A Programmable Substrate to Study Robots Jumping From Non-Rigid Surfaces

Sathvik Divi, Justin K. Yim, Mark Bedillion, Sarah Bergbreiter

Abstract: This study presents the development, characterization, and demonstration of a tunable substrate for small jumping robots. Jumping robots in the literature are typically evaluated when jumping from rigid surfaces, in contrast to surfaces with more significant compliance or damping that are encountered in the natural world. The aim of this work is to create a physical substrate, or ’ground', for which the effective mass, compliance, and damping can be programmed. This system enables quick testing of various substrate conditions and also allows for the introduction of complex nonlinearities to analyze the interactions between latch-mediated spring actuation (LaMSA) systems and their environment. A mathematical model for the substrate is defined and the system is built with a fast brushless DC motor and controller running on a real-time target machine. The results illustrate the range of compliance and damping that can be achieved, as well as example jumps from the substrate using a 4 g jumper and a 108 g jumping robot.