Möchtest du unsere Inhalte auf Deutsch sehen?


OPC UA in Practice – About Connecting Old and New Systems and the Context of Data


Click on the button to load the content from Spotify Player.

Load content

Listen to the IoT Use Case Podcast on Spotify.
Listen to the IoT Use Case Podcast on Spotify.
Listen to the IoT Use Case Podcast on other platforms.

IoT Use Case Podcasr #94 - OPC Foundation, Stefan Hoppe

OPC Unified Architecture (UA) is much more than just a protocol – especially in the context of industrial IoT. Miele, Equinor, Rosendahl, the Renault Group and many more. – in this podcast episode, six company projects show how they use and implement OPC UA. For this with Ing. Madeleine Mickeleit’s guest: Stefan Hoppe, President and CEO of the OPC Foundation.

Episode 94 at a glance (and click):

  • [16:44] Challenges, potentials and status quo – This is what the use case looks like in practice
  • [32:23] Results, Business Models and Best Practices – How Success is Measured

Podcast episode summary

In this episode, the Executive Director of the OPC Foundation – Stefan Hoppe – reports on how OPC UA is used in practice in the context of IoT. A rare insight in an otherwise technical subject. Because: OPC UA is much more than just a protocol!

“The future of automotive production lies in the control, optimization and profitability of cross-site manufacturing based on secure, real-time data” Renault Group recognized this several years ago and actively pushed the integration of OPC UA-enabled components and systems.

In this podcast episode, you will learn how these companies apply and implement OPC UA with which use cases on the basis of 6 further projects from Miele, Equinor and Rosendahl, among others. Not just an issue for manufacturing companies!

“Connecting old and new plants today really costs work”
The tasks for many companies is the connection of old and new plants as well as machines and components. That costs a lot of time and work today. Keyword: Brownfield. Some of the machines are from 1970 – the other machines just acquired last year in the factory. Many of the older machines are still equipped with old Siemens S5 controllers or other controllers from Beckhoff, FANUC, Heidenhain, Allen-Bradley (Rockwell Automation) or WAGO controllers. Today, for many manufacturing companies, there is often still no uniform access to data from different control types! That’s why companies are working to make the data available across the board. Because: In the best case, the controls or other data sources should only be handled once and then made available throughout. In practice, this means that not every machine should have 30 applications and interfaces. “It’s not like every car has its own new fuel nozzle.”

Standard tools and standardized connectivity are needed here, because this is the only way to scale digitalization. Today, interfaces and information are often not available identically across manufacturers! Today, the information models behind them are also often not designed across industries so that a provider can seamlessly use the same interface and information model. The answer to that? In the podcast!

“Data without any context is pretty useless”
In simple terms, OPC UA is the modernization of data communication according to the latest state of knowledge. Data is useless without context. Sensors provide values such as vibration, temperature and associated time stamps. However, humans and software cannot know whether the data is attributable to a pump, motor or valve. Most importantly, whether the data point refers to hydrocarbons at the inlet or outlet. A data exchange where a human “translation” must be interposed. This results in a correspondingly high time expenditure for adaptations or the integration of devices and systems. The answer to that? In the podcast!

But how can data be made available consistently? How does standardized connectivity work? What role does OPC UA play for the IIoT? For which use cases is this relevant?

Podcast interview

Many of you already know and use it: OPC UA. But what many don’t know is that OPC UA is much more than just a protocol, especially in the context of IoT. Today you will learn about six projects – including Miele, Equinor, Rosendahl or the Renault Group – how they apply and implement OPC UA, with which IoT use cases.

Today with none other than Stefan Hoppe, President and Executive Director of the OPC Foundation. What the whole thing has to do with the yellow pages, you can find out now!

Hi Stefan, welcome to the IoT Use Case Podcast, I’m so glad to have you on today! How are you and where am I reaching you right now?


I prefer to be well all the time and thank you for having me. I’m working from home today, but I’ve already been on the road a lot. The pandemic seems to be over and we have an exciting mix of events in person and also hybrid events to save the earth’s resources a bit. Sometimes it does make sense to meet people in the room and inspire each other with new visions.

I’m also back at some trade shows and look forward to shaking some hands and seeing faces again soon. Where are working from home?


