An Evolutionary Distributed Agent System: A Framework and Application to Manufacturing Design and Control

Information systems are subject to change from the environment. If systems are to be used beyond their initial purpose as conceived during the requirements definition stage, they must be designed to evolve. One approach to evolutionary complexity is to make systems dynamically configurable so that they can exhibit new behavior in the face of environmental change. In this paper, the authors propose a distributed agent approach to information systems development in order to overcome the inherent inflexibility of the agent systems developed with a traditional systems development methodology. They apply the framework to the development of alternative agent systems for manufacturing process design and scheduling problems. Through a simulation experiment, the authors demonstrate that the fit between the information system and its environment is one of the most important design considerations for the development of the evolutionary distributed agent system.