On the quantitative evaluation of functionality in distributed intelligence systems

A quantitative methodology is presented for identifying and evaluating the shortfalls and overlaps between the desired functionality of a distributed intelligence system, as characterized by the requirements and the functionality of a proposed or implemented system. The approach is based on developing compatible Petri net representations of the requirements and of the system architecture which form the basis for the comparison of the desired and actual functionalities. By using the theory of Petri net invariants, the functionalities in both representations are obtained. They can then be compared to determine shortfalls and overlaps. The methodology is applied to a distributed intelligence system that models a distributed command center