Task Partitioning in Swarms of Robots: Reducing Performance Losses Due to Interference at Shared Resources

摘要

The performance of large groups of robots is often limited by a commonly shared resource. This effect, termed interference, can have a large impact on robotic swarms. This article studies the issue of interference in a swarm of robots working on a harvesting task. The environment of the robots is spatially constrained, i.e., there is a commonly shared resource, the nest, which limits the group's performance when used without any arbitration mechanism. The article studies the use of task partitioning for reducing concurrent accesses to the resource, and therefore limiting the impact of interference on the group's performance. In our study, we spatially partition the environment into two subparts, thereby partitioning the corresponding harvesting task as well. We employ a simple method to allocate individuals to the partitions. The approach is empirically studied both in an environment with a narrow nest area and an environment without this constraint. The results of the task partitioning strategy are analyzed and compared to the case in which task partitioning is not employed.