In this paper, we look at the possible role for a Distributed Artificial Intelligence (Distributed AI) concept to be applied in a large scale networked training environment. Here, a battle conditions environment, would consist of many intelligent simulators, each of which is equipped with different, but possibly overlapping expertise. The goal is to coordinate these battle forces in such a way as to carry out an offensive attack. The motivation for developing and applying a Distributed AI concept seems clear since the problem of a large scale network simulation is, itself, inherently distributed.

We will first introduce the critical role of Distributed AI in a large scale simulation environment in terms of knowledge, goals, skills, and coordination for the intelligent simulators. Some basic concepts in Distributed AI will be presented together with a number of ongoing research studies in distributed intelligence. We will then give a few domain examples in battlefield simulation, and describe how Distributed AI methodologies may be explored in such diverse environments. Finally, we will look at the advantages and disadvantages of Distributed AI as a viable technology for distributed training. A large conceptual framework will be used in the analysis of difficulties and possibilities of Distributed AI as applied in the distributed training system environment.