Lifting as an Abstraction for Control Synthesis in Cyber-Physical Systems

Explore the concept of lifting as an abstraction in control synthesis for cyber-physical systems in this 29-minute talk by Necmiye Ozay from the University of Michigan. Delve into the comparison between discrete abstractions and continuous system simplification through lifting to higher-dimensional linear systems. Examine the similarities with Kernel embeddings in machine learning and investigate how these lifting techniques can be applied to controller synthesis. Discover the relationship between lifting and Koopman embeddings, and consider potential applications of lifting concepts in discrete reactive systems and software synthesis. Gain insights into this innovative approach to system simplification and control design presented at the Simons Institute's "Synthesis of Models and Systems" event.