Controlling Chaos Using Edge Computing Hardware

TL;DR

This paper demonstrates that a reservoir computing-based nonlinear controller can effectively manage chaotic systems on embedded hardware, offering accurate control with minimal power consumption, suitable for edge computing applications.

Contribution

It introduces a reservoir computing controller capable of controlling chaos on embedded FPGA hardware with low power use, advancing edge deployment of machine learning.

Findings

Controller successfully manages chaotic system to arbitrary states

Model operates on FPGA with minimal power consumption

First demonstration of efficient ML control on embedded hardware

Abstract

Machine learning provides a data-driven approach for creating a digital twin of a system - a digital model used to predict the system behavior. Having an accurate digital twin can drive many applications, such as controlling autonomous systems. Often the size, weight, and power consumption of the digital twin or related controller must be minimized, ideally realized on embedded computing hardware that can operate without a cloud-computing connection. Here, we show that a nonlinear controller based on next-generation reservoir computing can tackle a difficult control problem: controlling a chaotic system to an arbitrary time-dependent state. The model is accurate, yet it is small enough to be evaluated on a field-programmable gate array typically found in embedded devices. Furthermore, the model only requires 25.0 $\pm$ 7.0 nJ per evaluation, well below other algorithms, even without systematic power optimization. Our work represents the first step in deploying efficient machine learning algorithms to the computing "edge."