Erion Plaku, Motion Planning for Hybrid Robotics Systems

Nowadays robotic systems are designed to perform increasingly complex tasks. Robots are expected to explore unknown, dynamic, or possibly hazardous environments, quickly modifying their controls to respond to unanticipated changes in the environment, mishaps, or other failures. For example, a vehicle may be required to employe different controls over different terrains due to safety issues, and a reconfigurable robot may change its shape and use different gaits to climb, crawl, or walk fast. Such changes in the robot behavior are often realized by instantaneously switching to a different operating mode.

This talk presents an effective motion planner for these hybrid robotic systems that blends discrete and continuous searching. The multi-layered approach offers considerable improvements over existing methods addressing similar problems and is capable of planning motions for complex hybrid robotic systems.