Robots are usually programmed for repetitive tasks in a structured environment or particular tasks in a less structured environment with a considerable amount of hand-crafted tuning work. Recently, the attention of robotics is redirected from mass production to mass customization. Existing planning and control techniques have a very limited capacity to respond to such a trend: whenever a new type of tasks or a new robot with different dynamics is presented, the hand-crafted baseline planning and control algorithms for the robots usually have to be re-derived and the actions for robots to take have to be re-programmed. This talk will present my group’s recent studies on learning-based planning and control that aim to overcome such limitations and establish robotic systems that are capable to continuously learn from and collaborate with others. I will particularly discuss these studies within the applications to drones for their mass customization and applications, as well as collaborative robots for disassembly to improve efficiency and effectiveness of e-waste recycling and remanufacturing.
Dr. Minghui Zheng is currently an Assistant Professor in the Department of Mechanical and Aerospace Engineering, University at Buffalo (UB). Before joining UB, she received her Ph.D. degree in Mechanical Engineering from University of California, Berkeley in 2017. She received her B.S. degree in Engineering Mechanics and M.S. degree in Control Science and Engineering, both from Beihang University, Beijing, China. Her research interests include learning, planning, and control with broad applications in autonomous aerial vehicles and intelligent remanufacturing. She was the recipient of the National Science Foundation CAREER Award (2021).