Software Support for Organised Adaptation

Coordination of agent teams depends on the scale of the team. Teams comprising hundreds of agents tend to perform better with local computation and interactions (i.e. swarm intelligence, evolutionary computing, etc.) Teams comprising tens of agents tend to perform better with more sophisticated agents (e.g. BDI agents) with complex reasoning abilities. In the long term, our objective is to achieve agent teams comprising hundreds of sophisticated agents: then, a key aspect of coordination and control is the idea of organised adaptation. In this paper we present a new multi-agent programming environment, PreSage-MS, a rapid prototyping and animation tool designed to facilitate experiments with organised adaptation in ‘sophisticated’ agent teams. We describe the system architecture and functionality, and give a walkthrough of experimental design. We conclude with a discussion of several issues, including the migration from design-time tools for human users to run-time services for software agents. Categories and subject descriptors: I.2.5 [Artificial Intelligence]: Programming Languages and Software; I.2.11 [Artificial Intelligence]: Distributed Artificial Intelligence—Intelligent Agents General terms: Design, Experimentation