RL for Robotics

RL: Reinforcement Learning

Follow this topic Ask about this topic

Momentum

31 papers in the last four weeks, up 675% on the four weeks before. 0.3% of all new papers.

Jul 13Week of Sep 28

Latest papers 155

May 12, 2026cs.LG

Behavioral Mode Discovery for Fine-tuning Multimodal Generative Policies

We address the problem of fine-tuning pre-trained generative policies with reinforcement learning (RL) while preserving the multimodality of their action distributions. Existing methods for RL fine-tuning of generative policies (e.g., diffusion policies) improve task performance but often collapse diverse behaviors into a single reward-maximizing mode. To mitigate this issue, we propose an unsupervised mode discovery framework that uncovers latent behavioral modes within generative policies. The discovered modes enable the use of mutual information as an intrinsic reward, regularizing RL fine-tuning to enhance task success while maintaining behavioral diversity. Experiments on robotic manipulation tasks demonstrate that our method consistently outperforms conventional fine-tuning approaches, achieving higher success rates and preserving richer multimodal action distributions.

Alberta Longhini, David Emukpere, Jean-Michel Renders +1

Department of Computer Science, Stanford University, CA, USA

RL for Robotics Reinforcement Learning RL for Generative Models RL Fine-Tuning

May 11, 2026cs.RO

UniSteer: Unified Noise Steering for Efficient Human-Guided VLA Adaptation

Diffusion-based vision-language-action (VLA) models have emerged as strong priors for robotic manipulation, yet adapting them to real-world distributions remains challenging. In particular, on-robot reinforcement learning (RL) is expensive and time-consuming, so effective adaptation depends on efficient policy improvement within a limited budget of real-world interactions. Noise-space RL lowers the cost by keeping the pretrained VLA fixed as a denoising generator while updating only a lightweight actor that predicts the noise. However, its performance is still limited due to inefficient autonomous exploration. Human corrective interventions can reduce this exploration burden, but they are naturally provided in action space, whereas noise-space finetuning requires supervision over noise variables. To address these challenges, we propose UniSteer, a Unified Noise Steering framework that combines human corrective guidance with noise-space RL through approximate action-to-noise inversion. Given a human corrective action, UniSteer inverts the frozen flow-matching decoder to recover a noise target, which provides supervised guidance for the same noise actor that is simultaneously optimized via reinforcement learning. Real-world experiments on diverse manipulation tasks show that UniSteer adapts more efficiently than strong noise-space RL and action-space human-in-the-loop baselines, improving the success rate from 20% to 90% in 66 minutes on average across four real-world adaptation tasks.

Junjie Lu, Xinyao Qin, Yuhua Jiang +6

University of Technology Sydney · Work conducted at Microsoft Research. · Tsinghua University +1

RL for Robotics Robot Policy Adaptation Reinforcement Learning VLM Adaptation Human-in-the-Loop AI Robotic Manipulation+1

May 11, 2026cs.RO

Plan in Sandbox, Navigate in Open Worlds: Learning Physics-Grounded Abstracted Experience for Embodied Navigation

Vision-Language Models (VLMs) have demonstrated exceptional general reasoning capabilities. However, their performance in embodied navigation remains hindered by a scarcity of aligned open-world vision and robot control data. Despite simulators providing a cost-effective alternative for data collection, the inherent reliance on photorealistic simulations often limits the transferability of learned policies. To this end, we propose \textit{\textbf{S}andbox-\textbf{A}bstracted \textbf{G}rounded \textbf{E}xperience} (\textbf{\textit{SAGE}}), a framework that enables agents to learn within a physics-grounded semantic abstraction rather than a photorealistic simulation, mimicking the human capacity for mental simulation where plans are rehearsed in simplified physics abstractions before execution. \textit{SAGE} system operates via three synergistic phases: (1) \textit{Genesis}: constructing diverse, physics-constrained semantic environments to bootstrap experience; (2) \textit{Evolution}: distilling experiences through Reinforcement Learning (RL), utilizing a novel asymmetric adaptive clipping mechanism to stabilize updates; (3) \textit{Navigation}: bridging the abstract policy to open-world control. We demonstrate that \textit{SAGE} significantly improves planner-assisted embodied navigation, achieving a 53.21% LLM-Match Success Rate on A-EQA (+9.7% over baseline), while showing encouraging transfer to physical indoor robot deployment.

Zhixuan Shen, Jiawei Du, Ziyu Guo +5

School of Computing and Artificial Intelligence, Southwest Jiaotong University, China · Centre for Frontier AI Research, A*STAR, Singapore · School of Computer Science, University of Leeds, UK

RL for Robotics Embodied Navigation Large Language Model-Based Robot Planning Robotics Simulation Sim-to-Real Transfer

May 10, 2026cs.RO

Minimizing Worst-Case Weighted Latency for Multi-Robot Persistent Monitoring: Theory and RL-Based Solutions

We study multi-robot persistent monitoring on weighted graphs, where node weights encode monitoring priorities and edge weights encode travel distances. The goal is to design joint robot trajectories that minimize the worst-case weighted latency across all nodes over an infinite time horizon. The widely adopted worst-case latency objective evaluates team performance over the entire time horizon and therefore may fail to distinguish strategies with poor transient behavior but strong asymptotic performance. To address this limitation, we propose a family of tail-performance objectives that generalize the standard objective and study the resulting functional optimization problems. We establish several key theoretical properties, including the existence of optimal strategies, relationships among the proposed objectives and their corresponding optimization problems, approximation by periodic solutions to arbitrary accuracy, and reductions to event-driven decision models with discretized waiting times. Building on these results, we construct an equivalent event-driven Markov decision process (MDP), called the Tail Worst-case Latency-Optimizing Markov Decision Process (TWLO-MDP), which reformulates the tail-performance objective as a standard average-reward criterion. We then develop reinforcement-learning-based solution methods for the TWLO-MDP and introduce the multi-robot monitoring benchmark (M2Bench), a unified platform that supports the evaluation and comparison of heuristic and learning-based monitoring algorithms. Experiments on synthetic and realistic monitoring scenarios show that our methods effectively reduce the worst-case weighted latency and outperform representative baselines.