I am in the large town of Verl near Gütersloh, close to Bielefeld, even though some say Bielefeld doesn’t even exist. Up in East Westphalia, where many well-known automation companies are located and where I myself started my career 28 years ago at Beckhoff Automation.

Many of our partners, including soffico, in.hub, and others as well, often work and report on OPC UA as a standard that they use in many applications, and I’m glad you’re joining us in person today to talk from the field.

The OPC Foundation has been involved with the development and adoption of the OPC Information Exchange Standard for a very long time, since 2006, and you’re President and Executive Director of the OPC Foundation here; what’s that like in your role? What are your typical tasks and which clients do you work with?


We at the OPC Foundation have been working on the standard for a very long time, although you have to make a very fine distinction; we don’t do everything. I don’t think any organization in the world can claim to have the one hundred percent solution in today’s digitization challenge.

As a core competence, we take care of how to distribute standardized data with IT security. We work very closely with domain experts, from other partner organizations and they set the actual standardized data. Everyone benefits, not only the end customers, but also the suppliers if they sell via a standardized market. The market is just bigger than when I sell proprietary stuff.

And you guys are huge, you have hundreds of members.


We are very big, yes. We are very proud, we just welcomed the nine hundredth member. That’s “Procter & Gamble,” by the way, to put it in a nutshell. Every one of us, when we go to the bathroom and look at the package and tubes, we will often see Procter & Gamble as a brand. They have a lot of sub-brands that we’re not even familiar with, but it’s a giant that produces a lot.

Also Equinor, Norwegian oil and gas supplier. Foxconn, one of the largest manufacturers in the world with over 1 million employees Samsung, Volkswagen, these are all members of the OPC Foundation.

President and Executive. As president, I represent the organization and also work with others on future visions. Where is it going? What should we do? And as a Executive Director, it’s the same as for you, it’s operations. There is a budget, growth of the organization, cooperation with other organizations.

There is also a legal part, which I find less exciting, I say that quite openly, but that has to be and there I am also very proud of it, because the OPC Foundation stands out from other organizations. We do not only protect our own members, but all members have decided and signed that they also protect the integrators and the end users if they have OPC UA technology.

Exciting task!


This is an extremely great aspect! The Who’s Who of automation is a member of the OPC Foundation, but so are small IT companies like Amazon Web Services, Google, IBM, Microsoft Azure or SAP. If you think about how many patents they have in their patent pools, with all the big OT companies, the Siemens, ABB, Schneider Electric and others is enormous.

One might also ask, what is an AWS doing in an OPC UA world?


To put the icing on the cake, where my heart swells: by chance I met someone on some trip in China and learn that a small company like Foxconn with over 1.2 million employees is one of the largest consumer manufacturers in the world and they run OPC UA information models in their CNC machines around the world – they have 110,000 CNC machines; they say so.

I flabbergasted and asked myself: why are they doing this? They’re such a giant, they could also do their own Foxconn standard, but they use that themselves to transfer data to their Foxconn Cloud and they want to grow with the Foxconn Cloud into Europe and other areas as well, so standardization makes sense.

I am already very curious to see how the whole thing works in practice.

Can you share what use cases you guys deal with in practice?


Many of the success stories we have on the OPC Foundation website, but then they are in English and they are also more technical. We need to get better at highlighting the actual benefits.

Essentially, it’s that there are quite a few different international companies. An example from the automotive industry, discrete manufacturing: Renault Group is pushing data to Google Cloud, and they have a strategy for how they’re going to get data-driven digital transformation in their factory. This is also a very big aspect in Germany, because we still have a lot of manufacturing here, thank God. How do we get this step right now?

It’s not a small one either, Renault Group.


Renault is not a small one, it’s a whole group: Renault Group, to be precise. There are other areas as well. Equinor has created new oil and gas platforms. We all need energy, that’s just the way it is and there you have more than a million data points defined.

What makes me happy is that they use OPC UA, but everything that is missing and that has not yet been defined in terms of semantics, they have done themselves, but also offered publicly as open source. There’s also this idea: we take something from the community, but we also give back so that others can benefit – is quite strong. Equinor, by the way, delivers data to Microsoft Azure.

