Head to Head Autonomous Racing

Head-to-head autonomous racing research focuses on developing algorithms enabling vehicles to compete effectively and safely in direct, high-speed races against opponents. Current work emphasizes robust perception using sensor fusion (e.g., LiDAR and cameras) to inform model-based reinforcement learning (MBRL) or hierarchical control architectures, often incorporating game-theoretic approaches to handle strategic decision-making like overtaking and defensive maneuvers. These advancements improve the reliability and performance of autonomous racing systems, contributing to broader progress in multi-agent robotics, control theory, and the development of safer and more efficient autonomous vehicles.