Autonomous Parking

Autonomous parking research aims to develop systems enabling vehicles to park themselves without human intervention, focusing on safe and efficient maneuvers in diverse environments. Current efforts concentrate on improving localization accuracy using sensor fusion (e.g., visual-inertial odometry) and employing advanced path planning algorithms like Hybrid A*, RRT variants (often incorporating continuous-curvature paths), and neural network-based approaches (including imitation learning and generative models) to generate collision-free trajectories. These advancements are crucial for enhancing the safety and reliability of autonomous driving systems, particularly in challenging indoor settings like underground parking garages, and for optimizing the use of urban space.