Fabrication, Sensation and Control of Fluidic Elastomer Actuators and Their Application towards Hand Orthotics and Prosthetics

Due to their continuous and natural motion, fluidic elastomer actuators (FEAs) have shown potential in a range of robotic applications including prosthetics and orthotics. Despite their advantages and rapid developments, robots using these actuators still have several challenging issues to be addressed. First, the reliable production of low cost and complex actuators that can apply high forces is necessary, yet none of existing fabrication methods are both easy to implement and of high force output. Next, compliant or stretchable sensors that can be embedded into their bodies for sophisticated functions are required, however, many of these sensors suffer from hysteresis, fabrication complexity, chemical safety and environmental instability, and material incompatibility with soft actuators. Finally, feedback control for FEAs is necessary to achieve better performance, but most soft robots are still “open-loop”. In this dissertation, I intend to help solve the above issues and drive the applications of soft robotics towards hand orthotics and prosthetics. First, I adapt rotational casting as a new manufacturing method for soft actuators. I present a cuboid soft actuator that can generate a force of >25 N at its tip, a near ten-fold increase over similar actuators previously reported. Next, I propose a soft orthotic finger with position control enabled via embedded optical fiber. I monitor both the static and dynamic states via the optical sensor and achieve the prescribed curvatures accurately and with stability by a gain-scheduled proportional-integral-derivative controller. Then I develop the soft orthotic fingers into a low-cost, closed-loop controlled, soft orthotic glove that can be worn by a typical human hand and helpful for grasping light objects, while also providing finger position control. I achieve motion control with inexpensive, binary pneumatic switches controlled by a simple finite-state-machine. Finally, I report the first use of stretchable optical waveguides for strain sensing in a soft prosthetic hand. These optoelectronic strain sensors are easy to fabricate, chemically inert, and demonstrate low hysteresis and high precision in their output signals. I use the optoelectronically innervated prosthetic hand to conduct various active sensation experiments inspired by the capabilities of a real hand