Touch Localization with a Resistive Position Sensor-Gated Oxide Neuromorphic Transistor

Abstract

Abstract The ability to localize external mechanical stimuli is fundamental to human tactile perception, making its hardware implementation of significant interest for advanced human–machine interaction. Here, a resistive position sensor (RPS) is proposed for touch location recognition, exhibiting a high sensitivity of ∼55 kΩ/cm and excellent stability. A tactile perceptual system (TPS) is proposed by integrating the RPS with a chitosan-gated oxide ionotronic neuromorphic transistor. Benefiting from the strong proton relaxation effects of the chitosan-based electrolyte, the neuromorphic transistor can effectively process electrical signals generated from the RPS. The TPS can emulate various synaptic behaviors under mechanical stimuli. As a proof of concept, it demonstrates Braille code recognition. Furthermore, the TPS enables touch intention recognition through machine learning, demonstrating a high recognition accuracy of ∼99.2%. The results demonstrate great promise in human−machine interaction.