Humans continue to be an indispensable component of many systems of interest in spite of great advancements in automation technologies. Typical examples are supervisory control for remote unmanned systems and production systems where humans provide critical higher level decision support, and societal systems like traffic networks and smart grids where human decisions play a defining role in the overall performance of the system. This talk will present novel control policies to extract maximum efficiency from such human-centered systems. In particular, two illustrative scenarios will be discussed: task management policies that maximize productivity of human operators in a queueing setup, and route choice decisions of drivers that render maximum resilience to traffic networks. The results rely on novel tools and concepts at the intersection of dynamical systems, cognitive psychology, game theory and optimization, and provide key insights into the fundamental performance limits of such systems.
Sunday, April 10, 2011
Free and open to the public