Boolean network robotics: a proof of concept

TL;DR

This paper demonstrates how Boolean networks can be used as a control mechanism for robots, enabling complex behaviors like navigation and memory formation through an automated design process.

Contribution

It introduces a novel approach to robot control using Boolean networks designed via stochastic local search, showcasing their capability for complex task execution.

Findings

Successfully designed a Boolean network controlling robot navigation.

The network enables internal memory and response to light stimuli.

Automated design process simplifies creating complex behaviors.

Abstract

Dynamical systems theory and complexity science provide powerful tools for analysing artificial agents and robots. Furthermore, they have been recently proposed also as a source of design principles and guidelines. Boolean networks are a prominent example of complex dynamical systems and they have been shown to effectively capture important phenomena in gene regulation. From an engineering perspective, these models are very compelling, because they can exhibit rich and complex behaviours, in spite of the compactness of their description. In this paper, we propose the use of Boolean networks for controlling robots' behaviour. The network is designed by means of an automatic procedure based on stochastic local search techniques. We show that this approach makes it possible to design a network which enables the robot to accomplish a task that requires the capability of navigating the space using a light stimulus, as well as the formation and use of an internal memory.