Weizhen Wang, Ziheng Wang, Jianping He +2

School of Automation and Intelligent Sensing, Shanghai Jiao Tong University, Shanghai, China · Key Laboratory of System Control and Information Processing, Ministry of Education of China, Shanghai, China

RL for Robotics Markov Decision Processes Multi-Robot Systems Multi-Robot Coordination

May 6, 2026cs.RO

From Reach to Insert: Tactile-Augmented Precision Assembly under Sub-Millimeter Tolerances

High-precision assembly frequently involves tight-tolerance insertions, where even slight pose errors can cause jamming or excessive interaction forces, making robust and safe insertion policies difficult to obtain. This paper proposes a tactile-augmented two-stage method that combines Imitation Learning (IL) and Reinforcement Learning (RL) for precision insertion tasks. In the first stage, IL learns a reaching policy with position generalization that grasps the peg and brings it to the vicinity of the target region. In the second stage, RL executes the insertion and enables recovery from failures during contact-rich interactions. To better exploit tactile feedback, we introduce tactile group sampling to increase coverage of critical contact segments during training, and design a tactile critic to more accurately evaluate policy values, improving insertion performance while maintaining low contact forces. We conduct systematic experiments across five hole geometries and three clearance settings. Results show that our method substantially improves insertion performance across all settings; under the most challenging 0.05,mm clearance, it achieves a 67% success rate while keeping contact forces low, reducing the maximum interaction force by 60% and torque by 44%, thereby validating both effectiveness and safety for precision assembly.

Xinpan Meng, Siyao Huang, JingPu Yang +6

State Key Laboratory of Multimodal Artificial Intelligence Systems, Institute of Automation, Chinese Academy of Sciences, Beijing 100190, China · School of Artificial Intelligence, University of Chinese Academy of Sciences, Beijing 100049, China

RL for Robotics Contact-Rich Robotic Manipulation Robotic Insertion Tactile Sensing for Robotic Manipulation

May 4, 2026cs.LG

OGPO: Sample Efficient Full-Finetuning of Generative Control Policies

Generative control policies (GCPs), such as diffusion- and flow-based control policies, have emerged as effective parameterizations for robot learning. This work introduces Off-policy Generative Policy Optimization (OGPO), a sample-efficient algorithm for finetuning GCPs that maintains off-policy critic networks to maximize data reuse and propagate policy gradients through the full generative process of the policy via a modified PPO objective, using critics as the terminal reward. OGPO achieves state-of-the-art performance on manipulation tasks spanning multi-task settings, high-precision insertion, and dexterous control. To our knowledge, it is also the only method that can fine-tune poorly-initialized behavior cloning policies to near full task-success with no expert data in the online replay buffer, and does so with few task-specific hyperparameter tuning. Through extensive empirical investigations, we demonstrate that OGPO drastically outperforms methods alternatives on policy steering and learning residual corrections, and identify the key mechanisms behind its performance. We further introduce practical stabilization tricks, including success-buffer regularization, two-sided conservative advantages, and Q-variance reduction, to mitigate critic over-exploitation across state- and pixel-based settings. Beyond proposing OGPO, we conduct a systematic empirical study of GCP finetuning, identifying the stabilizing mechanisms and failure modes that govern successful off-policy full-policy improvement.

Sarvesh Patil, Mitsuhiko Nakamoto, Manan Agarwal +14

Carnegie Mellon University · University of California, Berkeley · University of Washington +1

RL for Robotics RL for Generative Models Residual Policy Learning RL Fine-Tuning Off-Policy RL

May 4, 2026cs.LG

Enhancing RL Generalizability in Robotics through SHAP Analysis of Algorithms and Hyperparameters

Despite significant advances in Reinforcement Learning (RL), model performance remains highly sensitive to algorithm and hyperparameter configurations, while generalization gaps across environments complicate real-world deployment. Although prior work has studied RL generalization, the relative contribution of specific configurations to the generalization gap has not been quantitatively decomposed and systematically leveraged for configuration selection. To address this limitation, we propose an explainable framework that evaluates RL performance across robotic environments using SHapley Additive exPlanations (SHAP) to quantify configuration impacts. We establish a theoretical foundation connecting Shapley values to generalizability, empirically analyze configuration impact patterns, and introduce SHAP-guided configuration selection to enhance generalization. Our results reveal distinct patterns across algorithms and hyperparameters, with consistent configuration impacts across diverse tasks and environments. By applying these insights to configuration selection, we achieve improved RL generalizability and provide actionable guidance for practitioners.

