Software bugs are increasingly the cause of catastrophic failures in automobiles and medical devices. The complexities arising from the interactions between a discrete software system and its continuous physical environment serve to defeat many testing and verification processes that are currently employed by the industry. A key challenge lies in the systematic choice of test inputs that can reveal the presence of defects in the system.

In this talk, we utilize optimization techniques to automatically steer the testing process towards likely violations. Our approach embeds the search for bad behaviors inside a numerical optimization scheme using a fitness function that measures how close or far away a system is from satisfying or violating a property expressed in temporal logic. Our framework can employ many off-the-shelf optimization algorithms that can guide the testing process. We illustrate the effectiveness of our approach through case studies involving the analysis of power train controllers in automobiles and human interactions with insulin infusion pumps. We conclude by discussing the grand challenge of modeling and analyzing human-machine interactions in cyber-physical systems.

Dr. Sriram Sankaranarayanan is an assistant professor of Computer Science at the University of Colorado, Boulder. His research interests include automatic techniques for reasoning about the behavior of computer and cyber-physical systems. He obtained a PhD in 2005 from Stanford University. Subsequently, he worked as a research staff member at NEC research labs in Princeton, NJ. He has been on the faculty at CU Boulder since 2009. Dr. Sankaranarayanan has been the recipient of awards including the Siebel Scholarship (2005), the NSF CAREER award (2009) and the Dean's award for outstanding junior faculty for the College of Engineering at CU Boulder (2012).