Current Institution: University of Washington
Bio: Sarah H.Q. Li is a Ph.D. candidate in Aeronautics and Astronautics Engineering department at the University of Washington, Autonomous Control Laboratory. She received her B.A.Sc. in Engineering Physics with minor in Honors Mathematics from the University of British Columbia. Inspired by large-scale autonomy settings including autonomous driving and urban air mobility, her research combines game theory, optimization, and learning to design safe and minimally congested networks of competitive autonomous agents. She is a 2020 Amelia Earhart Fellow and was a research intern at ONERA, the French Aerospace Lab, Loon, École Nationale de l’Aviation Civile, and Microsoft Research .
Abstract: Architecting Co-existence: Scalable Integration of Autonomy in Shared Spaces
Inspired by large-scale autonomous applications, my research aims to design multi-agent decision-making systems for uncertain environments. Due to physical space limitations and surging customer usage, large-scale autonomous systems are operating in increasingly congested environments with dynamically changing demands. In the near future, large-scale autonomy that does not address real-time resource constraints and demands will result in inefficient and congested operations. On the other hand, novel applications exemplified by urban air mobility and autonomous driving are unparalleled opportunities to reform real-time large-scale autonomy. Towards this goal, my research leverages techniques from control, game theory, and optimization to address key obstacles in this domain. Specifically, how to synthesize scalable decision-making algorithms that accommodate fluctuating supply and demand, design system infrastructure that supports competitive large-scale autonomy, and generate multi-agent learning networks that defend against adversarial attacks.