Human-Inspired Stretch and Joint-Bend Sensing System Based on Flexible Sensors

Bio-mimicking of human tactile sensing system using flexible and reliable sensors and electrical circuits for developing future robots and prosthetics with better sensory capabilities has gained importance in recent times. In addition, development of artificial proprioception systems capable of locating, monitoring and controlling different parts and joints in robots is also vital. In this work, using a highly flexible, reliable, and easy to fabricate ecoflex-graphite-graphene carbon paste (GCP) based piezoresistive strain sensor, a Wheatstone bridge and a simulated leaky-integrate-and-fire (LIF) neuron, an artificial tactile system was developed which also includes the proprioception functionalities. The developed system generated neural spike trains when the sensor was strained, and the firing rates of the presented system varied between 7-26 spikes per min while increasing the strain levels from 5% to 50%. The spiking behaviour resembles an afferent neuron connected to a slowly-adapting mechanoreceptor. The relationships between firing rate with different stretch and bending angle in the sensor were also calculated to develop an artificial proprioception system. The presented work could pave the way for making next generation smart robots, prosthetics, gloves, and tactile sensors.