One-dimensional organic artificial multi-synapses enabling electronic textile neural network for wearable...

Area of Science:

• Materials Science
• Neuroscience
• Electrical Engineering

Background:

• One-dimensional (1D) devices are crucial for wearable electronic technology due to their flexibility and integration into textiles.
• Maintaining device functionality under mechanical stress is a key challenge for e-textiles.

Purpose of the Study:

• To design and fabricate 1D fiber-shaped multi-synapses using ferroelectric organic transistors.
• To evaluate their performance as multisynaptic channels in an e-textile neural network for wearable neuromorphic applications.

Main Methods:

• Fabrication of 1D multi-synapses on a 100-μm silver (Ag) wire.
• Integration into NOR-type textile arrays for neural network formation.
• Testing device reliability under repeated stimuli and mechanical bending stress.

Main Results:

• The 1D multi-synapses exhibited diverse synaptic functions with high reliability (6000 stimuli).
• Textile arrays demonstrated integrated signal propagation without leakage.
• Achieved ~90% and ~70% recognition accuracy for MNIST and ECG patterns, respectively, in a single-layer network.

Conclusions:

• 1D fiber-shaped multi-synapses are suitable for wearable neuromorphic applications.
• The developed e-textile neural network shows robust performance under mechanical stress.
• This technology enables advanced functionalities in electronic textiles.