A knowledge engineering methodology for resource monitoring in the industrial domain

In industrial production, the efficient use of resources is becoming increasingly important due to the rising prices of resources and ecological awareness. A resource monitoring system can observe the individual resource consumption of the devices in an industrial plant to detect excessive resource consumption while considering context information such as the process steps or varying energy costs. A systematic approach for the semantic integration of relevant information is a key requirement to infer reliable monitoring states of the devices. Recent research explores how future industrial plants may use decentralized autonomous devices. To integrate such autonomous devices in a monitoring system, semantic models are required that describe the devices and their data, thus we need knowledge-based approaches. To date, the research on knowledge-based systems in the industrial domain has tended to focus on application-specific solutions rather than on generic solutions that integrate knowledge of multiple sources. In this paper, we introduce a knowledge-engineering methodology to build a generic, reusable resource monitoring system for industrial plants which includes multiple knowledge sources. This approach can also be used to develop knowledge-based systems for other applications such as diagnostic or prognostic systems.