IoT doesn’t stop at any machine or industry: The 55th episode of the IoT Use Case Podcast is about industrial shredding machines from UNTHA shredding technology that have learned to “talk” with a digitization solution from A1 Digital. Now they can perform condition monitoring and remote maintenance with real-time data from the cloud. These are just two skills from the machines’ newly learned skillset. With the help of engineering and technologies such as modems, SIM cards, gateways, clouds, edge devices, artificial intelligence, etc., a number of other functions have become possible for machine manufacturers and users. Sounds complicated? The podcast gets to the heart of everything – challenges, solution, added values – are explained in an understandable way!
Podcast episode summary
The A1 Telekom Austria Group is an internationally active expert in mobile connectivity and digitization. The global IoT business includes application fields such as predictive maintenance, asset management or machine learning. The base of 24 customers in Central and Eastern Europe is diverse: companies from the construction and mobility sectors, but also from the manufacturing industry. A1 Digital brought an important customer to this podcast episode and explains its industry-specific solutions with a success story from practice.
UNTHA shredding technology has been developing customized shredding machines in the premium segment for the industrial sector for 50 years. The machines are used in landfills or recycling plants. Topics that move the company: the recovery of valuable materials in recycling, the utilization of waste wood through to the processing of waste – all in the interests of sustainable resource conservation. The aim is to achieve the best possible use of energy while increasing machine availability and reducing downtime for the customer. For this purpose, there are now up to a hundred measuring channels on the shredders, which make data available in real time, such as the speed of the rotor, temperatures in the oil or in the gearbox, vibrations, oil quality or the quality of the power supply. An important KPI of this business is also the cost per ton of crushed material.
To hear how A1 Digital helped UNTHA shredding technology create a new business model and new services, listen to this podcast episode. The guests talking about this on Madeleine Mickeleit‘s microphone are:
• Florian Krcma (Presales Consultant Digital Business Solutions, A1 Digital)
• Philipp König (Product Marketing Manager IoT, A1 Digital)
• Robert Diosi (Product Manager, UNTHA shredding technology)
Florian, let’s start with a brief round of introductions, perhaps beginning with yourself and the core business of what exactly you do at A1 Digital.
My name is Florian Krcma. I’ve been with A1 Digital for just over four years now. I’m in the role of Presales Consultant in the VMS department – that stands for Verticle Market Solutions. As the name implies, our goal is to align the portfolio items as we offer them to meet the diverse needs of different industries. This means that we use a modular system for IoT solutions, consisting of hardware components, connectivity and then, of course, platforms. Whether this is an IoT platform or perhaps a machine learning platform based on it depends entirely on the project. My role is to find the right elements for our customers to best support them on their journey.
Then I, Philip, would hand over to you. Maybe you can also say a point or two about yourself and add a little bit to what Florian just elaborated on – what do you do at A1 Digital?
I am an economist by study and after graduation I spent several years in marketing and partly also in controlling in various companies. This was in the industrial, mechanical engineering, IT and telecommunications sectors. That’s why this mix fits quite well for what I’m doing now: I’m the Product Marketing Manager for IoT at A1 Digital and work quite closely with Florian. We, A1 Digital, are part of the A1 Telekom Austria Group. As you said, a company that is present in many countries. We have over 24 million customers. The A1 Telekom Austria Group itself is majority-owned by América Móvil – one of the largest mobile operators in the world. This means that our expertise in mobile connectivity is very high, also in the area of digitization. We also say we make digitization work for our customers. Our focus includes industry-specific applications: In the area of Internet of Things, i.e. IoT, in the area of security solutions and in the area of cloud hosting. In the latter, for example, we are also known as Exoscale, especially in Switzerland. We are a European cloud provider with hosting in Europe and are also a founding member of GAIA-X. That was a brief overview of our portfolio. Perhaps a few sentences about our very diverse customer base. These include companies from the construction sector, but also mobility and manufacturing companies. We are joined by one customer today: UNTHA shredding technology.