Totally interesting that you’re so much into Azure and also Google Cloud. You might think you’re very hardware-heavy, but that’s not the case at all; it’s totally exciting to see how these projects merge with the IT and OT worlds.


I do that deliberately, because what you say is a perception, a pre-determination in the brain of many. OPC UA is the last mile to the physical asset, I can physically touch that, to the robot, to the PLC, that’s in the heads. I want to use these very success stories to make it clear that data must scale.

There’s the real struggle between the OT and IT worlds, which makes everything a bit more complicated, that’s also worth its own podcast.

Just a month ago, we shared a Success Story. I didn’t know before that OPC UA is connected to the largest Green Energy Plans, i.e. the largest solar plant in the world in Egypt and to the largest wind turbine park. Two weeks ago I was in Vienna. I thought it was incredibly great that the CEO of Rosendahl Nextrom, as the leading machine builder in his field, invited us to the Industry 4.0 platform in Austria. I was invited and heard a fascinating presentation on OPC UA and all the benefits it has brought.

You have developed real fans. I had always thought that you guys were very much in the manufacturing industry, but that’s not the case. These are also other projects, for example, oil and gas, also solar. These are also other industries that you are in.


Absolutely! By the way, OPC used to be called something else: OLE for Process Control, so it started in the process industry. And in fact, as few people know, OPC UA is a member of the end-user organization OPAF, which is based in America and is a member of Namur Open Architecture. That’s the German process industry with big players like BASF, Bayer and many others; that’s the 4th largest in the world.

They have made OPC UA mandatory in the NAMUR Open architecture. MTP is another technology. These are technologies that have a different name out there because they are much more comprehensive. But they rely internally on OPC UA.

There is also a Companion Specification where an industry has come together for industrial catering equipment. There are 13 or 15 devices, also an industrial coffee maker or a deep fryer, for example. You can buy those here in Germany, a German manufacturer that has put OPC UA in, because they have exactly the same challenges in these areas. Different devices have to talk to each other; they have to exchange standardized data with integrated security.

Fast food chains don’t want to keep accessing different data and different protocols from different suppliers, so there’s a standard there.

We are just growing further into the energy world. We are already working on the largest solar project in the world. The Scatec company has modeled itself with other suppliers on the basis of OPC UA information. We have seen that there is a gap and we want to standardize this internationally.

Challenges, potentials and status quo – This is what the use case looks like in practice [11:56]

I’d like to get a little bit into the day-to-day life of your customers, so understanding who the contacts are that you talk to at Miele, for example, and what their typical tasks are.


Because OPC scales, it is not just the developer who has to write a protocol. You are one of the very few who just said OPC UA is not a protocol, but a framework or a technology to exchange information. Because it scales, a lot of people are involved. Then it’s the press plant manager, specifically at Miele it’s Mr. Frielinghaus, who was also involved in the Success Story. There is also someone from the IT world who is responsible for information technology, Christian Stickling, who has also appeared at the OPC Foundation press conference before.

There are also people who are responsible for engineering. We come together with several people, I hereby mentioned them and greet them and thank the people. I know Miele, of course, because I live around the corner here and my appliances are from Miele; after all, they last a long time. But this is a global company. They have eight production sites around the globe. In 2020, when Miele started, they sold over 6.3 million units worldwide. They have to be produced somewhere so that you can see the dimension.

To produce 6.3 million units, the plants must be designed accordingly. What are the tasks of such a Mr. Frielinghaus here?


We just take a washing machine and look at it from the outside. It first has a sheet and a lid and openings and so on. It all has to be produced somewhere. There arrives sheet metal, which must be punched. There are many machines that have to interact with each other. There are also people who have to have the right workpiece in the right place at the right time – as in any production – and wire it together.

This also means that you have a lot of machines and that you have connections. When you bring in technologies from old plants or from new plants, how do you tie them in? This also has something to do with government. So what’s the architecture like to put all this stuff together? As team leader, Christian Schickling has a lot to do with cybersecurity. How do I secure the production environment? At the same time I want to have a lot of connectivity, which the group needs on data.

There are an incredible number of tools that you have to bring in to plan a production, an MES system for example. Sometimes companies even have multiple MES systems because they have grown as well. They have production areas for different machines that they make and there are different MES systems for that. It all has to be brought together somehow.

