Event Information

Hybrid Seminar: Control for Human-Centric Autonomous Systems---Challenges and Solutions Monday, December 12, 2022 11:00 a.m. 2460 AV Williams Bldg and on Zoom For More Information: https://umd.zoom.us/j/94931160742 Hybrid Seminar Control for Human-Centric Autonomous Systems---Challenges and Solutions Sandra Hirche TUM Liesel Beckmann Distinguished Professor Technische Universität München Munich, Germany In-person: 2460 AV Williams Bldg Zoom: https://umd.zoom.us/j/94931160742 Passcode 295348 Abstract The symbiosis of AI and intelligent control has led to human-like AI for automation, aiming at human-level capabilities. The paradigm is limited and research is trending towards human augmentation through AI, rather than pure AI automation. Consequently the systems and control community has identified cyber-physical human systems as a key scientific challenge for the next decade. Sandra Hirche will present selected challenges and solutions for controlling human-centric autonomous systems. She will present her recent developments in online control for personalized robotic rehabilitation and assistance. This is motivated by a central societal challenge to prolong independent living for the elderly. Personalized robotic rehabilitation and assistance is considered an enabling technology, with control design playing a significant role. Focusing on sensorimotor rehabilitation and assistance, personalized control should be able to adapt to high interpersonal variability in human motor behavior and also to intrapersonal changes over time. Control adaptation is challenged by the sparsity of person-specific data, since calibration routines need to be brief for user acceptance. Above all, guaranteed safety is key. Hirche will present recent results on learning-based control with performance and safety guarantees for highly uncertain systems. Epistemic uncertainty because of limited training data will explicitly be taken into account to achieve robust uncertainty-aware behavior in a closed loop system. Online learning as well as real-time capabilities are also important and will be covered in this talk. Biography Sandra Hirche holds the TUM Liesel Beckmann Distinguished Professorship at the Technische Universität München in Munich, Germany. She also heads the Chair of Information-Oriented Control in the Faculty of Electrical and Computer Engineering at TUM since 2013. She received the diploma engineer degree in Aeronautical and Aerospace Engineering in 2002 from the Technische Universität Berlin, and the Doctor of Engineering Degree in Electrical and Computer Engineering from the Technische Universität München in 2005. Her main research interests are in learning, cooperative and networked control with applications to human-robot interaction, multi-robot systems and general robotics. She has published more than 200 papers in international journals, books and refereed conferences, and has been awarded multiple prestigious awards. Hirche is a fellow of IEEE and has received its Control System Society Distinguished Member Award.