Electro-active polymer hydrogels exhibit emergent memory when embodied in a simulated game environment
Vincent Strong, William Holderbaum, Yoshikatsu Hayashi
Abstract
The goal of artificial neural networks is to utilize the functions of biological brains to develop computational algorithms. However, these purely artificial implementations cannot achieve the adaptive behavior found in biological neural networks (BNNs) via their inherent memory. Alternative computing mediums that integrate biological neurons with computer hardware have shown similar emergent behavior via memory, as found in BNNs. By applying current theories in BNNs, can emergent memory functions be achieved with alternative mediums? Electro-active polymer (EAP) hydrogels were embedded in the simulated game-world of Pong via custom multi-electrode arrays and feedback between motor commands and stimulation. Through performance analysis within the game environment, emergent memory acquisition was demonstrated, driven by ion migration through the hydrogels.