Simulated Driving

Simulated driving environments are crucial for developing and testing autonomous vehicle (AV) systems, primarily focusing on validating planning algorithms and human-machine interfaces (HMIs) under controlled conditions. Current research emphasizes creating realistic and adversarial traffic simulations using techniques like reinforcement learning (RL), incorporating haptic feedback for improved driver control, and employing large language models (LLMs) to automate reward function design in RL. These advancements are vital for improving AV safety and efficiency, bridging the gap between simulated and real-world driving performance, and providing valuable insights into human-AV interaction.