Iterative Method of Labor Division for Multi-Robotic Systems

Abstract

Labor division in multi-robotic systems allows distributing tasks between agents in order to increase the efficiency of performing the global task. Collective decision-making methods allow agents to form the "agent-task" pairs. In this paper, we consider the case when the number of tasks significantly exceeds the number of agents. We propose an iterative method of labor division in multi-robotic systems. It uses collective decision-making to assign a cluster of subtasks to an agent. The paper examines different ratios between cluster size, number of clusters, and number of agents in order to find ratios that provide minimal average global task execution time and minimal average energy consumption.