Abstract: Although human interaction with autonomous systems is becoming ubiquitous, few tools exist for planning and control of autonomous systems that account for human uncertainty and decision making. We seek methods for probabilistic verification and control that can help ensure compatibility of autonomous systems with human decision making and human uncertainty. This requires the development of theory and computational tools that can accommodate arbitrary, non-Gaussian uncertainty for both probabilistic verification and control, potentially without high confidence models. This talk will focus on our work in probabilistic verification of ReLU neural nets, data-driven stochastic optimal control and stochastic reachability, and recent work in optimal “blameless” controllers that can facilitate operation under infeasible constraints in an ethical manner. Our approaches to probabilistic verification are based in Fourier transforms and chance constrained optimization, and our approaches to data-driven stochastic planning and control are based in conditional distribution embeddings. Both of these approaches enable computation without gridding, sampling, or recursion.
Bio: Meeko Oishi received the Ph.D. (2004) and M.S. (2000) in Mechanical Engineering from Stanford University (Ph.D. minor, Electrical Engineering), and a B.S.E. in Mechanical Engineering from Princeton University (1998). She is a Professor of Electrical and Computer Engineering at the University of New Mexico. Her research interests include human-in-the-loop control, stochastic optimal control, and autonomous systems. She previously held a faculty position at the University of British Columbia at Vancouver, and postdoctoral positions at Sandia National Laboratories and at the National Ecological Observatory Network. She was a Visiting Researcher at AFRL Space Vehicles Directorate, and a Science and Technology Policy Fellow at The National Academies. She is the recipient of the NSF CAREER Award, the NSF BRITE Fellowship, the Truman Postdoctoral Fellowship in National Security Science and Engineering, and a member of the 2022-2024 US Defense Science Study Group.