Variable Response Duration Promotes Self-organization in Decentralized Swarms

Decentralized Swarm Desynchronization via Inter-agent Variation for Logistical Resupply

“…Low threshold agents will respond when the task stimulus is low; high threshold agents will not respond until the task stimulus is higher. There are many examples in which variation in agent response thresholds successfully helps to coordinate decentralized agents, including studies on both static (dos Santos and Bazzan 2009;Kanakia, Touri, and Correll 2016;Krieger and Billeter 2000;Meyer et al 2015;Price and Tino 2004; and dynamic (Campos et al 2000;Castello et al 2018;de Lope, Maravall, and Quinonez 2015;Engholdt, Mathias, and Wu 2020;Goldingay and van Mourik 2013;Price and Tino 2004;Theraulaz, Bonabeau, and Deneubourg 1998) thresholds, as well as agents that behave deterministically (Krieger and Billeter 2000; and probabilistically (Castello et al 2013;Correll 2008;dos Santos and Bazzan 2009;Goldingay and van Mourik 2013;Kazakova and Wu 2018;Niccolini, Innocenti, and Pollini 2010;Nouyan et al 2005;Pang et al 2017;Price and Tino 2004;Yang, Chen, and Li 2010). Examination of whether the distribution of the threshold values affect swarm performance finds that simple uniformly distributed thresholds are effective and efficient on dynamic problems where task demands are not known in advance .…”

“…The response duration we study here, however, is an intrinsic characteristic of an agent, and not an external force. Initial studies on the effects of response duration suggest that it can effectively desynchronize agents in a swarm but may reduce the overall responsiveness of a swarm (Engholdt, Mathias, and Wu 2020).…”