TY - GEN
T1 - Self-organised aggregation in swarms of robots with informed robots
AU - Firat, Ziya
AU - Ferrante, Eliseo
AU - Cambier, Nicolas
AU - Tuci, Elio
N1 - Publisher Copyright:
© 2018, Springer Nature Switzerland AG.
PY - 2018
Y1 - 2018
N2 - In this paper, we study a swarm of robots that has to select one aggregation site in an environment in which two sites are available. It is known in the literature that, in presence of asymmetries in the environment, robot swarms are able to perform a collective choice and aggregate in one among two possible sites, for example the largest of the two. We focus on an aggregation scenario where the environment is morphologically symmetric. The two aggregation sites are identical with only one exception: their colour. In addition, in the swarm only a proportion of robots, that we call the informed robots, possess extra information concerning on which specific site the swarm is required to aggregate. The rest of the robots are non-informed, thus they do not possess the above mentioned extra information. In simulation-based experiments we show that, if no robot in the swarm is informed, the swarm is able to break the symmetry and aggregates on one of the two sites at random. However, the introduction of a small proportion of informed robots is enough to break the symmetry: the majority of the swarm aggregates on the site preferred by the informed robot. Additionally, the swarm is also able to completely aggregate on one of the two sites when only 30% of the robots are informed, independently from the swarm size among those we considered. Finally, we analyse how the time dynamics of the aggregation process depend on the proportion of informed robots.
AB - In this paper, we study a swarm of robots that has to select one aggregation site in an environment in which two sites are available. It is known in the literature that, in presence of asymmetries in the environment, robot swarms are able to perform a collective choice and aggregate in one among two possible sites, for example the largest of the two. We focus on an aggregation scenario where the environment is morphologically symmetric. The two aggregation sites are identical with only one exception: their colour. In addition, in the swarm only a proportion of robots, that we call the informed robots, possess extra information concerning on which specific site the swarm is required to aggregate. The rest of the robots are non-informed, thus they do not possess the above mentioned extra information. In simulation-based experiments we show that, if no robot in the swarm is informed, the swarm is able to break the symmetry and aggregates on one of the two sites at random. However, the introduction of a small proportion of informed robots is enough to break the symmetry: the majority of the swarm aggregates on the site preferred by the informed robot. Additionally, the swarm is also able to completely aggregate on one of the two sites when only 30% of the robots are informed, independently from the swarm size among those we considered. Finally, we analyse how the time dynamics of the aggregation process depend on the proportion of informed robots.
KW - Aggregation
KW - Informed leaders
KW - Self-organisation
KW - Swarm intelligence
KW - Swarm robotics
UR - http://www.scopus.com/inward/record.url?scp=85058545453&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85058545453&partnerID=8YFLogxK
U2 - 10.1007/978-3-030-04070-3_4
DO - 10.1007/978-3-030-04070-3_4
M3 - Conference contribution
AN - SCOPUS:85058545453
SN - 9783030040697
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 49
EP - 60
BT - Theory and Practice of Natural Computing - 7th International Conference, TPNC 2018, Proceedings
A2 - Martín-Vide, Carlos
A2 - Vega-Rodríguez, Miguel A.
A2 - Fagan, David
A2 - O’Neill, Michael
PB - Springer Verlag
T2 - 7th International Conference on the Theory and Practice of Natural Computing, TPNC 2018
Y2 - 12 December 2018 through 14 December 2018
ER -