Brain-inspired computing technology has entered an exciting new era thanks to a breakthrough from researchers at the Massachusetts Institute of Technology (MIT). Specifically, the team built a new computing platform that leverages nanoscale mechanics to mimic the firing behavior of human brain cells (neurons). Published in the journal Science Advances, Prof. Farnaz Niroui led this study alongside an interdisciplinary team from MIT.

In addition, the system integrates data processing, memory, and sensing capabilities within a single micro-device. Therefore, this brain-inspired computing innovation holds great potential for accelerating the development of edge artificial intelligence (edge AI), smart robotics, and low-power medical implants. Full details about this research appear on the Official MIT News Release.

1. Inspiration from Octopus Arms

Traditionally, computing systems rely on conventional architectures that separate the central processing unit (CPU) from memory storage. Consequently, this separation causes significant energy waste because data constantly travels back and forth.

However, biological systems in nature operate quite differently. An octopus, for example, features a distributed nervous system where roughly two-thirds of its neurons reside along its arms. As a result, the arms can sense the environment, store information, and react locally without waiting for instructions from a central brain. MIT’s brain-inspired computing platform applies this exact principle of distributed computing matter at the nanometer scale.

2. A “Nano-Spring” Solution to Adhesive Forces

Building mechanical devices at the nanoscale poses major physical challenges. Specifically, when two metal surfaces come extremely close together at the nanometer level, strong interatomic adhesive forces pull them together permanently (crashing).

To overcome this fundamental challenge, the researchers designed a specialized sandwich structure:

• Metal Electrodes: Top and bottom conductive layers serve as electrodes.
• PDMS Layer (Polydimethylsiloxane): A viscoelastic polymer layer measuring just 2 nanometers in thickness sits between the two electrodes.

Furthermore, the PDMS polymer acts as a “nano-spring” that balances these adhesive forces. When users apply an electrical voltage, the two electrodes compress the polymer layer and alter the electrical current flowing through it. Because PDMS is viscoelastic, it naturally remembers the history of applied voltages as it gradually relaxes back to its original shape.

3. Mimicking Neuron Behavior (Artificial Neuron)

In a biological neural network, neurons accumulate electrical charge step by step until they reach a threshold. After reaching that threshold, they fire an electrical impulse to pass information onward.

This nanoscale brain-inspired computing system mirrors that exact dynamic with high precision:

• Accumulation: Applied voltage gradually compresses the PDMS polymer over time.
• Firing: Once compression crosses a specific threshold, the electrical current spikes rapidly.
• Relaxation: The device gradually relaxes back to its initial state.

Moreover, because physical material properties directly handle both computing and memory functions, the system completely eliminates the need for external components like capacitors or complex auxiliary circuitry.

4. Potential Future Applications

Consequently, this platform opens up vast opportunities for next-generation technology:

• Smart Medical Devices & Prosthetics: Tactile sensors in artificial limbs can process pressure responses instantly without delay.
• Wearable Health Patches: Ultra-low-power skin patches analyze vital health indicators in real time.
• Robotics & Edge Computing: IoT sensors and mini-robots can process data locally without relying on cloud connections or draining heavy battery power.