This paper gives a theoretical foundation and implementation details of data and control structures in a diagnostic expert system with adaptive uncertainty management. The considered problem covers a class of expert systems and the application domain in which the operating personnel actively monitor and control the outcome of an industrial manufacturing process. Based on subjective ranking of the latent fault in the final output product, the operator queries the expert system for a recommended corrective action. Subsequent successful or unsuccessful corrective actions consistently modify confidence measures of all corrective recommendations from the same fault set, according to the embedded analytical model. The described approach is based on the relational data representation and procedural control structures, as opposed to a more traditional declarative, production-based system.

The application of Bayesian networks for monitoring and diagnosis of a multistage manufacturing process is described. Bayesian network “part models” were designed to represent individual parts in-process. These were combined to form a “process model,” a Bayesian network model of the entire manufacturing process. An efficient procedure is designed for managing the “process network.” Simulated data is used to test the validity of diagnosis made from this method. In addition, a critical analysis of this method is given, including computation speed concerns, accuracy of results, and ease of implementation. Finally, a discussion on future research in the area is given.