EAGER: A Synergistic Framework for Motion and Task Planning in Mixed Continuous and Discrete Spaces

点击翻译按钮获取中文摘要

This exploratory project is to increase the ability of robots to plan and act on their own in order to safely complete high-level tasks. Whether the task is to search, inspect, manipulate objects, or navigate to target destinations, it generally involves abstractions into discrete, logical actions, where each discrete action often requires complex collision-free motions in order to be implemented. This project establishes a unified framework for handling the discrete and continuous spaces that arise in these problems.Crucial to this goal, as robots are deployed into less and less structured environments, is their ability to reason and plan at multiple levels of discrete and continuous abstractions. This research makes it possible to specify sophisticated tasks using Planning Domain Definition Languages and automatically compute a sequence of control inputs such that the resulting trajectory satisfies the PDDL task specification, avoids collisions with obstacles, and obeys the motion dynamics and the constraints imposed by physics-based interactions with the environment.