Exactly, especially with regard to these industry-specific solutions and what you bring to the table from the core business, how that works exactly, that’s where we want to delve a little deeper today. That’s why I’m very pleased, Robert, that you’re with us today from the company UNTHA. Would you also like to introduce yourself briefly and say something about the company, what you do?
I work for UNTHA in the field of digitalization, i.e. development for digitalization. I started equipping UNTHA’s machines with sensors and digitalization about three years ago. I originally come from the field of mechanical engineering, having studied mechanical engineering and automation technology. Mechatronics is also the area where I am responsible. About UNTHA: As you said, founded in 1970, so we had our 50th anniversary last year. The company name UNTHA is composed of Unterwurzacher and Hasenbichler – these were the founding fathers. We are based at our headquarters in Kuchl near Salzburg and have approximately 300 employees. We develop and produce customized shredders for the industrial sector; we are not in the home application sector, but really in the industrial machinery sector. Here, it is very important to know the condition of the machines. It is about the recovery of valuable materials in recycling, about the utilization of waste wood up to the processing of waste. That’s about where we are in the industrial sector. We actively contribute to the conservation of resources – a very topical issue – and to the sustainable use of energy.
Shredder, I can already imagine something about it. Can you also give an overview to the listeners out there, what does a shredder like this actually look like and what is your daily business like?
Yes, in essence, there is a rotor inside the machine. There are single-shaft systems up to four-shaft systems. At the top is a hopper, where you throw coarse material in, and at the bottom it comes out in a certain grain size; this is called a fraction. Depending on how the customer wants it, the fraction is sometimes larger and sometimes smaller. A very important issue here is also that we are in the premium segment with our machine.
In other words, you have to imagine the recycling business in a very classic way? The machine is located somewhere on site – are they dumps then, or where are they located exactly?
Exactly, that is, for example, a garbage dump, or in a hall where recycling operations are located. This can be worldwide.
When I imagine a recycling machine like this, there’s an insanely large amount of energy behind it. What are such typical savings opportunities in the operation of this machine? What are potentials that you want to leverage? Also against the background of our topic IoT.
It’s all about making the best possible use of energy. We are very far ahead with our technology. The important point is to increase reliability and reduce downtime. Efficiency is, of course, a significant aspect. This means using the energy as well as possible and crushing the material.
Challenges, potentials and status quo - This is what the use case looks like in practice
Then I’d like to hear a little bit about the day-to-day and the challenges as we think in terms of IoT technology. You said it’s about reliable operations, downtime. What are classic challenges you encounter in everyday life? Does a customer call and say the machine is down? What typical challenges do you face?
The biggest challenge at the beginning three years ago was that we didn’t know anything about the machines that were being used all over the world. So it was a very important point that we bring connectivity and data acquisition into the machines. So that we – that is, not only we as a company UNTHA, but also the customer – know as best as possible about the condition of the machine at any time. The reason for the error can of course also be queried. That means, for example, you can look at where did downtime come from in the first place? Which values spiraled out of control that caused the machine to come to a standstill? In this way unplanned and expensive downtimes can be reduced as much as possible. In this business, the cost per ton is an essential aspect of such a machine. There are operating costs, energy costs, everything that has to do with the machine. We then tried to get to grips with that as best we could with digitization.
You’re talking about unplanned faults, or faults in general. That is, today a customer calls and says, here somewhere in the process – probably in the upstream and downstream process – an error has occurred, and you would then normally drive out or help the customer by phone to fix this problem?
Yes, you can imagine it roughly like that. The data is always available in real time in the cloud. This means that we can also look with the customer at what the current condition of the machine is, and can then also be available to him remotely in the best possible way and help to fix the fault.
You told us a bit about how a machine like that roughly works; there are forces at work. And what data does your machine provide that are of interest to you?
On our shredding machine, we have up to a hundred measurement channels available in real time. These include, for example, the speed of the rotor, temperatures in the oil or in the gearbox. Energy consumption is very important, and the performance of the machine as a whole is also a very important aspect. As I said, we have a hundred measurement channels. It’s also about oil quality and quality of power supply. But also vibration, the main drive train, is quite important. If I now imagine that you have the machine at the customer’s site.
