Predictive Maintenance Support System in Industry 4.0 Scenario

Abstract

The fourth industrial revolution that is being witnessed nowadays, also known as Industry 4.0, is heavily related to the digitization of manufacturing systems and the integration of different technologies to optimize manufacturing. By combining data acquisition using specific sensors and machine learning algorithms to analyze this data and predict a failure before it happens, Predictive Maintenance is a critical tool to implement towards reducing downtime due to unpredicted stoppages caused by malfunctions. Based on the reality of Commercial Specialty Tires factory at Continental Mabor - Indústria de Pneus, S.A., the present work describes several problems faced regarding equipment maintenance. Taking advantage of the information gathered from studying the processes incorporated in the factory, it is designed a solution model for applying predictive maintenance in these processes. The model is divided into two primary layers, hardware, and software. Concerning hardware, sensors and respective applications are delineated. In terms of software, techniques of data analysis namely machine learning algorithms are described so that the collected data is studied to detect possible failures.