Overview

Speaker: Astghik Hakobyan
Date: Friday, Aug 05, 2022 
Time: 5:00 PM
Venue: Museum Lecture Hall
Description: The lecture introduces recent advances in modern control theory, its integration with AI, as well as novel methods for safety assurance in robotic systems. Two state-of-the-art methods called model predictive control (MPC) and meta-reinforcement learning (meta-RL) will be presented with their application to robot control. Both approaches provide high-performance control actions with safety guarantees. Last but not least, the utility of both methods will be demonstrated via real-world experiments on a restaurant service robot and simulations in an autonomous driving scenario.