# Computing with Networks of Nonlinear Mechanical Oscillators

## TL;DR

This paper introduces a mechanical network of nonlinear oscillators that mimics neural networks to efficiently solve complex computational problems, combining sensing and computing in a compact, power-efficient device.

## Contribution

It presents a novel mechanical oscillator network that performs neural network-like computations, integrating sensing and processing in a single physical system.

## Key findings

- Successfully solved parity and spoken word classification tasks
- Demonstrated high parallelism and energy efficiency
- Proposed a new hardware paradigm for neuromorphic computing

## Abstract

As it is getting increasingly difficult to achieve gains in the density and power efficiency of microelectronic computing devices because of lithographic techniques reaching fundamental physical limits, new approaches are required to maximize the benefits of distributed sensors, micro-robots or smart materials. Biologically-inspired devices, such as artificial neural networks, can process information with a high level of parallelism to efficiently solve difficult problems, even when implemented using conventional microelectronic technologies. We describe a mechanical device, which operates in a manner similar to artificial neural networks, to solve efficiently two difficult benchmark problems (computing the parity of a bit stream, and classifying spoken words). The device consists in a network of masses coupled by linear springs and attached to a substrate by non-linear springs, thus forming a network of anharmonic oscillators. As the masses can directly couple to forces applied on the device, this approach combines sensing and computing functions in a single power-efficient device with compact dimensions.