Lingxiao Kong, Cong Yang, Oya Deniz Beyan +1

Fraunhofer Institute for Applied Information Technology FIT, Germany · Soochow University, China · University of Cologne, Germany +1

RL for Robotics Reinforcement Learning Robot Policy Generalization Interpretability in RL

May 4, 2026cs.RO

Do We Really Need Immediate Resets? Rethinking Collision Handling for Efficient Robot Navigation

Should a single collision necessarily terminate an entire navigation episode? In most deep reinforcement learning (DRL) frameworks for robot navigation, this remains the standard practice: every collision immediately triggers a global environment reset and is penalized as a complete task failure. While a collision during deployment naturally indicates task failure, applying the same treatment during training prevents the agent from exploring challenging obstacle configurations, which slows learning progress in the early training phase. In this work, we challenge this convention and propose a Multi-Collision reset Budget (MCB) framework that decouples local collision termination from global environment resets, allowing the agent to retry difficult configurations within the same episode. Simulation experiments show that MCB improves early-stage learning efficiency by reaching target success-rate levels with fewer interactions, with small collision budgets producing the most consistent gains. Real-world experiments on heterogeneous robot platforms further validate the deployability of the learned policies in cluttered environments.

Shanze Wang, Xinming Zhang, Siwei Cheng +4

College of Information Science and Technology, Eastern Institute of Technology, Ningbo, China · Department of Aeronautical and Aviation Engineering, The Hong Kong Polytechnic University, Hung Hom, Kowloon, Hong Kong · School of Computer Science and Technology, University of Science and Technology of China, Hefei, China +2

RL for Robotics Robot Navigation Sample-Efficient RL Robot Collision Avoidance

Apr 30, 2026cs.RO

LaST-R1: Reinforcing Robotic Manipulation via Adaptive Physical Latent Reasoning

Robotic foundation models require reasoning over complex visual scenes to execute adaptive actions in dynamic environments. While recent studies on latent-reasoning Vision-Language-Action (VLA) models have demonstrated the capability to capture fine-grained physical dynamics, they remain predominantly confined to static imitation learning, severely limiting their adaptability and generalization. In this paper, we present LaST-R1, a novel reinforcement learning (RL) post-training framework designed to effectively harness "latent reasoning-before-acting" policies. Specifically, we propose Latent-to-Action Policy Optimization (LAPO), a core RL algorithm that jointly optimizes the latent reasoning process and the action generation. By explicitly embedding latent Chain-of-Thought (CoT) reasoning directly within the RL optimization loop, LAPO stimulates profound physical world modeling, which in turn drives robust execution in interactive environments. Furthermore, an adaptive latent CoT mechanism is introduced, allowing the policy to dynamically modulate its reasoning horizon based on diverse environment states. Experiments show that LaST-R1 achieves a near-perfect 99.9% average success rate on the LIBERO benchmark with only one-shot supervised warm-up, significantly improving convergence speed and performance over prior state-of-the-art (SOTA) methods. In real-world deployments, LaST-R1 yields up to a 22.5% average improvement over SOTA supervised fine-tuning approach across four complex tasks, including both single-arm and dual-arm settings. Finally, LaST-R1 demonstrates strong generalization across simulated and real-world environments.

Hao Chen, Jiaming Liu, Zhonghao Yan +10

1The Chinese University of Hong Kong · 2State Key Laboratory of Multimedia Information Processing, School of Computer Science, Peking University · 3Simplexity Robotics

RL for Robotics Robot Foundation Models Robotic Manipulation RL for VLMs

Apr 29, 2026cs.RO

Learning Tactile-Aware Quadrupedal Loco-Manipulation Policies

Quadrupedal loco-manipulation is commonly built on visual perception and proprioception. Yet reliable contact-rich manipulation remains difficult: vision and proprioception alone cannot resolve uncertain, evolving interactions with the environment. Tactile sensing offers direct contact observability, but scalable tactile-aware learning framework for quadrupedal loco-manipulation is still underexplored. In this paper, we present a tactile-aware loco-manipulation policy learning pipeline with a hierarchical structure. Our approach has two key components. First, we leverage real-world human demonstrations to train a tactile-conditioned visuotactile high-level policy. This policy predicts not only end-effector trajectories for manipulation, but also the evolving tactile interaction cues that characterize how contact should develop over time. Second, we perform large-scale reinforcement learning in simulation to learn a tactile-aware whole-body control policy that tracks diverse commanded trajectories and tactile interaction cues, and transfers zero-shot to the real world. Together, these components enable coordinated locomotion and manipulation under contact-rich scenarios. We evaluate the system on real-world contact-rich tasks, including in-hand reorientation with insertion, valve tightening, and delicate object manipulation. Compared to vision-only and visuotactile baselines, our method improves performance by 28.54% on average across these tasks.

Pokuang Zhou, Yuhao Zhou, Quan Khanh Luu +7

Purdue University, West Lafayette, IN, USA. · Istituto Italiano di Tecnologia, Genoa, Italy. · Columbia University, New York, NY, USA.

Legged Robotics RL for Robotics Contact-Rich Robotic Manipulation Tactile Sensing for Robotic Manipulation

Apr 28, 2026cs.RO

Egocentric Tactile and Proximity Sensors as Observation Priors for Humanoid Collision Avoidance

Collision-free motion is often aided by tactile and proximity sensors distributed on the body of the robot due to their resistance to occlusion as opposed to external cameras. However, how to shape the sensor's properties, such as sensing coverage; type; and range, to enable avoidant behavior remains unclear. In this work, we present a reinforcement learning framework for whole-body collision avoidance on a humanoid H1-2 robot and use it to characterize how sensor properties shape learned avoidance behavior. Using dodgeball as a benchmark task, we ablate the properties of sensors distributed across the upper body of the robot and find that raw proximity measurements can substitute for explicit object localization provided the sensing range is sufficient and that sparse non-directional proximity signals outpace dense directional alternatives in sample efficiency.

