A distributed problem solving environment for multi-agent systems

Abstract

A distributed shell for the implementation of cooperative knowledge based sys tems, a general purpose architecture particularly suited for real world Distributed Artificial Intelligence (DAI) applications, in areas ranging from diagnosis, to the ones of classification and control, it is presented. Human domain experts are mapped to the system’s environment through agents. These particular entities are members of specialized communities that act as coherent and cooperative teams, react opportunisticly to events, being assyncronous and concurrent. The system’s modular design and functional mechanisms allows for the integration of new agents without affecting the overall system’s structure.