POS5-1388
Long-retention wavelength-tunable tactile neuromorphic display for vision-assisted robotic manipulation
When and Where
Nov 30, -0001
00:00 - 00:00
Presenter(s)
Woojoong Kim (Yonsei university)
Co-Author(s)
Abstract
Tactile-interactive neuromorphic displays with self-learning capabilities are critical for next-generation intelligent sensor systems that require adaptive perception and information encoding in real time. Here we present a tactile neuromorphic structural color display that simultaneously senses, stores, learns and visualizes external tactile stimuli within a single-device architecture. The platform integrates a periodically ordered block copolymer photonic crystal as a wavelength-tunable gate dielectric with an ion-gel-gated synaptic transistor featuring a tactile-interactive gate electrode. Tactile inputs modulate the gate electric field, triggering redistribution and electrochemical doping of non-volatile ions that concurrently reconfigure channel conductance and the lamellar periodicity of the block copolymer photonic crystal. The device exhibited synaptic plasticity with a reversible, non-volatile wavelength-encoded optical state at low power (~6 mW), reliable operations beyond 5,000 cycles, and long-term stability over 1 month. We further demonstrate a vision-assisted robotic gripper generating object-specific structural-color images for recognition of the shape as well as mechanical properties of the objects
