A computer model of operative reasoning

As machines become more complex, computers must play a larger role in their maintenance and operation. This is particularly evident in responding to faults. Already, prototype computer systems have been developed to diagnose failures in communications networks, digital circuits and power plants. As our ability to automate the type of reasoning that underlies diagnosis increases, more computer systems will be implemented in an even wider variety of domains.

The intent of this paper is to examine the role of knowledge in operating complex systems. Although operating any system consists of performing a number of varying tasks, we are primarily concerned with those that involve responding to faults. We first identify the conceptualizations that underly operative reasoning, and then construct a model that shows how the concepts interact with each other. Much of the model is encoded in a program for responding to faults in airplanes. The program, AIRMAN, is described along with several examples of its operation in a simulated airplane environment.