With the ‘BeaT4.0’ project, supported by CIM4.0, the SKF Group is significantly accelerating its Industry 4.0 objectives in its high-volume ball bearing production plants through a new management of production performance that impacts the digitalization of machinery, maintenance decision-making processes, worker organization, and product supply to the final customer. Central to all this is the development and synergy of manufacturing technology skills, Condition Monitoring, and Predictive Maintenance, supported by SKF’s advanced cloud infrastructure and Alten’s Artificial Intelligence expertise, which contributes to the development of dedicated Machine Learning algorithms, also with the support of the University of Turin. New models of ‘data-driven prognostic’ and ‘prescriptive maintenance’ will be correlated both to the operational data of the machinery and to the qualitative data of the finished product. Starting from historical series, using the most modern multidimensional statistical techniques, phenomena underlying deviations from the correct operation of machinery will be sought, thus recognizing hidden correlations or dynamic changes that allow for timely prevention of the loss of reliability of the production process along with any future discrepancies. With Data Analytics techniques, parameters not previously analyzed that nonetheless impact company productivity and production efficiency can be selected.
The adoption, in the production phase, of Machine Learning algorithms that process significant process data acquired in real-time will allow the identification of anomalous patterns and thus the prevention of potentially harmful incorrect operations for the efficiency of the production line and the final product. Two SKF Industrie plants, in Cassino and Villar Perosa, will adopt the ‘prescriptive maintenance’ developed with Machine Learning with proprietary software platform; other production sites of the Group will follow.
