Abstract: The ongoing evolution of enabling autonomy in many application domains is being fueled by the availability of high-performance embedded hardware platforms and significant advances in artificial intelligence techniques. Unfortunately, the critical software certification processes have not kept pace with this evolution, which is rather unsettling for safety-critical domains such as transportation and medical robots. A fundamental challenge in certifying autonomous system software is due to the difficulty of validating predictability in system behaviors, which encompass not only functionality but reaction and execution properties as well. Unpredictable, non-deterministic, and even erroneous behaviors are being observed from application-level decision-making frameworks to low-level system software. In this talk, I will describe our contributions on (partially) resolving the predictability challenge through multi-faceted research involving embedded and real-time systems, operating systems, and machine learning-driven autonomy.

Short Bio: Cong Liu is currently an associate professor in the Department of Computer Science at the University of Texas at Dallas. His research focuses on developing predictable and robust system software for emerging autonomous systems. He has received multiple awards from both academia and industry, including several outstanding/best paper awards from top-tier conferences and industrial research awards, the NSF CAREER award, and a UT-Dallas outstanding teaching award.