Run-Time Enforcers in Adversarial and Information-Limited Environments
|
ABSTRACT Recent advances in formal verification and synthesis have expanded the scalability of the methods and the expressivity of the models that can be expressed by these methods. We investigate two of these methods in the context of autonomous operations in adversarial and information-limited environments. The first of these methods is on the synthesis of run-time enforcers of safety-critical properties. The second is on the synthesis of robust finite-state policies in so-called uncertain partially observable Markov decision processes (POMDPs). We will describe the state of the art for these methods and discuss their strengths and shortcomings along with demonstrating examples. Slides found here. |
BIO
Ufuk Topcu is an Associate Professor in the Department of Aerospace Engineering and Engineering Mechanics at The University of Texas at Austin. He is a core faculty member at the Oden Institute for Computational Engineering and Sciences. Ufuk obtained his Doctor of Philosophy degree from the University of California, Berkeley in 2008. Prior to joining The University of Texas at Austin, he was with the Department of Electrical and Systems Engineering at the University of Pennsylvania. He was a postdoctoral scholar at California Institute of Technology until 2012. Ufuk’s research focuses on the theoretical and algorithmic aspects of the design and verification of autonomous systems, typically in the intersection of formal methods, reinforcement learning, and control theory. He takes a relatively broad view on autonomy and tends to tackle abstract problems motivated by challenges cutting across multiple applications of autonomy. His research contributions have been recognized by the NSF CAREER Award, the Air Force Young Investigator Award, the IEEE CSS Antonio Ruberti Young Researcher Prize, and Oden Institute Distinguished Researcher Award. He is a member of the Computing Community Consortium Council.