Softness- and Pressure-Perceptive Electronic Skin with Reservoir-Computed Central Nervous System

Simultaneous detection of softness and tactile pressure emulates the ability of human skin to discern objects and their texture. This functionality is crucial for developing human-interactive robots. Although electronic skin (e-skin) devices have been designed to sense tactile pressure and temperature via flexible or stretchable films, only a few sensors with a complicated structure simultaneously measure softness with other stimuli. This study presents a reservoir-computed, planar-type pressure sensor-integrated system, aiming to develop a simple, automated e-skin capable of perceiving softness and pressure. Six tactile pressure sensors are integrated into a single horizontal row. When mounted on a curved platform and robotic hand, the sensor's contact area varies with the object's softness, altering the pressure distribution. An echo state network algorithm-a reservoir computing platform-is developed to analyze time-series resistance data from the six sensors, enabling simultaneous softness and pressure prediction with high accuracy, analogous to the central nervous system. As a proof-of-concept, the sensor system is installed on a robotic hand, which demonstrates its ability to recognize softness and pressure. Owing to its simplicity, the sensor platform is adaptable to various tactile pressure sensors reported previously.