“Building Human-robot Symbiosis with Game-theoretic Intent Inference and Motion Planning” by Wenlong Zhang
Abstract: Human-robot interaction has become ubiquitous in workplace and daily life. Besides safety guarantee, human-friendly robots need to estimate the human intents in real time and dynamically adjust their actions to collaborate with humans. I will discuss our game-theoretic framework for intent estimate and motion planning for an autonomous vehicle to interact with a human-driven vehicle in a safe and socially-adept manner. I will also present some of my labโs recent effort to more efficiently estimate the humanโs intent and find optimal policies for future autonomous vehicles. Simulation results will be presented to demonstrate the human-like driving behaviors enabled by the developed algorithms. This talk will conclude with some open problems and possible directions for building human-robot symbiosis.
Bio: Dr. Wenlong Zhang is an Associate Professor in the School of Manufacturing Systems and Networks at Arizona State University (ASU). He received his M.S. in mechanical engineering, M.A. in statistics, and Ph.D. in mechanical engineering, all from the University of California, Berkeley. He is directing the ASU Robotics and Intelligent Systems Laboratory (RISE Lab). His research interests include designing compliant robots and dynamics-aware planning algorithms for robots to interact safely and efficiently with humans and complex environments. He is a recipient of the NSF CAREER Award and Bisgrove Junior Faculty Award from Science Foundation Arizona.