Stress-induced artificial neuron spiking in diffusive memristors.

Area of Science:

• Materials Science and Engineering
• Neuroscience and Neuromorphic Computing

Background:

• Diffusive memristors exhibit current spiking, making them suitable for artificial electronic neurons.
• Artificial neurons are crucial for advanced autonomous and robotic systems, particularly for sensor applications like object grasping and classification.

Purpose of the Study:

• To engineer a flexible, nanoparticle-based diffusive memristor capable of producing controllable electrical spiking.
• To investigate the influence of external mechanical stimuli on the memristor's spiking behavior for tactile sensing applications.

Main Methods:

• Fabrication of a silver nanoparticle-based diffusive memristor on a flexible polyethylene terephthalate substrate.
• Induction and manipulation of electrical spiking behavior using applied voltage and external mechanical impacts (varying magnitude and frequency).
• Development of a mathematical model to describe the memristor's operational principle and spiking characteristics.

Main Results:

• Demonstrated controllable electrical spiking in the diffusive memristor triggered by mechanical impact.
• Showcased the ability to modulate spiking characteristics by adjusting the magnitude and frequency of mechanical stimuli.
• Validated a mathematical model that accurately captures the observed spiking behavior.

Conclusions:

• The developed flexible diffusive memristor functions as a touch-perception sensor, converting mechanical pressure into electrical spikes.
• This technology offers a pathway for creating advanced robotic systems with tactile sensing capabilities.
• The memristor's ability to translate tactile information directly into neural-compatible spikes presents significant benefits for robotics.