Reset-Free Reinforcement Learning for Real-World Agile Driving: An Empirical Study

• URL: http://arxiv.org/abs/2604.07672v1
• Date: Thu, 09 Apr 2026 00:28:54 GMT
• Title: Reset-Free Reinforcement Learning for Real-World Agile Driving: An Empirical Study
• Abstract summary: This paper presents an empirical study of reset-free reinforcement learning (RL) for real-world agile driving.<n>We employ Model Predictive Path Integral control (MPPI) as both the reset policy and the base policy for residual learning.<n>Our results reveal a clear gap between simulation and real-world: SAC with residual learning achieves the highest returns in simulation, yet only TD-MPC2 consistently outperforms the MPPI baseline on the physical platform.
• Score: 3.038642416291856
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: This paper presents an empirical study of reset-free reinforcement learning (RL) for real-world agile driving, in which a physical 1/10-scale vehicle learns continuously on a slippery indoor track without manual resets. High-speed driving near the limits of tire friction is particularly challenging for learning-based methods because complex vehicle dynamics, actuation delays, and other unmodeled effects hinder both accurate simulation and direct sim-to-real transfer of learned policies. To enable autonomous training on a physical platform, we employ Model Predictive Path Integral control (MPPI) as both the reset policy and the base policy for residual learning, and systematically compare three representative RL algorithms, i.e., PPO, SAC, and TD-MPC2, with and without residual learning in simulation and real-world experiments. Our results reveal a clear gap between simulation and real-world: SAC with residual learning achieves the highest returns in simulation, yet only TD-MPC2 consistently outperforms the MPPI baseline on the physical platform. Moreover, residual learning, while clearly beneficial in simulation, fails to transfer its advantage to the real world and can even degrade performance. These findings reveal that reset-free RL in the real world poses unique challenges absent from simulation, calling for further algorithmic development tailored to training in the wild.

