Reset-Free Reinforcement Learning for Real-World Agile Driving: An Empirical Study
2026-04-09 • Robotics
Robotics
AI summaryⓘ
The authors studied how small robotic cars can learn to drive fast on slippery floors without needing people to reset them after mistakes. They tested different learning methods both in computer simulations and on real cars. They found that some approaches that worked well in simulations did not work as well on real cars, especially methods that tried to improve a baseline controller through extra learning. Their results show that learning to drive in the real world has unique problems that current algorithms don't fully solve.
reinforcement learningreset-free learningModel Predictive Path Integral (MPPI)residual learningPPOSACTD-MPC2sim-to-real transferagile drivingrobotic vehicles
Authors
Kohei Honda, Hirotaka Hosogaya
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.