How the application of big data has improved Pomini’s predictive maintenance thanks to the R&D project BigData@MA.
How much value do we lose everyday if we don’t collect and analyze big data related to our roll shop machines? Back in 2015, our Pomini team started to address this important issue with the idea of building a network among all roll grinding machines of a production site. The ultimate goal? To connect different sites with each other, in order to check the conditions of each machine remotely, from a centralized monitoring system.
A few years later, this fascinating and yet complex process was accelerated thanks to our participation in the MANUNET project BigData@MA (March 2018-September 2020) – an innovation-driven, close-to-market R&D initiative co-financed by the European Commission, the Lombardy Region and the Belgian Walloon Region as part of the MANUNET Transnational Call 2017, and led by Rina Consulting.
BigData@MA aimed to change the way industrial machines are used and maintained thanks to the application of big data technologies, which allow to merge high-dimensional data (pictures, videos, data sensors, etc.) with complex event processing to acquire significant information, detect plant/process/product anomalies and provide real-time support in decision. The project represented a great opportunity to boost Pomini’s research on big data technologies.
Three steps were fundamental to the development of the project.
Machine selection: We decided to include in the project Pomini roll grinding machines without the latest automation technologies. This choice reflected our intention to develop an advanced maintenance package even for clients who do not have the newest models installed (which are already designed to be equipped with advanced sensors), making this R&D activity applicable to the broadest number of cases.
Cloud infrastructure set-up: We selected Microsoft Azure as cloud computing service and, together with our digital experts, created the Human-Machine Interfaces and all necessary hardware and software modifications to get the machines connected with Tenova EDGE to gather data.
Data analysis: From the data collected, we made interesting discoveries regarding the operating conditions both of the electric motors and of the machines, as well as the entire machining process. The analyses allowed us to develop new maintenance algorithms.
A novel approach to machine maintenance
The experience of BigData@MA taught us that predictive maintenance demands continuous research and adjustments: failures, adaptations and changes were a natural part of the process, and unexpected results and upturns have to be taken into account.
Today, we benefit from the concrete achievements of the project: the novel system that has been developed is installed on all machines, ready to be used when internet connection is available on site. We are now capable to open up new perspectives for advanced and remote maintenance for our customers.
BigData@MA represented an exciting project, but only a first step into a fascinating journey we continue to pursue.