ServiceNavigator - A Bayesian Assistance System for Diagnosing Industrial Production Systems
Service technicians servicing machines in small-and medium-sized enterprises face the challenge to diagnose machines with increasing complexity in less time. To help them cope with the task of diagnosis (i.e. finding faults) this article introduces a novel fault diagnosis algorithm, and a web-based implementation for industrial fault diagnosis. When a fault occurs the algorithm proposes observations for the service technicians and generates likely causes according to the observations taken. This helps technicians to find faults faster, facilitates management of expert knowledge, and ultimately decreases system downtime. We have evaluated our approach with a Monte Carlo simulation of an industrial packaging machine and through the implementation of some prototype software. Both evaluations show that our approach is usable for real-world service technicians and operators of production machinery.