Data analysis and feature selection for predictive maintenance: A case-study in the metallurgic industry. (June 2019)

• Record Type: Journal Article
• Authors: Fernandes, Marta; Canito, Alda; Bolón-Canedo, Verónica; Conceição, Luís; Praça, Isabel; Marreiros, Goreti
• Is Part Of: International journal of information management. Volume 46(2019)
• Page Start: 252
• Page End: 262
• Publication Date: 2019-06
• Subjects: Predictive maintenance -- Data analysis -- Feature selection -- Rule-based model; Social sciences -- Information services -- Periodicals; Social sciences -- Research -- Periodicals; Information science -- Periodicals; Management information systems -- Periodicals; Knowledge management -- Periodicals; Sciences sociales -- Documentation, Services de -- Périodiques; Sciences sociales -- Recherche -- Périodiques; Sciences de l'information -- Périodiques; Systèmes d'information de gestion -- Périodiques; Information science; Management information systems; Social sciences -- Information services; Social sciences -- Research; Periodicals; Electronic journals; 025.52068
• DOI: 10.1016/j.ijinfomgt.2018.10.006
• Languages: English
• ISSNs: 0268-4012

Abstract

Highlights: Data analysis lead to a number of insights on the meaning of the features and how they describe machine behaviour. Some rules were derived from the associations found during the data analysis and were consolidated in a rule-based model. The rule-based model will be used to complement predictive maintenance models in the future. Combining feature selection methods with the data analysis insights helped reduce the feature space from 47 to 32 features. Abstract: Proactive Maintenance practices are becoming more standard in industrial environments, with a direct and profound impact on the competitivity within the sector. These practices demand the continuous monitorization of industrial equipment, which generates extensive amounts of data. This information can be processed into useful knowledge with the use of machine learning algorithms. However, before the algorithms can effectively be applied, the data must go through an exploratory phase: assessing the meaning of the features and to which degree they are redundant. In this paper, we present the findings of the analysis conducted on a real-world dataset from a metallurgic company. A number of data analysis and feature selection methods are employed, uncovering several relationships, which are systematized in a rule-based model, and reducing the feature space from an initial 47-feature dataset to a 32-feature dataset.