Complexity Measures: Open Questions and Novel Opportunities in the Automatic Design and Analysis of Robot Swarms.

Andrea Roli,Mauro Birattari,Antoine Ligot

doi:10.3389/frobt.2019.00130

Andrea Roli, Mauro Birattari + Show 1 more

Open Access

https://doi.org/10.3389/frobt.2019.00130

Copy DOI

Abstract

Complexity measures and information theory metrics in general have recently been attracting the interest of multi-agent and robotics communities, owing to their capability of capturing relevant features of robot behaviors, while abstracting from implementation details. We believe that theories and tools from complex systems science and information theory may be fruitfully applied in the near future to support the automatic design of robot swarms and the analysis of their dynamics. In this paper we discuss opportunities and open questions in this scenario.

Highlights

Metrics that quantify the complexity of a system and measure information processing are used in a wide range of scientific areas, including neuroscience, physics, and computer science
A possibility for using information theory in robotics is enabled by the notion of sensory-motor coordination (Pfeifer and Scheier, 1997) which emphasizes the role of the loop between sensors and actuators in robots performing cognitive tasks
While this way of estimating artificial neural network (ANN) complexity is common in machine learning and is not based on information theory, this paper provides a clear statement of the problem

Summary

INTRODUCTION

Metrics that quantify the complexity of a system and measure information processing are used in a wide range of scientific areas, including neuroscience, physics, and computer science. Let us consider a swarm of robots: we may ask what is the complexity of a function describing the overall behavior of the swarm, or what is the complexity of the problem of optimally assigning tasks to the robots, or what is the complexity of each of the tasks. These objectives require different measures, each addressing a specific question. The aim of this paper is to outline what we think are the most important open questions and to describe opportunities to use complexity measures for supporting the automatic design of swarms of robots and the analysis of their behaviors.

A CAPSULE INTRODUCTION TO COMPLEXITY MEASURES

COMPLEXITY MEASURES IN ROBOTICS

OPEN QUESTIONS AND OPPORTUNITIES FOR FUTURE RESEARCH