TY - GEN
T1 - Guiding aggregation dynamics in a swarm of agents via informed individuals
T2 - 2019 Conference on Artificial Life: How Can Artificial Life Help Solve Societal Challenges, ALIFE 2019
AU - Gillet, Yannick
AU - Ferrante, Eliseo
AU - FIRAT, Ziya
AU - Tuci, Elio
N1 - Publisher Copyright:
Copyright © ALIFE 2019.All rights reserved.
PY - 2020
Y1 - 2020
N2 - Self-organised aggregation, the formation of large clustersof independent agents, is an important process in swarmrobotics systems since it is the prerequisite for more com-plex collective behaviours. Previous work on self-organisedaggregation focused on the study of the individual mecha-nisms required to allow a swarm to form a single aggregate.In this paper, we discuss an analytical model which looks atthe possibility to use the concept of informed individuals toallow the swarm to distribute on different aggregation sitesaccording to proportions of individuals at each site arbitrarilychosen by the designer. Informed individuals are opinionatedagents that selectively prefer an aggregation site and avoid torest on the non-preferred sites. We study environments withtwo aggregation sites, and consider two different scenarios:one in which the informed individuals are equally distributedin numbers between the two sites; and one in which informedindividuals for one type of site are three times more numer-ous than those on the other site. Our objective is to find outwhether and for what range of model parameters the swarmdistributes between the two sites according to the relative dis-tribution of informed agents among the two sites. The analy-sis of the model shows that the designer capability to exploitinformed individuals to control how the swarm aggregatesdepends on the environmental conditions. For intermediatevalues of the site carrying capacity, a small minority of in-formed individuals is able to guide the dynamics as desiredby the designer. We also show that the larger the site carryingcapacity the larger the total proportion of informed individu-als required to lead the swarm to the desired distribution ofindividuals between the two sites
AB - Self-organised aggregation, the formation of large clustersof independent agents, is an important process in swarmrobotics systems since it is the prerequisite for more com-plex collective behaviours. Previous work on self-organisedaggregation focused on the study of the individual mecha-nisms required to allow a swarm to form a single aggregate.In this paper, we discuss an analytical model which looks atthe possibility to use the concept of informed individuals toallow the swarm to distribute on different aggregation sitesaccording to proportions of individuals at each site arbitrarilychosen by the designer. Informed individuals are opinionatedagents that selectively prefer an aggregation site and avoid torest on the non-preferred sites. We study environments withtwo aggregation sites, and consider two different scenarios:one in which the informed individuals are equally distributedin numbers between the two sites; and one in which informedindividuals for one type of site are three times more numer-ous than those on the other site. Our objective is to find outwhether and for what range of model parameters the swarmdistributes between the two sites according to the relative dis-tribution of informed agents among the two sites. The analy-sis of the model shows that the designer capability to exploitinformed individuals to control how the swarm aggregatesdepends on the environmental conditions. For intermediatevalues of the site carrying capacity, a small minority of in-formed individuals is able to guide the dynamics as desiredby the designer. We also show that the larger the site carryingcapacity the larger the total proportion of informed individu-als required to lead the swarm to the desired distribution ofindividuals between the two sites
UR - http://www.scopus.com/inward/record.url?scp=85085052771&partnerID=8YFLogxK
U2 - 10.1162/isal_a_00225
DO - 10.1162/isal_a_00225
M3 - Conference contribution
T3 - The 2019 Conference on Artificial Life
SP - 590
EP - 597
BT - Proceedings of the 2019 Conference on Artificial Life
PB - MIT Press
Y2 - 29 July 2019 through 2 August 2019
ER -