Title: Safety Assurance in Learning-enabled Autonomous Systems: A Contraction-based Approach
Date: Monday,March 4th, 2024
Time: 11:00 am – 12:00 pm
Location: EIT 3142
Abstract: This seminar by Saber Jafarpour focuses on certifying the safety of learning-enabled autonomous systems. It presents theoretical and algorithmic methods to ensure safety by studying reachability, using contraction and monotone system theory for efficient over-approximation of reachable sets, and integrating state-of-the-art machine learning verification for safety verification.
Speaker Bio: Saber Jafarpour is a Research Assistant Professor at the University of Colorado Boulder, with previous postdoctoral positions at Georgia Institute of Technology and the University of California Santa Barbara. He earned his Ph.D. from Queen’s University Canada and specializes in safe control and learning in autonomous systems.
This event addresses the urgent need for reliable safety measures in autonomous systems, offering novel insights and methods for researchers and practitioners.