Machine learning and reasoning for predictive maintenance in Industry 4.0: Current status and challenges

• Published in: Computers in industry Vol. 123; p. 103298
• Main Authors: Dalzochio, Jovani, Kunst, Rafael, Pignaton, Edison, Binotto, Alecio, Sanyal, Srijnan, Favilla, Jose, Barbosa, Jorge
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 01.12.2020
• Subjects: Artificial intelligence; Industry 4.0; Internet of Things; Ontology; Predictive maintenance; Systematic literature review; Industry 4.0; Ontology; Internet of Things; Artificial intelligence; Systematic literature review; Predictive maintenance
• ISSN: 0166-3615
• DOI: 10.1016/j.compind.2020.103298

•Predictive maintenance is a hot topic in the context of Industry 4.0.•The papers that bring novelty to the field are concentrated in the years 2018 and 2019.•We envision the necessity of implementing the theoretical frameworks found in the literature in real industrial environments.•Challenges related to big data are also of interest.•Issues like scalability, latency, and data security deserve further investigation. In recent years, the fourth industrial revolution has attracted attention worldwide. Several concepts were born in conjunction with this new revolution, such as predictive maintenance. This study aims to investigate academic advances in failure prediction. The prediction of failures takes into account concepts as a predictive maintenance decision support system and a design support system. We focus on frameworks that use machine learning and reasoning for predictive maintenance in Industry 4.0. More specifically, we consider the challenges in the application of machine learning techniques and ontologies in the context of predictive maintenance. We conduct a systematic review of the literature (SLR) to analyze academic articles that were published online from 2015 until the beginning of June 2020. The screening process resulted in a final population of 38 studies of a total of 562 analyzed. We removed papers not directly related to predictive maintenance, machine learning, as well as researches classified as surveys or reviews. We discuss the proposals and results of these papers, considering three research questions. This article contributes to the field of predictive maintenance to highlight the challenges faced in the area, both for implementation and use-case. We conclude by pointing out that predictive maintenance is a hot topic in the context of Industry 4.0 but with several challenges to be better investigated in the area of machine learning and the application of reasoning.