It’s always easy to build a new factory on a greenfield site. If Tesla builds a new factory in Berlin, then they have the chance to do everything anew for once. A normal company has existing plants and also builds new production plants to them. We thus have both brownfield and greenfield areas. Having standardized information, that’s the important thing. In the end, it doesn’t really matter which technology is used. The main thing is to have this standardized information; it’s not easy.

These standardized data formats that are needed, can you explain that a little bit? What does that mean today? What are the challenges from your customers to build such standardized connectivity?


Especially things, if you have existing plants, brownfield plants, then I just have to remember, being of advanced age, my earlier days. I actually worked there as an assistant scientist at the Fraunhofer Institute during my student days. Among other things, I have programmed Siemens building controls.

When I started working at Beckhoff 28 years ago, it was really the case that if you did projects – and that’s what I did, because I wasn’t born a present – and then exchanged data in a factory from a PLC to an industrial truck, which brought in material, for example, then two people who were responsible negotiated a human contract. They wrote down on a piece of paper the data on which they define with each other the communication between the two devices.

Then it really went like this: In the data word ten, In the data block five is always my current position in it and if you want to give me a new order, then you set me please the bit seven on page 20 of the data block 30 and I acknowledged that with the bit nine and if that is set…” and so on. This only fits exactly between these two devices. If you did anything with another device, then you took another data block and another offset and did the same thing all over again; a lot of time.

Could you also say that if, for example, I have meter A and another and I want a temperature value, one says Fahrenheit, the other says temperature, that just has to be defined, something like that?


That absolutely! But: What do the data actually mean? When you and I are talking about a beautiful Jaguar, I might be thinking about the car and you might be thinking about the animal. Therefore: What exactly does temperature mean? But it also goes further, this is what we call “metadata”. What is the minimum value of something? What is the maximum value where I may specify what?

So I have to contextualize it to the underlying data.


Exact! This used to be all defined manually on a piece of paper, it’s hard to document, it’s always been different, there’s no standard for that. That’s defined everywhere, even today in industry, everywhere on these bits and bytes and on these human mappings and Excel spreadsheets. This is incredibly time-consuming.

This data must be retrieved from various devices, from sensors, from machines, from a wide variety of assets. Can you explain how this human translation, this mapping works today?


Today with these old facilities, it is that you look up in an Excel spreadsheet what the meaning of the data is. That is maintained manually, that is practically the contract. I send someone the excel sheet and write to them, if they want to read that, then they need to read that excel sheet and understand where they are allowed to retrieve the data. Then they start programming to read that data. Then they must also have understood the data format. Is the temperature value four bytes or is that eight bytes? This is also different depending on the manufacturers.

So I can only guess at that.


That’s right, welcome to the club. Someone who maintains robots, from different manufacturers, they have variables with the same name, but where the meaning in the background is different for different suppliers. They really need to know that so that they don’t draw the wrong conclusions. It is important to understand: OPC UA solves this.

I was just about to make the transition to that, thank you very much.

Solutions, offerings and services – A look at the technologies used [22:30]

That’s exactly what we don’t want, these efforts that you have to handle data again and again, that you talk about something that is not defined and not standardized. Now the question is: How does it work?

How does Miele, for example, or other customers of yours do this in practice today, i.e. standardized connectivity?


First of all, you have to say, when you have brownfield scenarios, it’s exactly the question of how do you get this data integrated from different overlaid protocols and different production components from legacy systems, among others? A common step is to first put adapters before it, gateways so to speak. Either they are protocol turnovers, I rather advise against that, just subordinated has a device mode and I just transform that into OPC UA and then have a “byte sausage” in OPC UA.

This doesn’t help me much; I do have security and can route the data securely through my network as “byte sausage”, but that doesn’t really give me an advantage. The advantage only comes into play when I can write the meaning of the data into this gateway, into this adapter, i.e. the semantics, what the meaning of this “sausage” is.

I must confess, I forgot whether it is 255 words or bytes for Modbus S, but from that alone you can see, this is a completely unimportant know-how, that’s why I forgot. More important is the meaning of the data in there. Where is the actual position? What is the release for a new order? And this complex data description should be loaded into the gateway and then it fetches any proprietary data from below and outputs standardized OPC UA data with security upwards. That’s an IoT box, for example, that you can put upstream.

