Introduction

Biology has shown that complex global behaviors can arise from simple interactions between large numbers of relatively unintelligent agents. Swarm approaches to robotics, involving large numbers of simple robots rather than a small number of sophisticated robots, has many advantages with respect to robustness and efficiency. Such systems can typically absorb many types of failures and unplanned behavior at the individual agent level, without sacrificing task completion. At the same time, one can exploit parallelism and spatially distributed sensing and action. Emerging technologies are making it possible to cheaply manufacture small robots with sensors, actuators and computation. This makes swarm intelligence an attractive solution for many problem domains.

Shapebugs is a decentralized method for coordinating a swarm of identically-programmed mobile agents to spatially self-aggregate into arbitrary shapes while using only local interactions. The agents accomplish such complex group goals by composing simple individual local behaviors like coordinate system negotiation and locally reactive movement. Powerful features that emerge as a result of the distributed approach include:

1. The formation process adapts dynamically to different and changing numbers of agents.
2. The swarm self-repairs and retains its shape in cases of accidental agent death, failure or disorientation.
3. The formation compensates overall for errors and limitations in the robot model, being robust to large degrees of robot sensor and movement inaccuracies.

Applications for Shapebugs include laying out efficient sensor networks in hostile environments, exploring or mapping unknown terrains, or maintaining formations for defense or herding tasks. Shapebugs represents a self-preserving, connected, mobile sensor network. Thus it could serve as a robust foundation for more sophisticated distributed intelligence to address various kinds of problems.

The purpose of this web page is to introduce and describe Shapebugs' perspective on swarm intelligence, and to also provide resources (experimental data, software, etc.) to those interested in further studying this research topic.