Soft Multi-point Waveguide Sensor for Proprioception and Extereoception in Inflatable Fingers

Abstract

Disadvantages of conventional robotic systems include rigidity, multiple moving parts, and the need for elaborate safety mechanisms when used in human-machine interaction. Soft manipulators and grippers are gaining in popularity due to being able to handle large payloads whilst being lightweight, highly compliant, low-cost, and compactible or collapsible. Yet soft robots cannot make use of traditional rigid sensors to measure their pose or interaction with the environment. Perception in soft robotics needs to embrace alternative methods: sensors made from soft materials that perform robustly under compression and bending conditions; i.e, stretchable soft sensors that rely on (their) material and electrical properties to output signal measurements. However, many such sensors come with inherent drawbacks, including material incompatibility, fabrication complexity, and hysteresis. In this paper, we report on the use of multiple staggered optical waveguide sensors embedded in silicone. These stretchable optical waveguide sensors coated with a thin layer of gold were fabricated and integrated with a fabric-based, inflatable robot finger. An experimental study was performed to evaluate the sensor's responsiveness. We find that multi-curvature pose estimation (from 0.05-0.135 m-1) (from fully deflated to maximum inflation) can be acquired after integration with the inflatable robot finger. The sensor proves capable of measuring force information by way of interaction with the environment at multiple points along the gripper.