Theoretical Foundation of Learning-based Planning for Linear Temporal Logic Objectives

Abstract:  Modern autonomous systems and robots increasingly employ Linear Temporal Logic (LTL) to express complex planning objectives that involve logical reasoning over temporal events in various applications such as transportation, manufacturing, and military operations. However, traditional planning methods for LTL objectives often require full knowledge of the environment, which is typically unattainable in real-world scenarios. Recent works have shown the potential of reinforcement learning (RL) to handle LTL objectives through conversion to surrogate rewards. Still, these approaches face practical challenges such as lacking convergence guarantees and high computation complexity. This talk revisits the foundations of RL for LTL objectives on Markov decision processes and introduces recent advances in analyzing the surrogate rewards and thoughts on future research directions.

Bio: Yu Wang is an Assistant Professor in the Department of Mechanical and Aerospace Engineering at the University of Florida (UF) and the Group Lead in Autonomous and Connected Vehicles at UF Transportation Institute. He was a Postdoctoral Associate at the Department of Electrical and Computer Engineering at Duke University. He received his Ph.D. degree in Mechanical Engineering and M.S. in Statistics and Mathematics from the University of Illinois at Urbana-Champaign (UIUC). His research focuses on autonomy, cyber-physical systems, formal methods, and machine learning.

***Event will take place in hybrid format. The location for in-person attendance will be room 1500 EECS.   Attendance will also be possible via Zoom. The Zoom link and password will be distributed to the Controls Group e-mail list-serv.

To join this list-serv, please send an (empty) email message to [email protected] with the word “subscribe” in the subject line. To cancel your subscription, send an (empty) email to  [email protected] with the word “unsubscribe” in the subject line.  Zoom information is also available upon request to Shelly Feldkamp at [email protected].