Failure-Aware Iterative Learning of State-Control Invariant Sets
A learning-from-failure framework for learning the maximal control invariant set and its invariance-preserving inputs.
Bio
Ahmad Amine is a third-year PhD student in Electrical and Systems Engineering at the University of Pennsylvania, advised by Professor Rahul Mangharam. He previously obtained his MSE in Robotics at the University of Pennsylvania and his BE in Electrical and Computer Engineering with a minor in Mathematics at the American University of Beirut, Lebanon.
His research focuses on developing safe, learning-based control methods for autonomous robotic systems. In particular, he studies how to combine model predictive control with online learning and system identification to enable robots to improve their performance over time while maintaining safety guarantees, even under model uncertainty and in unexplored operating conditions.
Ahmad also co-instructs the Roboracer autonomous racing course at Penn and organizes the Roboracer competition at international robotics conferences including ICRA and CDC.