Maybe not digital in the past; now you’re moving toward digitization and also taking advantage of the data potential that a machine like this brings with it. Engine speed, engine, oil – these are all analyses that I have to make. I don’t know if this is then vibration analysis or also data pre-processing? How do I have to imagine this on the machine? Do you have a gateway somewhere on the device itself, which also passes the data upwards?
You can think of it as pre-processing already taking place in the machine. We have a fast measurement system called “condition monitoring”; an edge device. This is like a small industrial PC installed directly in the machine, and pre-processing of the quickly measured values already takes place there. Then only the relevant data is uploaded to the cloud.
Solutions, Offerings and Services - A Look at the Technologies Used
I would now like to talk about the solution. We learned there is an ePC, an edge device, that placed on the machine. Now I have to transfer this data to the cloud somewhere. Florian, how exactly does this work from a technical point of view? How do I record this data and what do you bring to it?
Basically, there are different ways to proceed. I mentioned at the beginning what our IoT portfolio usually essentially consists of. The first point I mentioned, about hardware, that’s exactly the relevant entry point. This means that there is a gateway somewhere, but it could just as well be a built-in sensor. The only thing that is relevant here is that you can get data from somewhere. Either determine them yourself or, should they already be obtained elsewhere, transmit them. That means we have different approaches, industrial gateways, gateways in general, in order to be able to connect to machines via the corresponding interfaces. There it is only very essential that there is simply the authorization for it. So it has to be open interfaces. One must actually be allowed to access this data. If that is the case and you have been given access to this information, there are two options. One is that, as Robert also just described, there is already some pre-evaluation of that data on the device itself. That is, this information is interpreted, certain algorithms are already running on the device itself and tell me whether this data is good or bad data, to put it banally. Only these findings are then transmitted to my platform, so to speak. The alternative is to go in and say, all the data that I’ve previously obtained is going to be transmitted to my platform, where then exactly what might otherwise happen edge-side is happening on the platform. This means, for example, that vibration patterns are analyzed, hydraulic pressure is evaluated, temperatures, noise, humidity. There you are very free to choose the relevant information you want to look at. Then you have the possibility to decide on the platform, what kind of consequence should one or the other event have for me? Do I want to access the machine directly? Do I want to maybe start or stop something? Or maybe I don’t want to access the machine itself, but just want to let someone know that there might be some unplanned maintenance here? These options exist. In other words, these three elements can also be found here: hardware, connectivity and platform. One major difference, of course, as far as the different approaches are concerned, is the amount of data transferred. This can be relevant for countries where you would have high costs for data transmission, for example. But of course, here you try to go into optimization as much as possible.
Let me get this straight: One has a hardware area. This would be the data recorded by the modem. In that case, UNTHA has its own modem; but of course, you have modems that you use if customers don’t have that, for example. Then I have the connectivity area; a SIM card that is installed accordingly. And on top of that the cloud level where the software runs and then all these issues Intelligent condition monitoring…
… you had given the example with the vibration patterns – that’s the top layer, so to speak. Have I understood this correctly?
Exactly. One is then on the IoT platform. What you ultimately do with the data on this platform is, of course, up to you. Very many of our customers use this data lake for some visualization. Others – and this is exactly what we offer in conjunction with the corresponding algorithms – utilize this data, which is already moving in the direction of machine learning, whereby interfaces are used to tap into this data, evaluate it and then feed it back again.
Robert, you said you have a modem on the machine, or an edge device. You are probably using the SIM card. Can you tell us about exactly how that works in practice with what A1 Digital brings to the table?
We have now equipped about a hundred machines in the field with digitization, and a modem is installed in each of them. That’s sort of the interface to the outside world to the cloud, so you can think of it that way. In addition, an edge device is installed, i.e. a small industrial PC where all data channels converge and are aggregated. There has to be a central point and fast measurement technology that also calculates mathematical functions – this data is then merged in the edge device, and it finally ends up in the cloud via the modem. An IoT SIM card is installed in the modem. The special thing is that this is a SIM cards pool. We have machines in operation all over the world, so it is important to be able to switch SIM cards on and off and possibly perform maintenance operations directly on the SIM card. We have our own platform for this.
