Soft and Stretchable Optical Fibers with Gradient Color Coding for Multipoint Bending and Tactile Perception in Dexterous Hands

Abstract

Robotic tactile sensing technology is crucial for the advancement of intelligent humanoid robotics. Recently, optical fiber-based tactile sensors have attracted significant attention, leading to rapid developments in the field. Inspired by the spatial distribution and multipoint sensing capabilities of tactile mechanoreceptors in human skin, we introduce a stretchable and flexible optical fiber sensor with gradient-colored segments embedded within its core alongside a fabrication method. Our findings demonstrate that the optical loss coefficients of absorbing bands in flexible optical fiber segments, which share the same color but have different doping concentrations, vary under bending or pressure, while nonabsorbing bands remain stable. Leveraging this property, we propose a multipoint stress and pressure decoupling method utilizing gradient color coding and multiwavelength referencing. We have successfully integrated this soft fiber optic sensor onto the fingers and back of robotic hands, enabling the precise measurement of bending angles, finger joint positions, and pressure localization on the robotic hand. The proposed sensor offers high design flexibility, ease of fabrication, and exceptional tensile performance, positioning it as a promising solution for applications in human-computer interaction and intelligent robotics.