Datasets: Ontology-based diagnosis reporting and monitoring to improve fault finding in Industry 4.0
datasetposted on 14.08.2020 by Iñigo Fernández del amo blanco, John ahmet Erkoyuncu, Maryam Farsi, Dominik Bulka, Stephen Wilding
Datasets usually provide raw data for analysis. This raw data often comes in spreadsheet form, but can be any collection of data, on which analysis can be performed.
This repository includes datasets on experimental cases of study and analysis regarding the research called "Ontology-based diagnosis reporting and monitoring to reduce no-fault-found scenarios in Industry 4.0".
Abstract: "Industry 4.0 is bringing a new era of digitalisation for complex equipment. It especially benefits equipment’s monitoring and diagnostics with real-time analysis of heterogenous data sources. Management of such sources is an important research challenge. A relevant research gap involves integration of experts’ diagnosis knowledge. Experts have valuable knowledge on failure conditions that can support monitoring systems and their limitations in no-fault-found scenarios. But their knowledge is normally transferred as reports, which include unstructured data difficult to re-use. Thus, this paper proposes ontology-based diagnosis reporting and monitoring methods to capture and re-use expert knowledge for improving diagnosis efficiency. It aims to capture expert knowledge in a structured format and re-use it in monitoring systems to provide failure recommendations in no-fault-found conditions. This research conducted several methods for validating the proposed methods. Laboratory experiments present time and errors reduction rates of 20% and 12% compared to common data-driven monitoring approaches for diagnosis tasks in no-fault-found scenarios. Subject-matter experts’ surveys evidence the usability of the proposed methods to work in real-life conditions. Thus, this paper’s proposal can be considered as a method to bridge the gap for integrated data management in the context of Industry 4.0."