Resilient, Adaptive, and Scalably Complex Soft Tensegrity Robots

Abstract

Completely soft robots are emerging as a compelling new platform for exploring and operating in unstructured, rugged, and dynamic environments. In principle they promise a level of mobility, resiliency, and configurability matched only by biological organisms. Unfortunately, the very properties which make soft robots so appealing also make them difficult to accurately model, scalably design, and robustly control. This talk will introduce tensegrity-based robots as a compelling low-cost entry-level platform with which to explore the fundamental soft robotic challenges of design and control. Despite their relative simplicity, tensegrities are an effective substrate for soft robotics research: they are continuum structures with a high dimensional configuration space; their pre-stress results in high compliance, and they are highly configurable -- capable of rapid shape change and tunable stiffness. This talk will show how we can employ state of the art machine learning techniques to discover robust and dynamic behaviors which exploit, rather than suppress, complex soft body dynamics. I will also show how to leverage automated design tools in to invent self-assembling soft robots which can scale in size and complexity.

Bio

John Rieffel is an Associate Professor in the Computer Science Department at Union College in Schenectady, NY. His research expertise is in soft robotics and genetic algorithms. He received in undergraduate degrees in Engineering (BS) and Computer Science (BA) from Swarthmore College, and his Ph.D. in Computer Science from Brandeis University. He has been a postdoctoral associate in Hod Lipson's lab at Cornell University, and in Barry Trimmer's lab at Tufts University.