A Systemic Metaphor of Multi-Agent Coordination in Living Systems
Uwe Borghoff, Paolo Bottoni, Pierre Mussio, Remo Pareschi
Living systems provide a good metaphor for the coordination of autonomous agents managing tasks according
to their role in the system. The study of such systems from the point of view of General Systems Theory would
benefit from computational experimentation. This paper proposes an agent-based methodology for system
modeling as well as computational tools to develop and test system models. Biological problems are general
enough to provide a metaphor by which one can model several problems in coordinating multi-agent behavior.
We show that systems theory and the study of living systems provide insight in the various problems of
multi-agent coordination, and that this insight produces computational models where we can express complex
coordination patterns involving highly-specialized agents, particularly those cases where the issue of sharing
limited resources is crucial.
Proc. of the 10th Europ. Simulation Multiconf. (ESM'96), June 2-6, 1996, Budapest, Hungary. San Diego, CA: The Society for Computer Simulation, pp. 245-253 (an earlier version was presented at the Intl. Workshop on Biologically Inspired Evolutionary Systems (BIES'95), May, 1995, Tokyo, Japan).
systemic-metaphor.pdf (146.16 kB)