Novel Planning

Novel planning research focuses on developing algorithms that enable agents, particularly robots and autonomous vehicles, to make effective decisions in complex, uncertain environments. Current efforts concentrate on incorporating elements like risk assessment, human-robot interaction modeling (often using epistemic logic and AND-OR search), and active information gathering (leveraging techniques like graph neural networks and value of information calculations) into planning frameworks. These advancements aim to improve the safety, efficiency, and robustness of autonomous systems across diverse applications, from navigation in unknown terrains to collaborative task execution with humans.