AWS Deepracer

AWS DeepRacer is a platform for developing and testing reinforcement learning (RL) models for autonomous driving, primarily using simulated and physical race tracks. Current research focuses on improving model performance and robustness, particularly addressing the "sim-to-real" gap and optimizing training efficiency through techniques like active testing and the use of specialized hardware like AWS Trainium and Inferentia2 chips. This work contributes to advancements in RL algorithms, particularly in areas such as object avoidance and mapless navigation, with implications for both autonomous vehicle development and broader RL research.