Florian, I have now the machines are probably distributed all over the world; in the USA, China … how does this connectivity work across countries? These different SIM cards must talk to your cloud, right?
Exactly, as Robert has already indicated: There are flexible options as far as the connectivity issue is concerned. We know of examples where you approach it very specifically and say you have operations in North America, then you might still have a certain customer focus in Southeast Asia or Europe. This means, of course, that you try to be as close as possible to the respective use case and design the tariffs according to this stored data. Depending on how large the expected data volume is – whether it is a few megabytes or gigabytes – different rate plans are put together. They don’t lump everything together, but actually create a separate pool for North America, for example, and exactly the same for Southeast Asia and then again for Europe. Alternatively, of course, if you say you have a certain focus – let’s say 90 percent in Europe, 10 percent perhaps spread over the rest of the world – you can just create a single data pool. What we always just pay very special attention to is that because of a few countries that may be very expensive, you don’t add up these mixed prices. Then, of course, there are some very specific countries that could become commercially uninteresting relatively quickly if approached via the mobile network. There, the approach is often to use the locally available network and not to use mobile communications at all. Now that I mention it, this is sometimes the preferred option anyway, not to primarily use the mobile network, but to use it at best as a kind of fallback in case the local network supply should no longer work or is temporarily interrupted.
Robert, in which countries are your machines located?
You can basically say Europe is a very big focus. But you can find us all over the world. For example, we’re in North America, we’re in India, we’re in Australia – so that stretches around the world a bit.
Theoretically, I would have to sign contracts with every local mobile provider to get my data to the cloud – normally. The bottom line is that it saves me the work if I have access to A1 Digital and can connect all my machines to this network via you, where you have already taken this work off my hands, right?
That’s exactly how you can think of it. It is a SIM card for the whole world. That means you don’t have to know beforehand where the machine is going – that will only become clear later, where it is sold to. Works then in any case, because the SIM card kicks in and can provide that there via roaming.
Results, Business Models and Best Practices - How Success is Measured
Florian, you offer more than just connectivity. You just mentioned, there is the layer where the data volumes are processed, where the value creation and data access really takes place. And you brought up your platform. How exactly does this visualization work there?
Basically, it’s probably best to distinguish between back-end and front-end. When we’re on the front end and the data is coming in in raw form, there’s an opportunity here to visualize that data with standardized tools. That is, if I now – again, for example, with the vibration pattern – get the data, the corresponding timestamps, there are so-called widgets. With these widgets it is possible for me to create an initial visualization. That means, I can display a certain graph, a speedometer or even revolutions. These are exactly the kind of tools that I don’t have to develop myself; the platform provides them out of the box. I’d say the majority of our customers make do with it, because a graph is actually used almost everywhere. But there is still the possibility, if he has a very specific requirement that perhaps could not be mapped with these widgets, to also initiate the corresponding development on the platform or to develop his own microservice on it in order to be able to interpret this data very specifically.
We have a Demo Center on our website A1.digital. We have freely accessible demos of the IoT platforms there. For example, you can look at a demo there called Asset Insight, or Industrial Insight. There you can see some examples, i.e. mock-ups, of how such an IoT platform works, which devices can be connected there and also what exactly you can see there.
Perfect, that’s very good that you can see the issues live right here. Robert, the question to you, you guys are a little bit more broadly positioned and you have, I think, also built up your own business model with it. You have a custom platform; at least you can see that online a little bit. What exactly are you doing there; how are you using this platform?
This customer platform – tailored to the customer – is called MyUNTHA. The customer can log in with his password and customer name and then access his data, i.e. real-time data and also historical data. For example, you can generate reports from the last week, last month and so on; you can also have time-dependent reports sent to you by e-mail on a weekly basis. But this goes even further, for example, that you can also find operating instructions for the machine in it, all alarms and messages that are relevant. In the future, there will also be a kind of online store there where customers can order their own spare parts.
