Claire Tomlin, University of California, Berkeley

EECS Prof. Claire Tomlin holds of the Charles A. Desoer Chair in Engineering. Her research interests include hybrid systems, distributed and decentralized optimization, and control theory, with an emphasis on applications, unmanned aerial vehicles, air traffic control and modeling of biological processes. She taught at Stanford University from 1998 to 2007 where she was a director of the Hybrid Systems Laboratory and held joint positions the Department of Aeronautics and Astronautics and the Department of Electrical Engineering. She was awarded a MacArthur Genius grant in 2006 and the IEEE Transportation Technologies Award in 2017 "for contributions to air transportation systems, focusing on collision avoidance protocol design and avionics safety verification"

A great deal of research in recent years has focused on robot learning. In many applications, guarantees that specifications are satisfied throughout the learning process are paramount. For the safety specification, we present a controller synthesis technique based on the computation of reachable sets, using optimal control and game theory. In the first part of the talk, we will review these methods and their application to collision avoidance and avionics design in air traffic management systems, and networks of unmanned aerial vehicles. In the second part, we will present a toolbox of methods combining reachability with data-driven techniques inspired by machine learning, to enable performance improvement while maintaining safety. We will illustrate these "safe learning" methods on robotic platforms at Berkeley, including demonstrations of motion planning around people, and navigating in a priori unknown environments.

Paulo Tabuada, University of California, Los Angeles

Paulo Tabuada (Fellow, IEEE) was born in Lisbon, Portugal, one year after the Carnation Revolution. He received the "Licenciatura" degree in aerospace engineering from the Instituto Superior Tecnico, Lisbon, Portugal, in 1998 and the Ph.D. degree in electrical and computer engineering from the Institute for Systems and Robotics, a private research institute associated with Instituto Superior Tecnico, in 2002. Between January 2002 and July 2003, he was a Postdoctoral Researcher at the University of Pennsylvania. After spending three years at the University of Notre Dame, as an Assistant Professor, he joined the Electrical and Computer Engineering Department, University of California at Los Angeles, Los Angeles, CA, USA, where he currently is the Vijay K. Dhir Professor of Engineering. Dr. Tabuada received multiple awards including the NSF CAREER award in 2005, the Donald P. Eckman award in 2009, the George S. Axelby award in 2011, and the Antonio Ruberti Prize in 2015. In 2009, he co-chaired the International Conference Hybrid Systems: Computation and Control (HSCC'09) and joined its steering committee in 2015; in 2012, he was Program Co-Chair for the 3rd IFAC Workshop on Distributed Estimation and Control in Networked Systems (NecSys'12); in 2015, he was Program Co-Chair for the IFAC Conference on Analysis and Design of Hybrid Systems; and in 2018, he was Program Co-Chair for the International Conference on Cyber-Physical Systems (ICCPS'18). He also served on the editorial board of IEEE Embedded Systems Letters and the IEEE Transactions on Automatic Control.

Deep neural networks have drastically changed the landscape of several engineering areas such as computer vision and natural language processing. Notwithstanding the widespread success of deep networks in these, and many other areas, it is still not well understood why deep neural networks work so well. In particular, the question of which functions can be learned by deep neural networks has remained unanswered. In this talk we give an answer to this question for deep residual neural networks, a class of deep networks that can be interpreted as the time discretization of nonlinear control systems. We will show that the ability of these networks to memorize training data can be expressed through the control theoretic notion of controllability which can be proved using geometric control techniques. We then add an additional ingredient, monotonicity, to conclude that deep residual networks can approximate, to arbitrary accuracy with respect to the uniform norm, any continuous function on a compact subset of n-dimensional Euclidean space by using at most n+1 neurons per layer. We will conclude the talk by showing how these results pave the way for the use of deep networks in the perception pipeline of autonomous systems while providing formal (and probability free) guarantees of stability and robustness.

Verena Wolf, Saarland University

Verena Wolf received the diploma degree in computer science from the University Bonn, in 2003 and the PhD degree from the University Mannheim, in 2008. She is a full professor with Saarland University since 2012 and leads the Modelling and Simulation Group at the Department of Computer Science. She is currently working on discrete stochastic modelling as well as efficient simulation methods and has been on the program committees of more than 50 international conferences.

In my talk, I will consider several probabilistic inference problems for a class of population-structured Markov jump processes, which capture stochastic interactions between groups of identical agents. The computation of quantities such as bridging and rare event probabilities or mean first passage times is notoriously challenging. Typically neither state-based numerical approaches nor methods based on stochastic sampling give efficient and accurate results. I will show that adaptations of popular approaches such as state-space lumping and truncation as well as moment-based analysis using semi-definite programming can give accurate approximations. In particular, we will consider methods for obtaining guaranteed bounds on the quantities of interest and apply them to complex models from biology and epidemiology.