Let’s stay with the example of Fahrenheit and temperature, that runs together there and is translated that way, can you say that?


Absolutely, we take this once again to the explanation: Below we have Modbus and there simply in four bytes only the process value of the temperature sensor is transferred. The information, whether that’s degrees Fahrenheit or degrees Celsius, and what’s the minimum value and what’s the maximum value where an alarm should be triggered, that’s on a piece of paper or in an Excel spreadsheet or somewhere.

In the gateway, OPC UA provides this additional meta information to the live value of the temperature and can be read out.

In the end, OPC UA is what is then provided with the semantics, that is, with the relationship, with the meaning of these individual data points that are underneath, that you create. The customers then use an adapter, a box, a gateway, that you put in between and that then gives you the possibility to have an OPC UA protocol at the end, which you can then use again; can you put it like that?


OPC UA as a protocol is not the crucial thing, but the ability to modulate in data, to model information, so that everyone doesn’t have to keep digging into the Excel spreadsheet of every piece of software and asking how to do it.

This is the plug-and-produce mechanism, which is a huge advantage and makes everything much more efficient. I no longer need to have all the IP addresses in an Excel spreadsheet. So where is the device and what is it called so I can connect it via connectivity in the first place? But OPC UA offers Discovery, where I can search the network, for devices.

It’s not always just devices, it can just be software running somewhere on the network. So we can get away, OPC UA would always be just the last mile to physical devices.

Plug-and-produce means you provide a mapping or this Discovery. You create a base where you can use information across manufacturers without having to do this mapping.


Perfectly correct so far, but there is more to it and many forget that and therefore sometimes compare protocols with OPC UA, where I say that it is much more. Not just the transfer of data, but that starts much earlier.

Discovery in the network, where are my OPC UA-enabled services? On devices or just the software? How do I connect those with security with IT security, the onboarding? I can “browse” into devices, into a sensor and then see what information it offers me? For example, a PLC has many more variables. I know from my own history. An injection molding machine like this has a million PLC variables.

Let’s say I’m a CO2 manager and I’ve made a few data points and I want to make CO2 measurements for these injection molding machines. Then I would go here and could search with the Discovery which data points are still possible to connect. Is that what you mean?


Exact! It’s like going into the yellow pages of the device and then finding out what all is offered to me; CO2 variants, but I can also do a robot procedure and other things. That is precisely the question: What data is available?

You can’t see all the data, but OPC UA has a lot of advantages over others because it’s also important to know which information is actually visible to which people and for which roles. A service technician will certainly have access to all data, but a SCADA system, visualization system will only have access to maybe 5000 variables. An MES system might still have access to maybe 80 or 100 variables, and a cloud system to even fewer.

Who owns data, who has visibility? And do I have data sovereignty? Which data is visible, which is also writable, so that IT, OT – operational people have a huge fear of the evil IT people breaking their machine, breaking the real-time cycle, writing on data that they’re not supposed to see and write on.

OPC UA also offers this via various protocols, real transport protocols. That’s what makes the flexibility that I can communicate within production.

Exactly, my energy manager does not want to see all data, he also needs certain values. In theory, I can use it to connect data that is then available in the network in a standardized form. That’s what you mean by flexibility, too, this architecture.


What comes to my mind with your example of the energy manager is that the energy manager would be very happy if not every machine, every robot and every RFID reader and every scale and whatever else you have as assets in production, if they all describe the energy data differently, but if they all agreed on the same format.

To say I’m transferring something and it’s no longer available in Excel spreadsheets and I have to manually look it up. Reading up on machines is one step, but if we agree on the same thing across manufacturers, that would be sensational. That’s exactly what’s happening, for example in cooperation with VDMA Europe’s largest machinery association, we are currently specifying an energy management interface; that’s why I just had to jump on it.

This happens so that all the industries that write compendiums specifications, so all the robot manufacturers agree on how to describe a robot and what variables it offers.

All this so that harmonized data can be defined across industries.

How does data recording work?

The next step is now to make the data available, so a Miele or even the others use Microsoft Azure, for example. Is that where you stop?