Carson Kohlbrenner, Niraj Pudasaini, William Xie +3

University of Colorado Boulder, 1111 Engineering Drive, Boulder, CO USA.

RL for Robotics Robotic Tactile Sensing Robotic Perception Humanoid Robotics Robot Collision Avoidance

Apr 24, 2026cs.LG

RL Token: Bootstrapping Online RL with Vision-Language-Action Models

Vision-language-action (VLA) models can learn to perform diverse manipulation skills "out of the box," but achieving the precision and speed that real-world tasks demand requires further fine-tuning -- for example, via reinforcement learning (RL). We introduce a lightweight method that enables sample-efficient online RL fine-tuning of pretrained VLAs using just a few hours of real-world practice. We (1) adapt the VLA to expose an "RL token," a compact readout representation that preserves task-relevant pretrained knowledge while serving as an efficient interface for online RL, and (2) train a small actor-critic head on this RL token to refine the actions, while anchoring the learned policy to the VLA. Online RL with the RL token (RLT) makes it possible to fine-tune even large VLAs with RL quickly and efficiently. Across four real-robot tasks (screw installation, zip tie fastening, charger insertion, and Ethernet insertion), RLT improves the speed on the hardest part of the task by up to 3x and raises success rates significantly within minutes to a few hours of practice. It can even surpass the speed of human teleoperation on some of the tasks.

Charles Xu, Jost Tobias Springenberg, Michael Equi +4

Physical Intelligence

RL for Robotics Robot Skill Learning Robotic Manipulation Data-Efficient Robot Learning Vision-Language-Action Models RL Fine-Tuning

Apr 22, 2026cs.RO

Navigating the Clutter: Waypoint-Based Bi-Level Planning for Multi-Robot Systems

Multi-robot control in cluttered environments is a challenging problem that involves complex physical constraints, including robot-robot collisions, robot-obstacle collisions, and unreachable motions. Successful planning in such settings requires joint optimization over high-level task planning and low-level motion planning, as violations of physical constraints may arise from failures at either level. However, jointly optimizing task and motion planning is difficult due to the complex parameterization of low-level motion trajectories and the ambiguity of credit assignment across the two planning levels. In this paper, we propose a hybrid multi-robot control framework that jointly optimizes task and motion planning. To enable effective parameterization of low-level planning, we introduce waypoints, a simple yet expressive representation for motion trajectories. To address the credit assignment challenge, we adopt a curriculum-based training strategy with a modified RLVR algorithm that propagates motion feasibility feedback from the motion planner to the task planner. Experiments on BoxNet3D-OBS, a challenging multi-robot benchmark with dense obstacles and up to nine robots, show that our approach consistently improves task success over motion-agnostic and VLA-based baselines. Our code is available at https://github.com/UCSB-NLP-Chang/navigate-cluster

Jiabao Ji, Yongchao Chen, Yang Zhang +4

UC Santa Barbara · Massachusetts Institute of Technology · Harvard University +2

RL for Robotics Multi-Robot Motion Planning Constrained Motion Planning Multi-Robot Systems Robot Task Planning Motion Planning

Apr 22, 2026cs.RO

Visual-Tactile Peg-in-Hole Assembly Learning from Peg-out-of-Hole Disassembly

Peg-in-hole (PiH) assembly is a fundamental yet challenging robotic manipulation task. While reinforcement learning (RL) has shown promise in tackling such tasks, it requires extensive exploration. In this paper, we propose a novel visual-tactile skill learning framework for the PiH task that leverages its inverse task, i.e., peg-out-of-hole (PooH) disassembly, to facilitate PiH learning. Compared to PiH, PooH is inherently easier as it only needs to overcome existing friction without precise alignment, making data collection more efficient. To this end, we formulate both PooH and PiH as Partially Observable Markov Decision Processes (POMDPs) in a unified environment with shared visual-tactile observation space. A visual-tactile PooH policy is first trained; its trajectories, containing kinematic, visual and tactile information, are temporally reversed and action-randomized to provide expert data for PiH. In the policy learning, visual sensing facilitates the peg-hole approach, while tactile measurements compensate for peg-hole misalignment. Experiments across diverse peg-hole geometries show that the visual-tactile policy attains 6.4% lower contact forces than its single-modality counterparts, and that our framework achieves average success rates of 87.5% on seen objects and 77.1% on unseen objects, outperforming direct RL methods that train PiH policies from scratch by 18.1% in success rate. Demos, code, and datasets are available at https://sites.google.com/view/pooh2pih.

Yongqiang Zhao, Xuyang Zhang, Zhuo Chen +3

Department of Engineering, King’s College London, Strand, London, WC2R 2LS, United Kingdom · Department of Informatics, King’s College London, Strand, London, WC2R 2LS, United Kingdom

RL for Robotics Contact-Rich Robotic Manipulation Robotic Manipulation Robot Imitation Learning Robotic Assembly

Apr 14, 2026cs.RO

FastGrasp: Learning-based Whole-Body Control Method for Fast Dexterous Grasping with Mobile Manipulators

Fast dexterous grasping while a mobile base remains in motion requires coordinated whole-body control and rapid adaptation to physical contact. We propose FastGrasp, a two-stage learning framework that integrates grasp guidance, whole-body control, and tactile feedback. First, a pretrained conditional variational autoencoder generates diverse grasp candidates from object point clouds. Candidates are filtered by approach direction and supporting-surface constraints, then ranked using an envelopment-based criterion combining grasp width and depth coverage measures. Second, a reinforcement learning policy jointly controls the mobile base, arm, and dexterous hand, using the selected grasp as guidance and tactile observations for online grasp adjustment. The policy is trained with domain randomization and deployed with command filtering and tactile-triggered grasp tightening. Simulation experiments show higher grasp success rates than the evaluated baselines under full and partial point-cloud observations, while real-world experiments demonstrate sim-to-real transfer across diverse object geometries.