Related papers

• Continual-RL for Generalization in Autonomous Racing on the RoboRacer Platform [6.737714968545043]
  A key challenge in modern robotics is to adapt to changing environments.<n>In particular, learning to drive on a new track-floor combination with the least amount of new experience naturally frames a continual learning problem.<n>This work proposes a continual RL framework based on Continual Backpropagation that is able to train a generalistic policy on a set of tracks and then fine- tune it within 15 minutes to outperform classical controllers.
  arXiv  Detail & Related papers  (2026-07-27T12:01:29Z)
• Object-Centric Residual RL for Zero-Shot Sim-to-Real VLA Enhancement [52.93267443851685]
  Vision-Language-Action (VLA) models can generalize across diverse manipulation tasks.<n>Can a reinforcement learning policy trained purely in simulation improve the robustness of real-world VLAs zero-shot?<n>We propose an object-centric residual RL framework that refines VLA actions using object poses.
  arXiv  Detail & Related papers  (2026-06-17T11:36:54Z)
• Physics-Informed Reinforcement Learning of Spatial Density Velocity Potentials for Map-Free Racing [73.88859384645264]
  Out-Of-Distribution (OOD) generalization to various racetracks utilizes Machine Learning (ML) to encode the mathematical relation between sensor data and vehicle actuation for end-to-end control.<n>This paper presents a DRL method that parameterizes nonlinear vehicle dynamics from the spectral distribution of depth measurements with a non-geometric, physics-informed reward, to infer vehicle time-optimal and overtaking racing controls with an Artificial Neural Network (ANN)<n>The policy outperforms human demonstrations by 12% in OOD tracks on proportionally scaled hardware, by maximizing the friction circle with tire dynamics that resemble an empirical Pacejka tire model.
  arXiv  Detail & Related papers  (2026-04-10T17:12:07Z)
• Incremental Residual Reinforcement Learning Toward Real-World Learning for Social Navigation [2.1204495827342438]
  Social navigation has emerged as a critical task, driving research into deep reinforcement learning (RL) approaches.<n>Real-world RL, in which agents learn while operating directly in physical environments, presents a promising solution to this issue.<n>In this study, we propose incremental residual RL (IRRL), which enhances learning efficiency by training only on the residuals relative to a base policy.
  arXiv  Detail & Related papers  (2026-04-09T08:08:37Z)
• Human-in-the-loop Online Rejection Sampling for Robotic Manipulation [55.99788088622936]
  Hi-ORS stabilizes value estimation by filtering out negatively rewarded samples during online fine-tuning.<n>Hi-ORS fine-tunes a pi-base policy to master contact-rich manipulation in just 1.5 hours of real-world training.
  arXiv  Detail & Related papers  (2025-10-30T11:53:08Z)
• Simulation Priors for Data-Efficient Deep Learning [56.525770511247934]
  SimPEL is a method that efficiently combines first-principles models with data-driven learning.<n>We evaluate SimPEL on diverse systems, including biological, agricultural, and robotic domains.<n>For decision-making, we demonstrate that SimPEL bridges the sim-to-real gap in model-based reinforcement learning.
  arXiv  Detail & Related papers  (2025-09-06T14:36:41Z)
• RIFT: Group-Relative RL Fine-Tuning for Realistic and Controllable Traffic Simulation [13.319344167881383]
  We introduce a dual-stage AV-centric simulation framework that conducts imitation learning pre-training in a data-driven simulator.<n>We then learn fine-tuning in a physics-based simulator to enhance style-level controllability.<n>In the fine-tuning stage, we propose RIFT, a novel group-relative RL fine-tuning strategy.
  arXiv  Detail & Related papers  (2025-05-06T09:12:37Z)
• Offline Robotic World Model: Learning Robotic Policies without a Physics Simulator [50.191655141020505]
  Reinforcement Learning (RL) has demonstrated impressive capabilities in robotic control but remains challenging due to high sample complexity, safety concerns, and the sim-to-real gap.<n>We introduce Offline Robotic World Model (RWM-O), a model-based approach that explicitly estimates uncertainty to improve policy learning without reliance on a physics simulator.
  arXiv  Detail & Related papers  (2025-04-23T12:58:15Z)
• A Platform-Agnostic Deep Reinforcement Learning Framework for Effective Sim2Real Transfer towards Autonomous Driving [0.0]
  Deep Reinforcement Learning (DRL) has shown remarkable success in solving complex tasks. transferring DRL agents to the real world is still challenging due to the significant discrepancies between simulation and reality. We propose a robust DRL framework that leverages platform-dependent perception modules to extract task-relevant information.
  arXiv  Detail & Related papers  (2023-04-14T07:55:07Z)
• Tackling Real-World Autonomous Driving using Deep Reinforcement Learning [63.3756530844707]
  In this work, we propose a model-free Deep Reinforcement Learning Planner training a neural network that predicts acceleration and steering angle. In order to deploy the system on board the real self-driving car, we also develop a module represented by a tiny neural network.
  arXiv  Detail & Related papers  (2022-07-05T16:33:20Z)
• Cloud-Edge Training Architecture for Sim-to-Real Deep Reinforcement Learning [0.8399688944263843]
  Deep reinforcement learning (DRL) is a promising approach to solve complex control tasks by learning policies through interactions with the environment. Sim-to-real approaches leverage simulations to pretrain DRL policies and then deploy them in the real world. This work proposes a distributed cloud-edge architecture to train DRL agents in the real world in real-time.
  arXiv  Detail & Related papers  (2022-03-04T10:27:01Z)
• Learning from Simulation, Racing in Reality [126.56346065780895]
  We present a reinforcement learning-based solution to autonomously race on a miniature race car platform. We show that a policy that is trained purely in simulation can be successfully transferred to the real robotic setup.
  arXiv  Detail & Related papers  (2020-11-26T14:58:49Z)
• Information Theoretic Model Predictive Q-Learning [64.74041985237105]
  We present a novel theoretical connection between information theoretic MPC and entropy regularized RL. We develop a Q-learning algorithm that can leverage biased models.
  arXiv  Detail & Related papers  (2019-12-31T00:29:22Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.