Modelling Knowledge about Data Analysis Processes in Manufacturing

Autoren
T. Neuböck, M. Schrefl
Paper
Neub15a (2015)
Zitat
Proceedings of the 15th IFAC/IEEE/IFIP/IFORS Symposium Information Control Problems in Manufacturing (INCOM 2015), May 11-13, 2015, Ottawa, Canada, IFAC-PapersOnLine, Volume 48, Issue 3, pp. 277-282, 2015.
Ressourcen
Kopie  (Senden Sie ein Email mit  Neub15a  als Betreff an dke.win@jku.at um diese Kopie zu erhalten)

Kurzfassung

In industry 4.0, analytics and business intelligence (BI) are of particular importance to increase productivity, quality, and flexibility. It is necessary to make right and quick decisions for effective and efficient problem solving and process improvements. Modern technologies allow to collect a large amount of data that can be analysed. Heterogeneity and complexity of industrial environments require considerable expert knowledge to perform meaningful and useful data analysis. BI analysis graphs represent expert knowledge about analysis processes. This knowledge can be modelled pro-actively at schema level and used at instance level. Analysis situations can be considered as multi-dimensional queries and represent nodes of a BI analysis graph. An arc between two nodes is a relationship between two analysis situations describing the difference of both. It represents a navigation step, e.g., an online analytical processing (OLAP) operation, of the analysis process. We demonstrate BI analysis graphs by a use case originated from manufacturing of brushes. Complex analysis paths, e.g., to analyse substitute material in the case of delayed delivery, are modelled by BI analysis graphs and can be used multiple times (also by non-experts). Reinvention of analysis knowledge is prevented - right and quick decisions for finding effective and efficient problem solutions can be made.


Keywords: decision making, knowledge representation, data models, process models, manufacturing, industry 4.0, data warehouses, business intelligence