Heng Tao, Yiming Zhong, Zemin Yang +1

ShanghaiTech University

RL for Robotics Robotic Grasping Mobile Manipulation Whole-Body Robotic Manipulation Whole-Body Control

Apr 5, 2026cs.RO

Learning Dexterous Grasping from Sparse Taxonomy Guidance

Dexterous manipulation requires planning a grasp configuration suited to the object and task, which is then executed through coordinated multi-finger control. However, specifying grasp plans with dense pose or contact targets for every object and task is impractical. Meanwhile, end-to-end reinforcement learning from task rewards alone lacks controllability, making it difficult for users to intervene when failures occur. To this end, we present GRIT, a two-stage framework that learns dexterous control from sparse taxonomy guidance. GRIT first predicts a taxonomy-based grasp specification from the scene and task context. Conditioned on this sparse command, a policy generates continuous finger motions that accomplish the task while preserving the intended grasp structure. Our result shows that certain grasp taxonomies are more effective for specific object geometries. By leveraging this relationship, GRIT improves generalization to novel objects over baselines and achieves an overall success rate of 87.9%. Moreover, real-world experiments demonstrate controllability, enabling grasp strategies to be adjusted through high-level taxonomy selection based on object geometry and task intent.

Juhan Park, Taerim Yoon, Seungmin Kim +10

Department of Artificial Intelligence, Korea University, Seoul, Korea · Korea University of Technology and Education, Cheonan, Korea · Naver AI Lab, Korea +4

RL for Robotics Robotic Grasping Dexterous Manipulation

Apr 1, 2026cs.RO

BAT: Balancing Agility and Stability via Online Policy Switching for Long-Horizon Whole-Body Humanoid Control

Despite advances in reinforcement and imitation learning, achieving both precise control and dynamic whole-body behaviors in long-horizon tasks remains challenging. Existing approaches typically follow two paradigms: coupled whole-body policies for global coordination and decoupled policies for modular precision. However, effectively combining these two types of controllers to achieve agility, robustness, and precision remains challenging. In this work, we propose BAT, an online policyswitching framework that dynamically selects between coupled and decoupled whole-body RL generalist controllers according to the evolving motion context. BAT employs two complementary switching predictors: OpVQ-VAE provides motion-token-based predictions with strong generalization, while OpHRL provides closed-loop, state-aware predictions. A Token-Familiarity Router (TFR) selects between their predictions, favoring OpHRL for familiar token sequences and OpVQ-VAE for unfamiliar ones when their predictions disagree. BAT achieves 85.2% success on seen long-horizon motion combinations, while retaining 71.3% success on unseen motions. BAT also outperforms existing whole-body controllers on individual motions and demonstrates successful zero-shot deployment on the Unitree G1 humanoid.

Donghoon Baek, Sang-Hun Kim, Sehoon Ha

Georgia Institute of Technology, Atlanta, GA, 30308, USA · Samsung Research

RL for Robotics Humanoid Robot Control Whole-Body Control Robot Policy Generalization

Mar 16, 2026cs.RO

CycleRL: Sim-to-Real Deep Reinforcement Learning for Robust Autonomous Bicycle Control

Autonomous bicycles offer a promising agile solution for urban mobility and last-mile logistics. However, conventional control strategies often struggle with underactuated nonlinear dynamics, suffering from sensitivity to model mismatches and limited adaptability to real-world uncertainties. To address this, we develop CycleRL, a comprehensive sim-to-real framework for robust autonomous bicycle control. Our approach establishes a direct perception-to-action mapping within the high-fidelity NVIDIA Isaac Sim environment, leveraging Proximal Policy Optimization (PPO) to optimize the control policy. The framework features a composite reward function tailored for concurrent balance maintenance, velocity tracking, and steering control. Crucially, systematic domain randomization is employed to reduce the reliance on precise system modeling, bridge the simulation-to-reality gap and facilitate direct transfer. In simulation, CycleRL achieves promising performance, including a 99.90% balance success rate, a heading tracking error of 1.15°, and a velocity tracking error of 0.18 m/s. These quantitative results, coupled with successful hardware deployment, validate DRL as an effective paradigm for autonomous bicycle control, offering superior adaptability over traditional methods. Video demonstrations are available at https://cpnt-lab.github.io/CycleRL/.

Gelu Liu, Teng Wang, Zhijie Wu +3

School of Electronics and Communication Engineering and Shenzhen Key Laboratory of Navigation and Communication Integration, Sun Yat-sen University, Shenzhen 518107, China

Robotic Control RL for Robotics Sim-to-Real Transfer Robust RL

Mar 10, 2026cs.RO

From Prior to Pro: Efficient Skill Mastery via Distribution Contractive RL Finetuning

We introduce Distribution Contractive Reinforcement Learning (DICE-RL), a framework that uses reinforcement learning (RL) as a "distribution contraction" operator to refine pretrained generative robot policies. DICE-RL turns a pretrained behavior prior into a high-performing "pro" policy by amplifying high-success behaviors from online feedback. We pretrain a diffusion- or flow-based policy for broad behavioral coverage, then finetune it with a stable, sample-efficient residual off-policy RL framework that combines selective behavior regularization with value-guided action selection. Extensive experiments and analyses show that DICE-RL reliably improves performance with strong stability and sample efficiency. It enables mastery of complex long-horizon manipulation skills directly from high-dimensional pixel inputs, both in simulation and on a real robot. Project website: https://zhanyisun.github.io/dice.rl.2026/.

