Fast and Flexible Multiagent Decision-Making
Naomi Ehrich Leonard
Mechanical and Aerospace Engineering
Friday, April 7, 2023
Light Engineering 250
Abstract: AI will present new theory and methodology for understanding and designing fast and flexible decision-making behavior for a group of agents that observe or communicate over a network. Our model-free theory shows how agreement and disagreement behaviors that capture real-world multiagent decision-making emerge through a bifurcation (transition) in which indecision is destabilized. To realize and study these behaviors, we define analytically tractable dynamics that are equivalent to classic linear opinion dynamics with a saturation applied to the exchanges of opinion states. We prove the role of network structure in the transition point, the post-transition opinion patterns, and the sensitivity of the transition to the distribution of inputs over the network. With the addition of state-feedback dynamics for model parameters, the model admits tunability of the network behavior in the face of changing external conditions. Our results provide new means for systematic study and design of dynamics on networks in nature and technology, including the dynamics of multiagent decision-making, critical transitions to oscillatory behavior, spreading processes, polarization, games, navigation, and task allocation. I will demonstrate with applications to multi-robot teams.
This is joint work with Anastasia Bizyaeva and Alessio Franci and based on the paper
Nonlinear Opinion Dynamics With Tunable Sensitivity with reference to other key papers with additional collaborators, including: Tuning Cooperative Behavior in Games With Nonlinear Opinion Dynamics and Breaking indecision in multi-agent, multi-option dynamics.
Bio: Naomi Ehrich Leonard is Edwin S. Wilsey Professor of Mechanical and Aerospace Engineering and associated faculty in Applied and Computational Mathematics at Princeton University. She is also Director of Princeton’s Council on Science and Technology and Founding Editor of the Annual Review of Control, Robotics, and Autonomous Systems. She received her BSE in Mechanical Engineering from Princeton University and her PhD in Electrical Engineering from the University of Maryland. She is a MacArthur Fellow, elected member of the American Academy of Arts and Sciences, and winner of the 2023 IEEE Control Systems Award. Leonard is Fellow of SIAM, IEEE, IFAC, and ASME. Her current research focuses on dynamics, control, and learning for multi-agent systems on networks with application to multi-robot teams, collective animal behavior, and other networked systems in nature, technology, and the arts.
Previous Distinguished Lectures in ECE