Perspectives on Multi-Impact Robotics

Impacts between robot and world, integral to dynamic walking, running, and high-speed manipulation, are notoriously difficult to predict or control. Simultaneous contact, for instance between two feet and the ground, is particularly challenging to manage. In this talk, I will discuss our recent and ongoing efforts to better understand and leverage the effects of multiple impacts. First, calling back to David Remy's 2016 Dynamic Walking presentation, I will show the conclusion of our work to model the non-unique outcomes that arise during simultaneous contact, culminating in a single differential inclusion for the continuous and discrete dynamics. Next, I will discuss an approach to control during these impact events. To deal with the uncertainty prevalent when striking the ground, common approaches use heuristics to blend control strategies or reduce gains. Here, based on a representation of uncertainty, we develop an impact-invariant strategy that reduces sensitivity to impacts while maintaining some control authority and demonstrate this strategy for jumping with the Cassie robot. I'll conclude the talk with highlights of our work on ContactNets, where we leverage physical models of discontinuity to efficiently learn contact dynamics.

Dr. Michael Posa