Zhanyi Sun, Shuran Song

1Stanford University.

RL for Robotics Robot Policy Learning Robot Skill Learning Data-Efficient Robot Learning Residual Policy Learning Policy Learning+2

Mar 5, 2026cs.RO

Latent Policy Steering through One-Step Flow Policies

Offline reinforcement learning (RL) allows robots to learn from offline datasets without risky exploration. Yet, offline RL's performance often hinges on a brittle trade-off between (1) return maximization, which can push policies outside the dataset support, and (2) behavioral constraints, which typically require sensitive hyperparameter tuning. Latent steering offers a structural way to stay within the dataset support during RL, but existing offline adaptations commonly approximate action values using latent-space critics learned via indirect distillation, which can lose information and hinder convergence. We propose Latent Policy Steering (LPS), which enables high-fidelity latent policy improvement by backpropagating original-action-space Q-gradients through a differentiable one-step MeanFlow policy to update a latent-action-space actor. By eliminating proxy latent critics, LPS allows an original-action-space critic to guide end-to-end latent-space optimization, while the one-step MeanFlow policy serves as a behavior-constrained generative prior. This decoupling yields a robust method that works out-of-the-box with minimal tuning. Across OGBench and real-world robotic tasks, LPS achieves state-of-the-art performance and consistently outperforms behavioral cloning and strong latent steering baselines.

Hokyun Im, Andrey Kolobov, Jianlong Fu +1

Department of Artificial Intelligence, Yonsei University · Microsoft Research

RL for Robotics Latent Action Learning Offline RL

Feb 16, 2026cs.RO

AdaptManip: Learning Adaptive Whole-Body Object Lifting and Delivery with Online Recurrent State Estimation

This paper presents Adaptive Whole-body Loco-Manipulation, AdaptManip, a fully autonomous framework for humanoid robots to perform integrated navigation, object lifting, and delivery. Unlike prior imitation learning-based approaches that rely on human demonstrations and are often brittle to disturbances, AdaptManip aims to train a robust loco-manipulation policy via reinforcement learning without human demonstrations or teleoperation data. The proposed framework consists of three coupled components: (1) a recurrent object state estimator that tracks the manipulated object in real time under limited field-of-view and occlusions; (2) a whole-body base policy for robust locomotion with residual manipulation control for stable object lifting and delivery; and (3) a LiDAR-based robot global position estimator that provides drift-robust localization. All components are trained in simulation using reinforcement learning and deployed on real hardware in a zero-shot manner. Experimental results show that AdaptManip significantly outperforms baseline methods, including imitation learning-based approaches, in adaptability and overall success rate, while the learned estimator keeps tracking the object when visual observations are intermittent. We further demonstrate fully autonomous real-world navigation, object lifting, and delivery on a humanoid robot.

Morgan Byrd, Donghoon Baek, Kartik Garg +5

Georgia Institute of Technology, Atlanta, GA, 30308, USA · Georgia Tech Research Institute, Atlanta, GA, 30308, USA

RL for Robotics Robotic Manipulation

Feb 9, 2026cs.RO

High-Speed Vision-Based Flight in Clutter with Safety-Shielded Reinforcement Learning

Quadrotor unmanned aerial vehicles (UAVs) are increasingly deployed in complex missions that demand reliable autonomous navigation and robust obstacle avoidance. However, traditional modular pipelines often incur cumulative latency, whereas purely reinforcement learning (RL) approaches typically provide limited formal safety guarantees. To bridge this gap, we propose an end-to-end RL framework augmented with model-based safety mechanisms. We incorporate physical priors in both training and deployment. During training, we design a physics-informed reward structure that provides global navigational guidance. During deployment, we integrate a real-time safety filter that projects the policy outputs onto a provably safe set to enforce strict collision-avoidance constraints. This hybrid architecture reconciles high-speed flight with robust safety assurances. Benchmark evaluations demonstrate that our method outperforms both traditional planners and recent end-to-end obstacle avoidance approaches based on differentiable physics. Extensive experiments demonstrate strong generalization, enabling reliable high-speed navigation in dense clutter and challenging outdoor forest environments at velocities up to 7.5 m/s}.

Jiarui Zhang, Chengyong Lei, Chengjiang Dai +4

Institute of Cyber-Systems and Control, College of Control Science and Engineering, Zhejiang University, Hangzhou 310027, China · Differential Robotics, Hangzhou 311121, China.

RL for Robotics UAV Navigation Safe RL Vision-Based Robot Control Robot Collision Avoidance

Jan 30, 2026cs.LG

PlatoLTL: Scaling LTL-Guided Multi-Task RL

Linear temporal logic (LTL) has emerged as a powerful formalism for specifying structured, temporally extended tasks in multi-task reinforcement learning (RL). However, while existing approaches in LTL-guided multi-task RL demonstrate success in simple environments, they suffer from challenges in representation learning and exploration efficiency when applied to high-dimensional environments with parameterized specifications. We present PlatoLTL, which elevates state-of-the-art methods to address both challenges. We model atomic propositions as instances of atomic predicates and inject task parameters directly into the goal embedding to enable efficient generalization. We also leverage simple priors on closeness to predicate satisfaction to enable and accelerate learning of complex tasks without biasing the optimal policy. We validate our approach on challenging environments including robotic manipulation and multi-drone navigation.

