Design Strategies for Controlling Motion in Soft Robotic Systems

Abstract

The field of soft robotics has attracted much interest due to its promise of machines that are cheaper, safer, and more robust than their rigid counterparts. Despite rapid progress, there is an open question of how best to control these new types of robots. Applying control techniques from traditional robotics is far from trivial, largely due to the fundamentally different body plans of rigid and soft robots, and the highly nonlinear mechanics of elastomeric materials. This thesis proposes a holistic approach to control systems for soft robots, in which the controller is designed in parallel with robot morphology, actuation strategy, and overall system requirements. We distinguish between passive control strategies, in which some form of morphological computation is employed, and active control strategies, in which a user (or the environment) provides input. In the first part of this thesis we examine passive control. We present two novel fabrication paradigms that each establish a new class of soft robots, and show how fabrication and control are tightly coupled in each case. In the second part, we discuss how previous work in microfluidic logic can be adapted and improved to make it a viable option for active control in soft robotics. We conclude by outlining future research directions, and propose a way in which both passive and active control may be combined into a single comprehensive control strategy.