference horizon. By applying preference optimization on these future states, expert corrections are propagated into the safety-critical regions where the agent is expected to explore, significantly improving learning efficiency and reducing human demonstrations needed. We evaluate our approach with experiments on both autonomous driving and robotic manipulation benchmarks and demonstrate its efficiency and generality. Our theoretical analysis further shows that selecting an appropriate preference horizon L balances coverage of risky states with label correctness, thereby bounding the algorithmic optimality gap. Demo and code are available at: https://metadriverse.github.io/ppl

Explicit information seeking is essential to human problem-solving in practical environments characterized by incomplete information and noisy dynamics. When the true environmental state is not directly observable, humans seek information to update their internal dynamics and inform future decision-making. Although existing Large Language Model (LLM) planning agents have addressed observational uncertainty, they often overlook discrepancies between their internal dynamics and the actual environment. We introduce Information Seeking Decision Planner (InfoSeeker), an LLM decision-making framework that integrates task-oriented planning with information seeking to align internal dynamics and make optimal decisions under uncertainty in both agent observations and environmental dynamics. InfoSeeker prompts an LLM to actively gather information by planning actions to validate its understanding, detect environmental changes, or test hypotheses before generating or revising task-oriented plans. To evaluate InfoSeeker, we introduce a novel benchmark suite featuring partially observable environments with incomplete observations and uncertain dynamics. Experiments demonstrate that InfoSeeker achieves a 74% absolute performance gain over prior methods without sacrificing sample efficiency. Moreover, InfoSeeker generalizes across LLMs and outperforms baselines on established benchmarks such as robotic manipulation and web navigation. These findings underscore the importance of tightly integrating planning and information seeking for robust behavior in partially observable environments. The project page is available at https://infoseekerllm.github.io