Event Details


Center of Autonomy and AI, Center for Cyber-Physical Systems and the Internet of Things, and Ming Hsieh Institute Seminar Series

Wed, Sep 28, 2022
2:00 PM - 3:00 PM
Location: EEB 132
Speaker: Giulia Pedrielli, School of Computing and Augmented Intelligence (SCAI) at Arizona State University.

Talk Title: Going Inside the Box: Bayesian Optimization for Verification of Cyber Physical Systems with Varying Levels of System Knowledge

Series: Center for Cyber-Physical Systems and Internet of Things

Abstract: Systems across automotive, bio-pharma, aerospace, energy, have become increasingly complex, and simulation represents a standard tool to evaluate their performance independently from the purpose of the analysis being optimization, control, certification. As a result, black-box optimization, that can embed simulation to perform a wide range of analyses, has attracted a lot of attention from the science and engineering communities. This talk centers around Black-box optimization methods, focusing on random search approaches (such randomness is injected in the search independently from the problem being affected by noise) in the broad area of verification of Cyber Physical Systems. In this context, the problem of falsifying properties is translated into the minimization of a robustness function. This is a metric function that quantifies how far a CPS execution is from violating a property of interest.

We first focus on control and acceleration of the explore/exploit process for the falsification of safety requirements without exploiting any property of the system under analysis. Our approach alternates local and global search using local knowledge while exploring the space of possible solutions. The performance of the proposed approach is analyzed, and key future directions are discussed in the context of Cyberphysical systems safety evaluation.

In the second part of the talk, we present algorithms developed in the scope of certification of safety critical systems that in some form exploit some structure of the problem at hand. Part-X is a family of partitioning informed Bayesian optimizers that can identify regions in which the system can present safety concerns (bugs in the case a software is analyzed). In this sense, the algorithm learns structure of the robustness function used to find falsification. We also produce a global estimate of the falsification volume. The algorithm min-BO works to identify faults in systems that have complex requirements that can be decomposed into a set of simpler requirements that need to be simultaneously satisfied by the system (conjunctive requirements). Finally, we show the basic ideas behind the design of algorithms that can exploit, when available, instrumented source code for the CPS to verify.

Biography: Giulia Pedrielli (https://www.gpedriel.com/ ) is currently Associate Professor for the School of Computing and Augmented Intelligence (SCAI) at Arizona State University. She graduated from the Department of Mechanical Engineering of Politecnico di Milano. Giulia develops her research in design and analysis of random algorithms for global optimization, with focus on improving finite time performance and scalability of these approaches. Her work is motivated by design and control of next generation manufacturing systems in bio-pharma and aerospace applications, as well as problems in the design and evaluation of complex molecular structures in life-science. Applications of her work are in individualized cancer care, bio-manufacturing, design and control of self-assembled RNA structures, verification of Cyberphysical systems. Her research is funded by the NSF, DHS, DARPA, Intel, Lockheed Martin.

Host: Pierluigi Nuzzo, nuzzo@usc.edu

Webcast: https://usc.zoom.us/j/98083929768?pwd=SUJreHk0N0ZXbk5QZ1ZPUkRlM3FmZz09