Online Replanning
Online replanning focuses on dynamically adjusting plans for robots and autonomous systems in response to unexpected changes in the environment or execution errors. Current research emphasizes efficient algorithms, such as model predictive control and diffusion models, to generate and adapt trajectories in real-time, often incorporating techniques like waypoint management and trajectory optimization to minimize computational cost and ensure safety. This field is crucial for enabling robust and adaptable autonomous systems in complex, dynamic environments, with applications ranging from drone racing and robotic manipulation to multi-agent coordination and delivery systems.
Papers
February 7, 2024
October 14, 2023
June 26, 2023
March 30, 2023
March 2, 2023
July 19, 2022