Our audience comes from very different backgrounds that are also mechanical engineers. Or rather, I always hear similar issues. The things you provide there are exactly what the end user is interested in – having reports, instructions directly available online, which used to be kept in paperwork or had to be generated manually. This brings an insane advantage. Florian has just spoken about timestamps. Robert, what would that mean in your case? You said that you have, among other things, a measurement system, which means that you set certain timestamps, so to speak, which you then map in the cloud, in terms of data points – or how does that work?
Exactly, we also have historical data in our cloud. You can imagine where the data is captured, it all has to be synchronized. That is, the hundred channels should be related to each other, so that you can also perform analysis. That’s why it needs timestamps in any case. They are also very important in that they are internationally intertwined. We are also talking about time shifts. That means it’s a different time in Mexico or in Australia; it all has to be manageable. A very exciting topic.
Florian, you just brought up the topic of Exoscale, which is sort of, if I understood correctly, your cloud computing platform. What exactly is it all about?
Exoscale is our cloud environment. It is, as we call it, the European Cloud – after all, of course, the question always comes up when you talk about platforms when it comes to data processing: Where is the data located? What assurance do I have as a company that this data is not only secure, but also protected from access by different governments, for example. We also operate Exoscale because we ourselves want to be able to offer our platforms in such an environment, simply to be able to make this commitment that we guarantee where this data is located and where the servers are. That’s why it’s exclusively in Europe. We have data centers in Frankfurt, in Switzerland, with us in Vienna, in the Arsenal. All platforms run there. We are of course talking primarily about the IoT platforms today, but that is just as true for our SIM management platform and just as true for our machine learning platform – so that is all within Europe.
Robert, cloud security is a big issue. How do you see it; how do your customers see it?
That was a very important point for us from the very beginning, that the data is stored somewhere in Europe; not somewhere in the Asian server area, but really located in Europe and protected as well as possible. Another very important issue is that the data cannot be misused. It’s all about performance between customers and so on. Under no circumstances should there be a breach that allows one customer to read any data from another.
A leap into practice; I always like to talk about best practices as well. If I want to start tomorrow – I think it’s exciting, I’m also in mechanical engineering, I want to do it now. Exactly which components of the solution do I need? Florian, can you tell us more about that, from hardware to connectivity to cloud?
That’s a great question, of course, and of course you always want to know, how do you get started? This differs massively from customer to customer. Philipp briefly mentioned the Demo Center. Not to advertise again, but this just gives a first overview as far as visualization is concerned: Often people say they would like to have data on an IoT platform here, but they don’t really know yet what kind of possibilities this can open up. What forms of visualization? What forms to be able to gain some knowledge at first glance? If you’re reasonably clear about this, or even say, I might not know yet, but would like to launch the whole thing as a test balloon: Then the recommended approach is not to try to launch the perfect system already. But that you start as small as possible. There are customers who already have a lot of experience in this area, so you might start somewhere else. And then there are customers who say, I just really want to try this for the first time and see where it takes me. We offer different approaches in this context, depending on where the customer is on his journey. When it comes to getting data on a platform and gaining initial insights, we usually look at the as-is situation, what kind of machines are there? Is this from one manufacturer or are they from different ones? Is this machine already smart? Might sound a bit strange, but not everywhere this data gain is already possible locally. That means, would data have to be obtained here in the first place, would we perhaps even have to work with one or the other sensor technology as a supplement? Or is it really just about tapping into what’s already there? So to put it more simply, the first approach is usually, how do we get the data from a handful of machines onto the platform? A kind of test environment, test instance, is then set up on this platform for a defined period of time. The data then arrives on it, and the customer has the opportunity to develop an initial feeling for it. Of course, we provide support when it comes to how to get this into a form where you can say, you can actually decide with this, this is successful and I would like to continue. That would then be the impetus for saying you’re moving further from your proof of concept here towards a possible roll-out.
