I successfully defended my thesis on August 12, 2005. My dissertation deals with the problem of decentralized decision making in a team of robots with limited communication. These types of problems can be modeled as partially observable stochastic games (POSGs) with common payoffs; however, the optimal solution to such games is computationally intractable. The Bayesian Game approximation algorithm (BaGA), is a method for finding approximate solutions to POSGs by interleaving planning and execution. Domains covered in my dissertation include multi-robot herding problems and variations of laser-tag as well as the more canonical Lady and the Tiger and Multiple Access Broadcast Chanel

An abstract of my dissertation, Game-Theoretic Control for Robot Teams, can be found here.

My thesis advisors were Sebastian Thrun and Jeff Schneider. For my first two years at CMU I was advised by Tucker Balch who is now at Georgia Tech. With Tucker, I participated in RoboCup 2000 and 2001.

CONTACT INFORMATION

Address:
Computer Science Department
Stanford University
353 Serra Mall
Gates Building 116
Stanford, CA 94305

Work Phone: (650)725-8790

Email:remery [at] cs [dot] cmu [dot] edu