Two Paradigms for the Design of Artificial Collectives

Summary Artificial collectives are systems composed of multiple autonomous information or software agents, mobile robots, or nodes in a sensor or communication network. In the future, such systems will be responsible for many important tasks, such as highway traffic control, disaster response, toxic spill monitoring and cleanup, and exploration of other planets. Because such systems will have to function in environments with unreliable communication channels, where agents are likely to fail, they will have to be reliable, scalable, robust, adaptable, and amenable to quantitative mathematical analysis. The last property is important because analysis is crucial to understanding the issues of the design, control, adaptability, and dynamics of collective behavior. We describe two approaches to distributed control of artificial collectives and study them quantitatively. The first, biologically based control, relies on local interactions among many simple agents to create desirable collective behavior. The second approach allows collectives to maximize their world utility using market-based mechanisms. We present two applications—foraging in a group of robots and resource allocation in dynamic environments—that use these control paradigms and perform an analysis of each problem.