The digital transformation of manufacturing has been underway for a number of years and has been greatly accelerated by the Industry 4.0 Plan, which gave a major boost to innovation in the sector. It was with the pandemic, however, that even the most skeptical realized they had to keep up with the times: remote monitoring and maintenance were no longer a luxury, but a necessity. And, as traffic restrictions eased, the solutions put in place during emergencies were in some cases enhanced, because those who adopted them realized that they could provide greater productivity and significantly cut costs by reducing the travel of service technicians to the necessary minimum, but also by anticipating failures through predictive maintenance solutions. These monitor all parameters, and alert operators when some of them are outside optimal ranges, a sign of a possible breakdown. By intervening early, it is thus possible to greatly limit downtime, all to the benefit of production efficiency.
We talked about smart monitoring here, and here we will elaborate on the concept by going into SAP solutions to enable smart monitoring and some concrete use cases.
IIoT, the parameters that matter
Predictive maintenance can be enabled both on newer machines, which are already equipped with all the necessary sensors and connectivity, and on older machines, even those decades old, to which inexpensive sensors should be added that can detect a number of key parameters, offering insights into the operation of the equipment. Which parameters are most important? At the top of the list is vibration monitoring. High vibration levels could be caused by bearing problems, while small variations in vibration patterns could indicate misalignment of some component.
Infrared cameras installed of the vicinity of the machines, on the other hand, enable the real time temperature monitoring, making it possible to check the temperatures of each individual component at any instant, so as to quickly identify those that are starting to overheat and could therefore compromise the reliability of the instrument. Analysis of the sounds emitted by the machinery is also very important, since if entrusted to machine learning algorithms it can signal the level of wear and tear of certain parts.
Far more effectively than the human ear, since the sensors can also acquire ultrasonic frequencies.
Those presented here are just a few examples of the key parameters, but certainly not the only ones: fluid analysis makes it possible to verify the proper lubrication of various parts, while other sensors make it possible to identify in advance corrosion of materials, cracks in machinery, and problems with electrical circuits.
Trenitalia improves train maintenance with SAP Predictive Maintenance and Service
The cooperation between SAP and Trenitalia has been going on for many years. Trenitalia is a company that has about 30,000 vehicles including locomotives, electric trains, passenger and freight vehicles, and about 8,000 of its trains run on the rail network every day. This is not an easy fleet to keep up, and to optimize maintenance operations, the company has adopted SAP solutions. Specifically, SAP Predictive Maintenance and Service makes it possible to analyze data from IoT sensors installed on board vehicles and monitor the status of vehicles remotely. The SAP HANA-based predictive analytics system finally processes this huge amount of information through machine learning algorithms, so as to check for problems before they occur and plan maintenance more effectively, so as to limit unplanned downtime as much as possible and cut maintenance costs by 8 percent to 10 percent.
How? The data captured (we’re talking 5,000 values recorded per second, about 450 billion values daily) allowed over time to better understand how many opening-closing cycles the doors could endure before exhibiting malfunctions, but also how long the brakes would last. Previously, this was based on a mileage estimate, but SAP solutions have helped the company more accurately estimate wear and tear by measuring the heat dissipated during braking. The adoption of the Smart Monitoring solution has also made it possible to identify correlations that would have been very difficult to unearth, such as the different performance of certain types of batteries on different trains.
The next step to real time monitoring: the digital twin
Acquiring real-time data enables predictive maintenance and remote monitoring, but it is also possible to extend this concept further, going so far as to create a digital twin, a digital twin, of a specific asset, or an entire factory. The advantage is that you will have a digital replica that is identical in every way to the original asset, partly because the data that feeds this twin are those acquired in real time from the machinery. What are digital twins used for? There are many scenarios for use, but the main advantage is that they make it possible to simulate alternative scenarios without touching the production tools, so as to improve process efficiency, reduce energy consumption, and generally optimize production and distribution.
One example comes from the giant Unilever, which in 2019 launched a pilot project on one of its production plants (soap, specifically), with plans to extend it to all 300 factories. The result has been positive, and the adoption of the digital twin has enabled the plant to increase productivity while reducing energy consumption. Specifically, Unilever claims that adoption at one plant alone has resulted in savings of $2.8 million.
Smart monitoring is a key part of any digital transformation initiative. By collecting data and monitoring consumption, you can contain costs and better understand how the company leverages resources and uses technology.
Contact us to receive more information and book your free consultation.
