The Neuromechanics of Soft-bodied Locomotion in Caterpillars and Robots

Human-built machines are usually efficient, fast and powerful, and are largely constructed from stiff materials. In order to cope with complex environments, the most recent robotic devices have begun to incorporate compliant joints and control systems based on impedance rather than force and position monitoring. However, even these advanced machines cannot perform with the robustness and adaptability found in living animals. A major challenge is that the design, fabrication and control of highly deformable structures is still poorly understood. Our research is directed at understanding how the movements of soft animals are controlled and in applying these findings to the development of soft robots. The guiding framework is that morphological computation (embodied intelligence) is essential for soft animals to move and manipulate in the natural world, and that soft machines need to incorporate the same strategies.

We have used the caterpillar as a tractable model system to understand how neural commands and nonlinear material properties interact to create useful movements. Some of these concepts have been implemented in a family of simple elastomeric robots (Softworms) that can move with a variety of caterpillar-like gaits. The next challenge is to make these robots climb in complex branched structures.