Michael Matthesius, Head of Global Industry Management Machinery
WIN! Temperature, pressure, energy consumption, vibration – more and more sensors in the field mean continually growing volumes of data. How can this mass of data be kept in check so that valuable information can ultimately be obtained from it?
Matthesius: The trick is filtering out the relevant data from this enormous volume. To do so, we look for patterns, and occasionally manage to uncover hitherto unknown and astonishing relationships. On this basis, we are able to understand the normal behaviour of the machine or production facility, which is then compared with the learned model during operation. If the current machine data deviates from this model, we have an anomaly. If the anomaly recurs repeatedly, this then indicates a potential fault. By following this process, we can predict machine states and enable our customers to respond proactively.
WIN! Not all machinery and production facilities currently provide data and information for further analysis. Will your customers first have to upgrade their sensors before data analysis is possible?
Gaukstern: In most cases there is sufficient data and the useful information just needs to be filtered out. We start quite simply with all the existing process and machinery data, from which we first learn the machinery behaviour offline. We then boil it down to the few pieces of data that tend to be required to actually understand the machine, and configure an application-specific analytics engine depending on the type of machine. This entails software that can monitor and predict the machinery behaviour. That’s why we also refer to advanced analytics. Because it is not just about analysing historic data, but also making predictions based on this data.
WIN! In other words, you gaze into your crystal ball?
Gaukstern: That would be an overly speculative way of putting it. Rather we to use tangible mathematical and statistical processes to describe and analyse the machinery behaviour. The processes within the machine may well be complex – but with the necessary tools they can be predicted without having to be a clairvoyant.
WIN! Publishing data is often a sensitive issue. How do you deal with this?
Matthesius: The data always belongs to the customer and protecting that data is our prime concern. Our industrial security routers with VPN technology create secure communication paths. They ensure that different communication networks harmonise smoothly and under the strictest security standards.
WIN! What role does the user play in your solution?
Gaukstern: Users are the centre of our analysis – their know-how is extremely
Tobias Gaukstern, Industry Portfolio Manager Factory Automation
important. The analytics engine may well be able to predict a fault with a certain probability – but to do so it must have been classified beforehand. Only the user can assess whether an anomaly must actually be classified as a critical fault.
WIN! Advantages such as preventing downtime through predictive maintenance primarily relate to the operator. In what way does the machinery manufacturer benefit?
Gaukstern: The vision in machine construction is an entirely new customer relationship based on a changed data-driven business model. Currently, if a machine stops working, the operator must call the manufacturer, who then sends out a service technician. This relationship will be reversed in future, whereby the machinery manufacturer will contact the customer proactively and point out a current problem or an imminent fault to them based on the analysed data.
“Advanced Analytics” means generating valuable information and consequently clear added value from machinery and process data
Matthesius: With customisable remote maintenance solutions based on u-link, our web-based remote access service, we now offer machinery manufacturers the necessary tool for tailor-made services which cover a machine’s entire life cycle. Machinery manufacturers can also further develop future machinery models very specifically because they know exactly how their machines are used and what demands are placed on these machines.
WIN! What makes Weidmüller stand out as a partner for advanced analytics?
Matthesius: We know our customers’ applications and are therefore able to offer specific analytics functions for machinery and plant engineering. What is special in this respect is that we do not make predictions for individual components – the particular manufacturer is best placed to do this. Rather, we combine all the available information and are able to obtain a comprehensive, holistic view of the machine. Our customers benefit from advice, engineering, software and hardware, which is tailored to their individual needs. The analytics engine can be used both on our automation components and in the cloud. What the topology ultimately looks like is decided in close dialogue with our customer. Our aspiration is not a standard package, but always a flexible, scalable overall solution.
Successful pilot project with Ferag
Advanced analytics allows the behaviour of a wide range of applications to be monitored and predicted. Weidmüller has already tested this in its own production using an injection moulding machine and is currently conducting initial successful pilot projects with its customers – such as with the Swiss Ferag AG, global specialist for intelligent processing systems for newspaper and magazine production.
“We were surprised by the information that can be derived from the machinery and process data and what value is ultimately concealed in that data. Together with Weidmüller, we developed a solution to implement predictive maintenance. The combination of Weidmüller’s analytics and application know-how and our knowledge of complex conveying and processing systems in the graphical industry were key to the success of this project. On the basis of the results, Ferag will put
With intelligent systems for newspaper and magazine production, Ferag AG is making an important contribution towards ensuring that printed publications reach their readers on time
together new data-driven service models, which will shape our business models as well as our relationship with customers over the long term.” – Uwe Groth, Management Ferag Electronic
You might also be interested in:
To the Indsutrial Analytics Webspecial
Industry 4.0: Heading towards the Smart Factory together
Published in April 2016
+49 5231 14-0
+49 5231 14-292083