Learning to Drift with Individual Wheel Drive: Maneuvering Autonomous Vehicle at the Handling Limits

TL;DR

This paper introduces a reinforcement learning framework with domain randomization and GPU simulation to enable precise drifting control of an autonomous vehicle with individual wheel drive, successfully transferring from simulation to real-world RC car experiments.

Contribution

It presents a novel RL-based drifting control method that effectively bridges the simulation-to-reality gap using systematic domain randomization and parallel simulation on a custom IWD RC platform.

Findings

Successful transfer of drifting policies from simulation to real RC car.

Achieved precise trajectory tracking during complex maneuvers.

Maintained controlled sideslip angles in various scenarios.

Abstract

Drifting, characterized by controlled vehicle motion at high sideslip angles, is crucial for safely handling emergency scenarios at the friction limits. While recent reinforcement learning approaches show promise for drifting control, they struggle with the significant simulation-to-reality gap, as policies that perform well in simulation often fail when transferred to physical systems. In this paper, we present a reinforcement learning framework with GPU-accelerated parallel simulation and systematic domain randomization that effectively bridges the gap. The proposed approach is validated on both simulation and a custom-designed and open-sourced 1/10 scale Individual Wheel Drive (IWD) RC car platform featuring independent wheel speed control. Experiments across various scenarios from steady-state circular drifting to direction transitions and variable-curvature path following demonstrate that our approach achieves precise trajectory tracking while maintaining controlled sideslip angles throughout complex maneuvers in both simulated and real-world environments.