Coordination of Communication in Robot Teams by Reinforcement Learning - Foundations on Natural and Artificial Computation - page 155

Coordination of Communication in Robot Teams
by Reinforcement Learning

Dar ıo Maravall 1 ,JavierdeLope 1 , 2 ,andRaul Dom ınguez 1

1 Cognitive Robotics Group
Dept. of Artificial Intelligence
Universidad Politecnica de Madrid
2 Dept. Applied Intelligent Systems
Universidad Politecnica de Madrid
[e-mail],[e-mail],[e-mail]

Abstract. In Multi-agent systems, the study of language and communi-
cation is an active field of research. In this paper we present the applica-
tion of Reinforcement Learning (RL) to the self-emergence of a common
lexicon in robot teams. By modeling the vocabulary or lexicon of each
agent as an association matrix or look-up table that maps the meanings
(i.e. the objects encountered by the robots or the states of the environ-
ment itself) into symbols or signals we check whether it is possible for
the robot team to converge in an autonomous, decentralized way to a
common lexicon by means of RL, so that the communication eciency of
the entire robot team is optimal. We have conducted several experiments
aimed at testing whether it is possible to converge with RL to an optimal
Saussurean Communication System. We have organized our experiments
alongside two main lines: first, we have investigated the effect of the team
size centered on teams of moderated size in the order of 5 and 10 indi-
viduals, typical of multi-robot systems. Second, and foremost, we have
also investigated the effect of the lexicon size on the convergence re-
sults. To analyze the convergence of the robot team we have defined
the team's consensus when all the robots (i.e. 100% of the population)
share the same association matrix or lexicon. As a general conclusion we
have shown that RL allows the convergence to lexicon consensus in a
population of autonomous agents.

Keywords: Multi-agent systems, Multi-robot systems, Dynamics of ar-
tificial languages, Computational semiotics, Reinforcement learning, Self-
collective coordination, Language games, Signaling games.

1 Introduction

In a multi-robot system obtaining a common lexicon or vocabulary is a basic
step towards an ecient performance of the whole system [4]. In this paper we
present the application of Reinforcement Learning (RL) to the emergence of a
common lexicon in a team of autonomous robots. We model the vocabulary or
lexicon of each robot as an association matrix or look-up-table that maps the