Predictive Maintenance in the Metallurgical Industry: Data Analysis and Feature Selection
As a consequence of the increasing competitivity in the current economic environment, Proactive Maintenance practices are gradually becoming more common in industrial environments. In order to implement these practices, large amounts of heterogeneous information must be analysed, such that knowledge about the status of the equipment can be acquired. However, for this data to be of use and before it can be processed by machine learning algorithms, it must go through an exploratory phase. During this step, relationships in the data, redundancy of features and possible meanings can be assessed. In this paper, a number of these procedures are employed, resulting in the discovery of meaningful information. Moreover, a subset of features is selected for future analysis, allowing for the reduction of the feature space from 47 to 32 features.
Keywords:Predictive maintenance Data analysis Feature selection
The present work has been developed under the EUREKA - ITEA2 Project INVALUE (ITEA-13015), INVALUE Project (ANI|P2020 17990), and has received funding from FEDER Funds through NORTE2020 program and from National Funds through FCT under the project UID/EEA/00760/2013.
- 4.Selcuk, S.: Predictive maintenance, its implementation and latest trends. Proc. Inst. Mech. Eng. Part B: J. Eng. Manuf., 0954405415601640 (2016)Google Scholar
- 8.Snijders, C., Matzat, U., Reips, U.-D.: “Big Data”: big gaps of knowledge in the field of internet science. Int. J. Internet Sci. 7(1), 1–5 (2012)Google Scholar
- 10.InValuePt. InValuePT – Home (2017), http://www.invalue.com.pt/. Accessed 21 Nov 2017
- 11.Canito, A., et al.: An Architecture for proactive maintenance in the machinery industry. In: International Symposium on Ambient Intelligence. Springer (2017)Google Scholar
- 12.Isabelle, G.: Feature Extraction foundations and applications. Pattern Recognit. (2006)Google Scholar