We had talked about the topic of IoT devices. That’s where I would recommend: Since October, we have launched an IoT Center on our website A1.digital, which can also be found in the Demo Center. There we describe the concept, i.e. which modules do we generally offer? Hardware, connectivity, IoT platform, cloud. And also our expert services. This is for companies that really want to have an end-to-end solution. But what we still have there is a so-called IoT enabler kit. What you can order there is in fact a package consisting of a development kit with a chip from Nordic Semiconductor, our SIM card and also access to the IoT platform. That means you can order this package there, you get tutorials as well as support from us, also for an onboarding, and in principle you can also connect sensors et cetera to the development kit. You get access to the IoT platform and can integrate your devices there right away.
Thank you for the addition. Robert, then I would take up the question: How did you start in practice and with which components from A1 Digital?
For us, this was a relatively simple undertaking. The interfaces to A1’s cloud are very well documented. In the beginning, I tried to send individual channels, temperature or anything non-critical, to the cloud. It was immediately important that each machine has its own password and identifier. This means that the security of the transmission was already very good. I also looked at other cloud systems, and that’s where A1 won with their technology. And once you’ve started with one channel, it’s relatively obvious that you’ll use a second; a speed or some other relevant measurement of the machine. Then you keep increasing like this and at some point you have a hundred measuring channels and alarms and messages, which has finally resulted in a total system.
I’m also always interested in what is the business case? After all, the point is to generate added value in the end. We want to leverage potential through data. Robert, what is the business case for you guys? You have built up your own business model with it.
For us, it is definitely the reliability of the machine. This means we can proactively address the customer when needed. It’s all about aftersales issues. It’s about quality issues. It’s about warranty issues that come into it as well. That’s a very, very wide area that we’ve been able to open up with this.
In other words, you also use it to optimize certain processes and to further improve quality and reliability. You are thus also an innovative pioneer who is leading the way and simply offering the customer an additional service.
You can definitely say that. In addition, customers can configure notifications themselves, for example. This means that if a fault condition is brewing somewhere on the machine, they can be notified via SMS, for example. This can be done in real time. This means that the machine has as little downtime as possible.
What comes on top maybe, service technicians … a customer calls, you have to go on site – you save time too, right?
Exactly, maintenance operations are now much more plannable, of course. This means we can set routes and then see which machines we approach first and which are in sequence in this route. That’s also a really great side effect of the whole thing.
Transferability, Scaling, and Next Steps - Here's how you can use this use case.
Florian, we have now discussed this one project of yours. At the beginning, you and Philipp mentioned that you are positioned quite broadly, with different customers. How can this use case be transferred? We probably also have mechanical engineers with us or other companies from completely different areas. What other use cases do you see? How might that be expanded as well?
The use cases that are mapped are actually very different. What they nevertheless have in common, what they have in common, are essentially the core elements. That is, we talk to machine builders where it comes to data transmission, information transmission. In the same way, we have customers who come from completely different sectors, from completely different industries. Where, for example, it also goes in the direction of Smart City. In fact, it is always essential to find out what kind of information the customer wants. What kind of decisions should be made and how can the portfolio as we have it support this? That is, the beautiful example of UNTHA shredding technology shows how something like this can work – but whether it’s a shredding system or a machine for wood processing, for example, or a production line, I have to say quite honestly, the elements are often very similar. It’s either how do I get that data from my local system to my platform – that’s where we’re back to the industry gateway. And what I alluded to earlier, if there are machines, for example, that don’t yet have the ability to acquire this data, you can of course think about retrofitting. This means that you may already have some of your machines smart, others would have to become so first, and you can still create a common data basis in order to then be able to map the complete truth on the platform side.
Philipp, you had already mentioned the Demo Center. What can I see there exactly? What can I do?
We have mapped some use cases in our Demo Center. For example, there is the Asset Insight or the Industrial. This shows what you can connect to machines like this in the industrial environment and what you see there. And we briefly mentioned the IoT Center there earlier. There you also have the possibility to start an individual demo for 90 days. There, you get your own individual access to our IoT platform and can, in principle, connect your own devices.
Thank you very much!