New Approaches to the Modeling and Control of Complex Dynamics

Google Tech Talks October 16, 2008 ABSTRACT Complex phenomena such as animal morphology, human motion, and large fluid systems challenge even our most sophisticated simulation and control techniques. This talk presents some fundamentally new methods to approach such high dimensional and nonlinear problems, with applications to fluid dynamics, crowd simulation, human motion, animal morphology and protein folding. Speaker: Adrien Treuille I am an Assistant Professor in the computer graphics group at Carnegie Mellon University. I received my PhD under Zoran Popovic in the computer graphics group at the University of Washington, and was a postdoc in the Baker Group under Zoran Popovic and David Baker. I was one of the creators of Foldit, the computer game where users contribute to science by folding proteins. I also pursue research in the simulation and animation of very high-dimensional nonlinear phenomena like animal morphology, human motion, and large fluid systems. One thread of my research addresses the complexity of such systems by developing model reduction tools that generate compact representations. A complementary thread seeks to control such systems, which means learning to set inputs to produce desired effects. While I seek theoretical advances, I am also deeply interested in the implications for science and engineering of these techniques, from fluid dynamcs to laying down a joint cognitive and biomechanical basis for animal motion.