Semantic Heterogeneity Reduction for Big Data in Industrial Automation
Václav Jirkovský, Marek Obitko · ITAT · 2014
The large amounts of diverse data collected in industrial au- tomation domain, such as sensor measurements together with informa- tion in MES/ERP 1 systems need special handling that was not possible in past. The Big Data technologies contribute a lot to the possibility of analyzing such amounts of data. However, we need to handle not only data volume, which is usually the major focus of Big Data research, but we also need to focus on variety of data. In this paper, we primarily focus on variety of industrial automation data and present and discuss a possible approach of handling the semantic heterogeneity of them. We show the process of heterogeneity reduction that exploits Semantic Web technologies. The steps include construction of upper ontology describ- ing all data sources, transformation of data according to this ontology and finally the analysis with the help of Big Data paradigm. The pro- posed approach is demonstrated on data measured by sensors in a passive house.