This is a very important aspect, where, by the way, today there is a battle between the IT and OT worlds, here we are with the topic: Who defines the data and the IT people? The cloud people in particular want to have all the data in their cloud and want to normalize it there so that everything is there. If it is there, then it always costs in access to retrieve the data.

We at the OPC Foundation are more of the opinion that the data should be made available in a standardized form as close as possible to the data source. For example, the flow meter should provide the data standardized with OPC UA if possible in the flow meter, if it can. Then all the levels above can also already access the data in a standardized way and do not have to go back to the cloud. This also means that I can omit layers in communication.

A flow sensor can of course route the data to a controller, but it can also route it to the cloud in parallel without routing it to the MES system because it is not even needed in this use case.

Past that.


Past that! I don’t believe in always believing that there has to be one central data pool in the world and that everything has to go into it. To make a point: I don’t believe that everything will always go through the administration shell and that this will become the central hub of the world, but it is an important and further aspect.

Many don’t know that OPC UA scales all the way to the cloud to transport data. Here, too, a factory operator, like at Miele, wants to have their freedom, and they don’t want to be tied into a certain Microsoft Azure ecosystem, even if it’s absolutely great.

After all, the important thing is to protect the customer, to standardize data once and then send it up in a specific format. And that all these cloud providers understand that directly, like a USB plug, like plug and produce, so that when they push that to another cloud provider, they don’t have to send the data in a different format. Then they would have to adjust their system again.

I’m really proud of that because the OPC Foundation really scales up to the cloud via MQTT. Many people ask me if they should do OPC UA to the cloud or MQTT. The truth is that the OPC Foundation loves MQTT, which is included in our spec as a transport channel. But MQTT itself does not have interoperability.

It’s like Modbus; I can move the data over from left to right, but Modbus doesn’t define the semantics of how it’s in there either. The great thing is, this standard that we adopted in 2018, all the major cloud users have agreed to that; there’s also a 2022 press release about it. Amazon Web Services, IBM, Google, Microsoft Azure, Siemens or SAP that they support OPC UA over MQTT and Amazon Web Service and Azure already do. We have a dashboard and people can just push data in there and it’s automatically in there on two dashboards from different companies.

This closes the circle, because I said at the beginning that our partners in the network keep mentioning you, because the software companies that build their solutions on the basis of Microsoft Azure as a Platform as a Service also use precisely this connectivity.

On the subject of administration shells: There is an episode with Volkswagen on this very subject, feel free to listen to it; I’ll link to it in the show notes. This is a whole separate topic about the digital twin.

It can be stated that the classic automation pyramid is dissolving to a certain extent. Not everything has to go up through the MES, but there are various data points that are only interesting for certain use cases. That’s where the whole thing breaks down accordingly, and the architecture becomes wider than what was originally thought, right?


I think that when people say the automation pyramid is dissolving, they only mean in terms of communication. Of course, we still have a pyramid, there is a control level and an MES level. The question is: where does it take place? Sometimes, as things stand today, the MES part is very close to the production line because it is also real-time communication. Decisions have to be made in a few milliseconds and then it is usually very close to the plant.

Where is a specific functionality? At what level does that take place? This will be much more flexible. If I produce brake pads, then of course I have my PLC on the machine where the brake pads are produced. And if I want to lower the temperature in a building over the weekend, that PLC “chugging along” can run without real time somewhere in a cloud. And if the cloud isn’t there, it’s not so bad. Then you’ve lost a little money over the weekend, but it’s not that bad.

“It’s disintegrating” was perhaps a bit of a provocative way of putting it.


There is still one boss for one company, logically. Talking abot dissolving pyramid, one must always add: what is actually dissolving?

Results, Business Models and Best Practices – How Success is Measured [39:08]

Can you say what the business case is for your end customers, that is, for a Miele?


It is not only the advantage for the users, they also specify it. So Miele contacted its suppliers in advance, said: We are going to use OPC UA, get ready for it and please deliver that. That is an important step; I am enormously pleased about that. Other large corporations are doing the same, including Volkswagen.

