Machine manufacturer, MPS Systems designs and builds printing presses for the label and flexible packaging industry. The mission of MPS is to enable the operators of their machines to deliver the best quality and productivity. This is only possible if the machines are used optimally. But how do you ensure that your customer obtains maximum value out of his expensive machine? Data is key here.

MPS: Challenges in the market call for new services

MPS began its ‘data journey’ a few years ago. The company started looking for new ways of providing services, with which it could offer more value-added services to customers, , thus securing long-term relationships and business continuity. Atze Bosma is CEO at MPS Systems and the driving force behind this strategy. Atze: “MPS operates in a niche and global market, big yet , with a limited number of competitors. In terms of price and quality, we are pretty much the Rolls Royce or Porsche of the market. But we have been experiencing price pressures for some time, partly because the technology of the various manufacturers is becoming more comparable. Should we want to escape this, we have to create new markets or think about alternative service delivery methods This is where we saw a huge opportunity.”

MPS helps their customers to perform at a optimum levels

Atze: “For some time, we suspected that our customers were not using their machines optimally and were not as efficient as they could be. Which is actually quite strange. After all, we sell an expensive machine that has to run at high volumes, with many changeovers. If that does not happen, the customer does not get the benefit, the efficiency and quality and ultimately the potential value out of it. That is where we believe there is an opportunity: if we help our customers to perform at peak performance, then we can distinguish ourselves in the market.

We expect that our customers will also benefit greatly from this. Many of these companies still have a relatively stable and comfortable market position. But risks like price pressure and consolidation also play a role in their market. Plus the demand for, for example, sustainable biodegradable products, biodegradable ink and plastics is increasing. Can they meet this demand? We can help them with this challenge.”

Pioneering with data

To test the suspicion about the inefficient use of the machines, MPS decided to install measuring equipment, in order to collect as much information as possible about their usage. Rob Nijland is Manager ICT and Software Engineering at MSP Systems, and is responsible for both IT and Controls Engineering. In other words, the department that manages and develops the hardware and software around the machines. Rob has been closely involved with the developments around data and services from the start.

Rob: “To enable our customers to be more efficient and productive, we needed to know more about the use of the machines. We started very pragmatically, by equipping our machine with a device to collect all kinds of usage data. So how is the machine operated, how often does the valve open, when does the roller change, and so on. In this way, we collected a lot of event-based data, which we then used to make our own analyses. We set up a development environment and started making analyses and the first visualisations.” Since January 1, 2020, every machine has been equipped with a Microsoft IoT edge computer, allowing MPS to substantially grow the database with relevant data.

Atze: “The suspicion that each operator operates the machine differently was indeed confirmed by the data. Setting up the machine depends on the operator and each operator does this differently. This has a lot of influence on the production, possible waste, production time and so on.”

MPS takes next step in IoT applications

Atze: “We found out that there are still very few companies that think and work in this way. So, we had to get going ourselves and look for expertise to test our hypotheses. We expected there to be companies that were already a lot further ahead. But that turned out not to be the case!

We see that the IoT technology that is being applied is still very much focused on ‘high volume and low intelligence’ and mainly on spotting deviations in a digital process. This is perfectly suitable if you want to optimise maintenance and stay ahead of failures. What we are doing now is therefore still relatively new in the IoT world, and that is mainly due to the human component. That results in a lot of logic in the models, think of ‘if… then’…. When discussing with customers, we try to test these patterns with expertise from the field. But this is a lot more difficult than making a report based on simple measurement data and signals.”

Rob: “Our data factory is about interpretation. We are already thinking more in the direction of machine learning to look for the logic. We are looking for patterns, but we are dealing with so many variables that it is very complex.”

Looking for a new business model

Atze: “Ever since we started, we had to start over again three or four times. Setting up the data structure, choosing software and so on. We have now reached the point where we are talking to customers and starting to use our potential. But the circle is not complete, yet.

After all, as far as we are concerned, data and reports do not make a product. You only have a product when the customer trusts us to think about potential improvements to their machine, based on the view that we are not necessarily the better for it, but that the customer is the better for it. We will only be proud if our machines are making a lot of money for our customers. After all, a company will then continue to hire us and our relationship and continuity will be guaranteed.

Our goal is to learn from the data and thus develop our own benchmark for optimal use. By combining data and knowledge about our machines, we think we can then significantly increase functionality. At that point we can guarantee service, for example, and then you move towards a revenue model such as a price per metre.”

Rob: “All the information we collect can also be a driver for further R&D developments. Because we see that, for example, the interface can be improved. So there is a lot of value in the data on many different domains.”

Collaboration with HSO for stable data architecture and continuous data flow

Rob: “We had already made a lot of progress, but ran into a number of bottlenecks in the data structure. Then we contacted HSO. We first did a review: what do we have now, where do we want to go and what is needed to solve the gaps we have? Then we set up a new architecture in Microsoft Azure cloud. We now equip every machine with an Azure device. The result is a continuous data flow, based on a stable architecture.”

“What we liked about the collaboration, is that HSO has various disciplines in-house and therefore serves a broad playing field, also in terms of technology. We had an expert in the field of data disclosure in the team. Next, we hired someone who knew everything about Databricks, our analysis tool. And right now, we are working on a portal based on Microsoft Power Apps. It is very nice that there are short lines of communication, despite the fact that HSO is quite a large organisation. This helps to quickly switch and arrange things. This broad playing field and expertise across various disciplines is definitely a positive point. It is very nice to work with one company that communicates well with each other.”

Atze: “We also experienced the team’s focus as very pleasant, especially in the initial phase. A lot came out of it in a short time. Being MPS, we do think it is important to look beyond the boundaries of the Microsoft stack. If other technology works better for our applications, we are open to that too.”

Looking to the future

How does MPS look to the (near) future? And what does the company expect to be able to offer customers in the future? Atze: “We want to visualise more than just the production part as soon as possible. It seems logical to continue with data about the speed of production lines, time and metres. We know all about that by now. But we want to offer more than just insight into productivity and move towards a holistic approach. Think of connections with the ERP system, managing the energy management of the line and looking at the quality instead of the quantity of production. But also think about improving planning and personally addressing or coaching the operator, so that we ultimately help them to do the job better.”

MPS: Getting past the ideation and POC stage

Many companies looking for new business models and new forms of service are struggling with this change. How do you get past the idea phase? Atze: “From the beginning, we knew very well what we wanted and that is crucial. You must ensure that you have a good understanding of your market and that you have a very clear idea of what your goal is. Because that is what you need to keep going, even when things are not going your way.

Rob: “You can chew the business case to pieces, but you must also dare to take the step somewhere. And you mustn’t want to close your business case completely. After all, you never know in advance exactly where you will end up. But the direction must be clear, so that you can take the organisation with you.”

Atze: “We have already invested quite a lot in data-related projects. But our shareholders, like us, do not doubt that this is the future. It is important, however, to include the rest of the organisation in this culture change. After all, developing, marketing and selling data is different from building and selling a machine. All 100 employees go through the car wash to make everyone data-minded. Curiosity has become an absolute core value for us. After all, it is only when you are curious that you go looking for information, and data becomes your driving force.”