Jacques Cloete, Mathias Jackermeier, Ioannis Havoutis +1

Oxford Robotics Institute University of Oxford Oxford, United Kingdom · Department of Computer Science University of Oxford Oxford, United Kingdom

RL for Robotics Constrained RL Multi-Task RL Generalization in RL Goal-Conditioned RL

Jan 28, 2026cs.RO

OGPO: One-Step Generative Policy Optimization for Real-Time Robot Control

Real-time robot control demands fast action generation. Diffusion and flow matching policies for robot control require multi-step sampling, limiting their deployment in real-time scenarios. Natively reducing the sampling steps to one sacrifices representation quality and task performance, creating a trilemma among speed, fidelity, and performance. We present One-Step Generative Policy Optimization (OGPO), a systematic framework to resolve this trilemma. OGPO first pairs a lightweight architecture with the interval velocity principle for distillation-free one-step inference, while representation spreading prevents representation quality degradation. It then performs on-policy reinforcement learning (RL) fine-tuning on this fast, stable policy to break the imitation learning ceiling. Experiments on RoboMimic and OpenAI Gym benchmarks show that OGPO matches or exceeds multi-step baselines while achieving a 5-20 times inference speedup and over 120Hz control frequency. Physical deployment on a Franka-Emika-Panda robot validates real-world applicability. Project page: https://ogpo-project.github.io/

Guowei Zou, Haitao Wang, Hejun Wu +3

School of Computer Science and Engineering, Sun Yat-sen University, Guangzhou, China · Guangdong Key Laboratory of Big Data Analysis and Processing, Guangzhou, China

Robotic Control RL for Robotics Robot Policy Learning Robot Imitation Learning

Jan 27, 2026cs.RO

Reversible Simplex Supervision with Post-Action Debt Accounting for Goal-Reaching RL

Deploying reinforcement learning (RL) on multi-tonne robots calls for supervisory mechanisms that address both operational safety and progress toward task completion. However, repeated switching need not preserve task progress when a learned action increases storage before recovery takes control. We introduce reversible Simplex supervision with post-action debt accounting for a frozen finite-state policy and robust-adaptive recovery. Under exact sampled-state information and stated model and certificate conditions, we prove that recovery repayment exceeding a uniform triggering-edge debt bound guarantees finite switching and finite-sample goal entry. We formulate reachability-based certificate constructions for establishing these sufficient conditions. In 20 matched simulations, goal-entry counts are 20 with debt gating and 18 without it; the two remaining runs terminate under the supervisor's admissibility stopping rule. On an experimental 6000 kg robot, 24 asphalt and soft-terrain trials evaluate 50 ms supervision above a 1 kHz actuator stack; all eight triggered recoveries complete debt-gated re-entry. The experiments demonstrate the supervisory mechanism in the tested trials.

Mehdi Heydari Shahna, Joongheon Kim, Jouni Mattila

Faculty of Engineering and Natural Sciences, Tampere University, Tampere, Finland · Department of Electrical and Computer Engineering, Korea University, Seoul, Republic of Korea

RL for Robotics Robot Safety Runtime Assurance Safe RL

Jan 22, 2026cs.RO

AION: Aerial Indoor Object-Goal Navigation Using Dual-Policy Reinforcement Learning

Object-Goal Navigation (ObjectNav) requires an agent to autonomously explore an unknown environment and navigate toward target objects specified by a semantic label. While prior work has primarily studied zero-shot ObjectNav under 2D locomotion, extending it to aerial platforms with 3D locomotion capability remains underexplored. Aerial robots offer superior maneuverability and search efficiency, but also introduce new challenges in spatial perception, dynamic control, and safety assurance. In this paper, we propose AION for vision-based aerial ObjectNav without relying on external localization or global maps. AION is an end-to-end dual-policy reinforcement learning (RL) framework that decouples exploration and goal-reaching behaviors into two specialized policies. We evaluate AION on the AI2-THOR benchmark and further assess its real-time performance in IsaacSim using high-fidelity drone models. Experimental results show that AION achieves superior performance across comprehensive evaluation metrics in exploration, navigation efficiency, and safety. The project is available at https://github.com/Zichen-Yan/AION.

Zichen Yan, Yuchen Hou, Shenao Wang +3

Department of Electrical and Computer Engineering, National University of Singapore, Singapore.

RL for Robotics Aerial Robotics Vision-Based Robot Control Object-Goal Navigation

Nov 14, 2025cs.RO

Humanoid Whole-Body Badminton via an Annealed Reinforcement Learning Curriculum

Humanoid robots have demonstrated strong capabilities for interacting with static scenes across locomotion and manipulation, yet dynamic real-world interactions remain challenging. As a step toward fast-moving object interactions, we present an RL training pipeline that yields a unified whole-body controller for humanoid badminton, coordinating footwork and striking without motion priors or expert demonstrations. In badminton, locomotion and striking are tightly coupled, making the final hitting objective difficult to optimize directly due to sparse rewards and conflicting gradients. We address this with an annealed curriculum that first stabilizes learning through auxiliary locomotion objectives, then progressively removes them to focus optimization on the final hitting objective. For deployment, we use an Extended Kalman Filter (EKF) to estimate and predict shuttlecock trajectories for target striking, and also develop a prediction-free variant that removes the EKF and explicit prediction. We validate the framework in simulation and on hardware. In simulation, two robots sustain a rally of 21 consecutive hits. In real-world tests with both machine-fed shuttles and human-robot rallies, the robot achieves outgoing shuttle speeds up to 19.1 m/s. Moreover, the prediction-free variant attains comparable performance to the EKF-based policy. Overall, our approach enables dynamic yet precise goal striking in humanoid badminton and suggests a path toward more dynamics-critical whole-body interaction tasks.