It’s not just a push for device providers, but I think they themselves benefit from selling into a large market without always having to describe their own data plug. It simply helps to say: I offer a device, an RFID reader for example, and I have OPC UA as an interface. Then I already know that my counterpart will say: This is an accepted interface, by the way, it has IT security in it, and then it is clear to me that with this OPC connector, various software packages, MES packages or even SAP packages can be connected directly.

Miele has linked certain data, production plan data with actual plant data. SAP has implemented OPC UA since 2008 and is on the board. The connection to SAP is super easy and goes quickly. Attaching new devices with OPC UA and SAP takes less than fifteen minutes.

You have time savings, there are groups of people who work with it that today have manual efforts that we want to avoid, of course. This is the issue of saving time. Flexibility, also with the connection and integration of specific IT systems, but also the possibility to access different data. Do you have anything else to add?


One more, that was also in the Miele use case. Miele has talked about having extreme savings by not having to pay old licensing fees for proprietary software. It’s not just that we can transfer data from A to B securely. This is an ongoing process on the part of the OPC Foundation. Every week, 20 IT security people meet there and reactively look at what happened. Have we been attacked anywhere? The proactively act, what new security things should be integrated into the standard?

That’s something where end users benefit from; from this worldwide ecosystem where experts around the globe have agreed on something, because it can’t be solved by one person; this community behind it, that’s an incredible advantage. This is also a pressure for the suppliers, because they have to do it. And on the other hand, this is the advantage for users, because Miele has three different PLC suppliers, for example. Beckhoff, Siemens, and Mitsubishi, and they know they all supply OPC UA, though, and can do it anywhere.

Miele is working with Microsoft Azure, but if they run into idiotic pricing models, Miele has the freedom to switch to AWS. You don’t have to do anything else; it’s the same protocol.

Today was all about learning from practice, what kind of projects you have, how your customers use that or even how your partners deal with that. Thank you for the exciting insights, also into the projects!

What else is coming in the future? Where is the market going?


We are definitely growing into more industries and information models, for example in the energy sector. We definitely need to do even better marketing. Many do not know that we offer other interfaces with OPC UA, for example. The IT world does not know that an OPC server can also have a Rest API interface.

That’s why so many things in the IT world are constantly being redefined. If I see OPC UA as a Lego building block system and every Lego brick has a functionality, and then some guy finds a Lego brick that he is missing. Then it is often the case that a new ecosystem is made around this one missing Lego brick and all the other Lego bricks are rebuilt again instead of contacting the OPC Foundation.

For the Hannover Messe 2023 I recommend to visit the OPC Foundation; come by! We’re showing live connectivity to Data Spaces, also a really exciting topic. By the way, I personally find this much more exciting than other areas. Data Spaces means: Many different companies share data in a data space, the Data Space, but are able to do this not only technically, but also have an agreement, a contract about the use of the data and about the duration of use of the data.

So I can say I’m supplying this data for a week, my partner can calculate on it, but they are not allowed to copy it. They don’t have them available offline afterwards either, and I continue to be the master of my data. There, too, we show how to connect OPC UA to Data Spaces. We will show a live demo with the Metaverse; there will not be the one Metaverse. Of course, in the future we will have glasses on with either complete virtual reality or augmented reality.

Or simply lenses.


Exactly, and then the question is also: How do we connect the physical existing world, the real existing machine to this additional enriched information that we see in the glasses? We will show this live at the Hannover Messe.

The nice thing is, we’ve proven that for 16 years so far, there is a framework, this Lego puzzle. And that is always expanded with more puzzle pieces without the others becoming incompatible. OPC UA has not become incompatible over the last 16 years and I can still connect to a software package today with the first application in 2006. That’s the value, that stability, securing the investment for the industry.

Come on by, check out the OPC Foundation. We ourselves from the IoT Use Case Ecosystem will also have an IoT meeting there. That’s that Thursday, 4:00pm, feel free to talk to me and share with others. Stefan if you feel like it, cordial invitation of course.

Thank you so much for joining us, and I would love to hear from you again on this.


Many thanks from my side that I was allowed to be there. I’ll be happy to come back and hear from you again!

Please do not hesitate to contact me if you have any questions.

Questions? Contact Madeleine Mickeleit

Ing. Madeleine Mickeleit

Host & General Manager
IoT Use Case Podcast