Mechanical Engineering Fall Seminar Series: Class 530.803
“Robust Decision and Control of Autonomous Robotic Vehicles”
Presented by Professor Marin Kobilarov
Department of Mechanical Engineering, Johns Hopkins University
This talk will consider the problem of reliable autonomous navigation of robotic vehicles, such as unmanned aerial vehicles, autonomous underwater vehicles, or self-driving cars. The decision and control algorithms for such systems must be designed to efficiently achieve a mission objective, such as reaching a goal location in a cluttered and dynamic environment, but also remain safe and robust to the uncertainties of the real world. We will propose computational-theoretic methods to analyze this trade-off and discuss ongoing development of algorithms with built-in robustness based on statistical learning theory.
Marin Kobilarov is an assistant professor in mechanical engineering and computer science at the Johns Hopkins University, and the director of the Autonomous Systems, Control and Optimization Lab. He received his Ph.D. from University of Southern California in 2008 and was the Keck post-doctoral fellow in control and dynamical systems at the California Institute of Technology during 2009-2012. His group focuses on computational theory, algorithms and software development, for decision making, planning, control and system integration of robotic systems operating in complex environments. Their work has been nominated for best papers at RSS and ICRA. Current applications are: unmanned aerial vehicles operating through contact with the environment, self-driving robotic vehicles, autonomous underwater vehicles for environmental assessment, and robot-assisted eye surgery.