Evolving Dyadic Strategies for a Cooperative Physical Task

Autor:	Sheybani, Saber, Izquierdo, Eduardo J., Roth, Eatai
Rok vydání:	2020
Předmět:	Computer Science - Neural and Evolutionary Computing Computer Science - Multiagent Systems
Druh dokumentu:	Working Paper
Popis:	Many cooperative physical tasks require that individuals play specialized roles (e.g., leader-follower). Humans are adept cooperators, negotiating these roles and transitions between roles innately. Yet how roles are delegated and reassigned is not well understood. Using a genetic algorithm, we evolve simulated agents to explore a space of feasible role-switching policies. Applying these switching policies in a cooperative manual task, agents process visual and haptic cues to decide when to switch roles. We then analyze the evolved virtual population for attributes typically associated with cooperation: load sharing and temporal coordination. We find that the best performing dyads exhibit high temporal coordination (anti-synchrony). And in turn, anti-synchrony is correlated to symmetry between the parameters of the cooperative agents. These simulations furnish hypotheses as to how human cooperators might mediate roles in dyadic tasks. Comment: 6 pages, 4 figures, IEEE Haptics Symposium 2020
Databáze:	arXiv
Externí odkaz:	http://arxiv.org/abs/2004.10558 Zobrazit plný text záznamu View this record from Arxiv