Chenhao Liu, Leyun Jiang, Ningyuan Tian +4

RL for Robotics Reinforcement Learning Humanoid Robot Control Curriculum RL Robotics

Oct 30, 2025cs.RO

Thor: Towards Human-Inspired Whole-Body Reactions for Intense Contact-Rich Environments

Maintaining whole-body stability and motion tracking under large interaction forces remains challenging for humanoids. We present Thor, a reinforcement learning framework for forceful humanoid loco-manipulation. Thor jointly trains lower-body, waist, and upper-body policies with shared whole-body observations and body-specific rewards to coordinate locomotion and force adaptation, waist posture regulation, and upper-body motion tracking. We further introduce a force-adaptive torso-tilt (FAT2) objective that derives a load-dependent horizontal center-of-mass offset reference from quasi-static moment balance. Capacity-matched simulation blations show that the three-policy architecture improves tracking under large external force disturbances, while real-world ablations demonstrate that FAT2 increases peak pulling capability. On the Unitree G1, Thor achieves mean peak dual-hand pulling forces of 167.7 N and 145.5 N during backward and forward locomotion, exceeding the best-performing baseline by 68.9% and 74.7%, respectively. Real-world demonstrations include opening a fire door with one hand using approximately 60 N of pulling force and towing a 1.7-ton passenger car.

Gangyang Li, Hongzhe Shi, Qing Shi +5

Intelligent Robotics Institute, School of Mechatronical Engineering, Beijing Institute of Technology · Beijing Academy of Artificial Intelligence

RL for Robotics Humanoid Loco-Manipulation Contact-Rich Robotic Manipulation Humanoid Robot Control Whole-Body Control

Oct 16, 2025cs.RO

CBF-RL: Safety Filtering Reinforcement Learning in Training with Control Barrier Functions

Reinforcement learning (RL), while powerful and expressive, can often prioritize performance at the expense of safety. Yet safety violations can lead to catastrophic outcomes in real-world deployments. Control Barrier Functions (CBFs) offer a principled method to enforce dynamic safety -- traditionally deployed online via safety filters. While the result is safe behavior, the fact that the RL policy does not have knowledge of the CBF can lead to conservative behaviors. This paper proposes CBF-RL, a framework for generating safe behaviors with RL by enforcing CBFs in training. CBF-RL has two key attributes: (1) minimally modifying a nominal RL policy to encode safety constraints via a CBF term, (2) and safety filtering of the policy rollouts in training. Theoretically, we prove that continuous-time safety filters can be deployed via closed-form expressions on discrete-time roll-outs. Practically, we demonstrate that CBF-RL internalizes the safety constraints in the learned policy -- both enforcing safer actions and biasing towards safer rewards -- enabling safe deployment without the need for an online safety filter. We validate our framework through ablation studies on navigation tasks and on the Unitree G1 humanoid robot, where CBF-RL enables safer exploration, faster convergence, and robust performance under uncertainty, enabling the humanoid robot to avoid obstacles and climb stairs safely in real-world settings without a runtime safety filter.

Lizhi Yang, Blake Werner, Massimiliano de Sa +1

Caltech MCE

RL for Robotics Control Barrier Functions Constrained RL Safety Filtering Safe RL

Oct 14, 2025cs.RO

Pretraining in Actor-Critic Reinforcement Learning for Locomotion

The pretraining-finetuning paradigm has facilitated numerous transformative advancements in artificial intelligence research in recent years. However, in the domain of reinforcement learning (RL) for robot locomotion, individual skills are often learned from scratch despite the high likelihood that some generalizable knowledge is shared across all task-specific policies belonging to the same robot embodiment. This work aims to define a paradigm for pretraining neural network models that encapsulate such knowledge and can subsequently serve as a basis for warm-starting the RL process in classic actor-critic algorithms, such as Proximal Policy Optimization (PPO). We begin with a task-agnostic exploration-based data collection algorithm to gather diverse, dynamic transition data, which is then used to train a Proprioceptive Inverse Dynamics Model (PIDM) through supervised learning. The pretrained weights are then loaded into both the actor and critic networks to warm-start the policy optimization of actual tasks. We systematically validated our proposed method with 9 distinct robot locomotion RL environments comprising 3 different robot embodiments, showing significant benefits of this initialization strategy. Our proposed approach on average improves sample efficiency by 36.2% and task performance by 4.3% compared to random initialization. We further present key ablation studies and empirical analyses that shed light on the mechanisms behind the effectiveness of this method.

Jiale Fan, Andrei Cramariuc, Tifanny Portela +1

Section of Microengineering, EPFL, 1015 Lausanne, Switzerland · Robotic Systems Lab, ETH Zurich, 8092 Zürich, Switzerland · ETH AI Center, ETH Zurich, 8092 Zürich, Switzerland

RL for Robotics Reinforcement Learning Robot Locomotion Actor-Critic Methods