Mechanoreception of Soft Robots via Multi-Channel Optical Sensing Using Ambient Light and Internal Infrared Light

摘要

Soft robots exhibit high deformability, enabling compliant interactions between the robots and unstructured environments. Mechanoreception of the robot state and its surroundings is essential for feedback control and decision making in these interactive scenarios, while remaining in significant challenges owing to the complexity of deformation and environment. In this paper, we introduce a multi-channel optical sensing approach using ambient light and internal infrared light to detect the contact force and contact position of soft robots. We demonstrate that it is possible to simultaneously measure multimodal physical quantities (external contact position, and interactive forces) with a concise optical waveguides design. This concept is verified in a soft multi-channel optical sensor design that embeds optical sensing units for receiving both environmental visible light and infrared light. The entire methodology from the core sensing principle to conceptual design, modeling, and experiments is developed and demonstrated in a soft robotic gripper capable of providing interaction sensory information, such as object detection, bending angles, and force measurements. This approach may inspire new soft sensor and robot designs to obtain multimodal sensory information for robotic applications.