Issues in the scaling of multi-robot systems for general problem solving

Abstract

Problem solving using multi-agent robotic systems has received significant attention in recent research. Complex strategies are required to organize and control these systems. Biological-inspired methodologies are often employed to bypass this complexity, e.g. self-organization. However, another line of research is to understand the relationship between low-level behaviors and complex high-level strategies. In this paper, we focus on understanding the interference caused in multi-robotic systems for the problem of search and tagging. Given a set of targets that must be found and tagged by a set of robots, what are the effects of scaling the number of robots and sensor ranges? Intuitively, increasing robot numbers, or sensor strength would seem beneficial. However, experience suggests that path and sensor interference caused by increased robots, increased targets, and sensor range will be harmful. The following investigation uses several abstract models to elucidate the issues of robot scaling and sensor noise.