Applying Contextualization for Data-Driven Transformation in Manufacturing

Abstract

Manufacturing is highly distributed and involves a multitude of heterogeneous information sources. In addition, Production systems are increasingly interconnected, hence leading to an increase in heterogeneous data sources. At present, data available from these new type of systems are growing faster than the ability to productively integrate them into engineering and production value chains of companies. Known applications such as predictive maintenance and manufacturing equipment management are currently being continuously optimized. While these applications are designed to help companies manage their manufacturing and engineering data, they only use a fraction of the total potential that can be realized by linking manufacturing and engineering data with other enterprise data. In the future, the context in which the data can be set will play an essential role. A meaningful added value in manufacturing can be achieved only with context specific data. Against this background, this paper presents three main areas of application for contextualizing data (semantics, sensitivity and visualization) and explains these applications with the help of a contextualization architecture. The concept is also evaluated using an industrial example. Furthermore, the paper describes the theoretical background of contextualization and its application in industry. The major challenges of the ability of engineers to adapt their activities and the integration of process knowledge for semantic linking are addressed as well.