The Polymer Society of Korea

Abstract

Although advanced robots mimic human aesthetics, they lack the ability to adapt or evolve in response to external experiences. To address this, we propose an innovative control system using parallel-processable, retention-engineered synaptic devices that simulates human-like learning without complex computational hardware. Synaptic retention was modulated by adjusting the density of sprayed Ag/AgCl ink, which altered the voltage drop at the gate-electrode/electrolyte interface. Furthermore, the unrestricted movement of ions enhanced the signal multiplexing of the ion gel, enabling device-level parallel processing. Integrating these synaptic devices with actuators allowed us to emulate human-like workout processes, including feedback between acute and chronic responses. This approach provides a robust framework for reducing system complexity while achieving sophisticated learning in biomimicry.