A skin-like sensor for intelligent Braille recognition

Abstract A flexible skin-like tactile sensor as one of the key components in the next generation for robots should have the ability to perform real-time feedback, continuous measurement, and quantization of weak target signals in the real applications (e.g. intelligent Braille recognition). Herein, inspired by human skin, a novel flexible piezoresistive tactile sensor with high sensitivity and linearity is designed and fabricated. It is composed of three main parts including the random Gaussian distribution (RGD) spinosum on polydimethylsiloxane as the top substrate, the multilayer Ti2C-MXene film as the intermediate conductive filler, and the commercial polyimide-based interdigital electrodes as the bottom substrate. The working mechanism of RGD spinosum and Ti2C-MXene films in the flexible tactile sensor are clarified by COMSOL Multiphysics simulations and density functional theory calculations, respectively. The assembled sensor demonstrates ultrahigh sensitivity, large linearity, excellent cycling stability, and fast response speed in the milliseconds. Moreover, the development of advanced neural network technology makes it possible to realize intelligent skin-like sensor. In this study, it is the first time to realize real-time Braille effective intelligent recognition by such a flexible skin-like tactile sensor with the random decision forests algorithm. This study is of great significance to solve the social and public issues of information exchange for the vision-impaired or even the blind that has been highly concerned in recent years.