Ming Cao - Decision-Making Dynamics in Mixed Teams of Humans and Robots

Profitable integration of human and robot decision-making dynamics should take advantage of strengths of human decision-makers as well as strengths of robotic agents.   A major challenge in achieving this goal is understanding how humans make decisions and what are their associated strengths and weaknesses. Correspondingly, a central tenet of this work is to leverage the experimental and modeling work of psychologists and behavioral scientists on human decision-making. We focus on a well-studied class of sequential binary decision-making tasks. We introduce a decision-making problem associated with a collective robotic foraging task that integrates human and robotic decision making dynamics with feedback. To explore the integrated decision dynamics, we present two models of human decision-making and with these models we prove convergence of the human behavior to the observed aggregate decision-making. We also show how adaptive laws for the robot feedback that use only local information can be applied to help the human make optimal decisions.