Robotics Motion Planning and Learning
Robotics Motion Planning and Learning
Abstract
A very successful approach for finding the motions for a robot to move from some initial configuration to a goal configuration is sampling-based motion planning. In this approach, the planner performs a systematic exploration, through sampling, of the configuraton space in order to build a connectivity map that the robot can follow. These planners trade completeness for probabilistic completeness, which means that given enough time, they will find existing paths with high probability, although they are not able to tell when there is no such path. In this talk, I will discuss how, while exploring the configuration space, sampling based planners are also able to extract features that can be useful to produce better maps. I will also briefly talk about my work on multi-agent planning and autonomous navigation.
Bio
Marco Morales Aguirre is a Teaching Associate Professor at the Department of Computer Science at the University of Illinois Urbana-Champaign and he is currently on leave as an Associate Professor at the Department of Computer Science at Instituto Tecnológico Autónomo de México (ITAM). He has also been a Visiting Professor at Texas A&M University and a Lecturer at Universidad Nacional Autónoma de México (UNAM). He holds a Ph.D. in Computer Science from Texas A&M University, a M.S. in Electrical Engineering and a B.S. in Computer Engineering from UNAM. His main research interests are in motion planning and control for autonomous robots, artificial intelligence, machine learning, and computational geometry.