A central challenge in the design of cyberphysical systems arises from the coupling between computational and mechanical components. In this talk, I focus on the dynamical interaction that occurs when a robot or animal moves through its environment. It has proven difficult to engineer such neuromechanical systems despite the fact that organisms provide proof-of-concept designs in many situations where robots struggle. Studying the piecewise-defined (“hybrid”) dynamics governing locomotion and manipulation, I prove that models of periodic behaviors generically exhibit reduction in the number of mechanical degrees-of-freedom due solely to the interaction between the body and the environment. Furthermore, I exploit the structure of the reduced-order dynamics to derive a scalable algorithm for parameter identification from motion capture data and apply this technique to perturbation recovery in running cockroaches. Finally, I combine these analytical, computational, and experimental tools to propose a foundation for systematically engineering neuromechanics.
Sam Burden earned his BS with Honors in Electrical Engineering from the University of Washington, Seattle. Currently, he is a PhD candidate in Electrical Engineering and Computer Sciences at the University of California, Berkeley and expects to graduate in May of 2014. He is broadly interested in applying control and dynamical systems theory to study neuromechanical and cyberphysical systems. Specifically, he focuses on discovering and formalizing principles that enable dynamic locomotion and dexterous manipulation in robotics, biomechanics, and human motor control. He is a recipient of the NSF Graduate Research Fellowship and collaborator in the ARL Micro Autonomous Systems and Technology CTA. In his spare time, he teaches robotics to students of all ages in K-12 classrooms, Maker Fairs, and campus events.