Epple Druckfarben & Innomat: Digital Batch Traceability with Kalisto IoT

Today, I’m bringing you an exciting project from the world of chemical production. Joining me is Epple Druckfarben – a long-standing family business with roots going back to 1870. For decades, Epple has stood for continuous technological innovation – and is now taking the next important step toward digitalization. I’m very excited to have them here today.


The focus of this episode is the use case of batch traceability – a central topic many of you are probably familiar with. But what exactly is the digitalization potential for Epple and their customers? How is Epple taking on a pioneering role in the area of sustainable products? And what does all this mean for the packaging print industry and the wider market? What best practices can you take away for your own implementation? How did Epple approach the project – and what challenges had to be overcome?
To answer these questions, I’ve invited Tobias Friedrich, who leads digitalization projects at Epple. Also on board: their IoT partner for implementation – Innomat-Automation AG, represented today by Marco Müller, who is the technical manager.
You’ll now hear what exactly was implemented and which best practices emerged from this journey. As always, you’ll find all the implementation details at www.iotusecase.com and in the show notes. Enjoy – and let’s head into the studio!

Hello Marco, hello Tobias! Tobias, let me start with you – how are you today? And where are you joining us from?

Tobias

I’m doing well, thanks! It’s Friday – right before my vacation – so the mood is definitely relaxed.

Fantastic! Just a quick follow-up: where exactly are you located today? According to Google, Epple Druckfarben is based in Neusäß?

Tobias

That’s right – I’m working from home today. Epple Druckfarben is based in Neusäß, which is near Augsburg.

Nice! Greetings to your team if anyone’s tuning in from there – and of course to everyone else listening as well. Marco, how about you – how are you, And where are you joining us from?

Marco

I’m at our Innomat branch office in Frauenfeld. The weekend is just around the corner – just wrapping up a few tasks before we can ease into it.

Fantastic! And for those who don’t know Frauenfeld – can you quickly tell us where you’re located?

Marco

Frauenfeld is near Winterthur, in the Lake Constance region.

Ah, very nice. Great to have you both here – I’m really looking forward to this conversation. To get us started: how did the two of you end up here together on the podcast? I know from our prep talk that you’ve been working together for a while. Can you tell us a bit about how your companies first got in touch – is there a bit of a background story there?

Tobias

I think Marco should tell the background story. When I joined Epple seven years ago, Innomat was already a well-known partner.

Marco

Right. At Innomat – back then still including the company Asprotec, which we later acquired – we installed the first fully automated dosing systems over 15 years ago. At the time, we were already receiving orders directly from the ERP system and producing them automatically on the dosing systems. All batch data was recorded and fed back – across three different machines or production lines. Since then, we’ve been in continuous contact with Epple.

Got it – so Epple is still using your dosing systems today. Very interesting! Let’s stay with Epple for a moment. Tobias, what’s your current role? You’re a leading provider of printing inks – can you give us an idea of your daily work and what operations look like at your plant?

Tobias

Let me start with the harder part of the question – my daily routine. You introduced me earlier as a project lead – currently, I’m actually Head of Logistics. That’s part of the journey I’ve been on over the past seven years at Epple. I originally joined the company as an industrial clerk and completed my apprenticeship there. From there, things evolved: I worked as a production planner for a while, completed a part-time degree, and eventually moved into the role of project lead for digitalization.
The project we’re discussing today was actually a central part of that role.
As for my daily work – no two days are the same. Fortunately, I’m always dealing with exciting challenges coming directly from the shop floor – mostly from logistics and production. It’s all about solving problems, improving processes, and ensuring that we manufacture and deliver our printing inks as efficiently as possible.

Okay, so looking at your product portfolio – from coatings to different systems and packaging inks – do you produce everything on site at your facility?

Tobias

Right. On the one hand, we are specialized – we only produce sheet-fed offset printing inks. This means that our products are designed for use on sheet-fed offset industrial printing presses. Within this segment, we see ourselves as a full-range provider. We offer a complete portfolio of inks – from standard to custom solutions – as well as all the necessary printing additives, like the coatings you just mentioned, and other auxiliary materials used in this field.

Interesting! So your typical customers are probably the ones buying and processing those inks and coatings – is that right? Who are your main customer groups?

Tobias

Basically, we have two customer groups: Either it’s the printing companies themselves, who use our inks directly, or distributors who resell our products – also to printing companies.

[05:43] Challenges, potentials and status quo – This is what the use case looks like in practice

What’s currently happening in your market that’s prompting a company like Epple to invest in this area and push ahead with such projects? Is it something driven by your customers, or is it more of an internal initiative? Tell us – what’s happening at your company, and what exactly is your digitalization project?

Tobias

That’s actually quite interesting when you talk about digitalization – because we produce a product that fundamentally stands in contrast to that: it’s ultimately used in an analog process. You might think less is being printed nowadays – for example, you see fewer flyers or brochures in your mailbox. But printing is actually increasing, especially in the packaging sector. As you mentioned earlier, that’s an important area for us – one where we’re already strong and want to grow even further.
And that’s exactly where the topic of batch traceability becomes especially relevant. Almost everything is packaged these days, but the requirements for packaging are constantly increasing – including due to legal regulations. There are increasingly strict guidelines that require complete batch traceability, particularly when packaging comes into direct or indirect contact with food.
That’s one driver. On the other hand, batch traceability is also important for us internally. We can learn a lot about the quality of our products, for example, when we notice fluctuations in certain quality parameters – such as differences in raw materials. If we know that these fluctuations remained consistent within a batch, we can draw conclusions about the end product – for example regarding characteristics like viscosity.

That’s really interesting! I’d like to dig deeper into that in a moment. So, does batch traceability in your case mean complete traceability of products and materials? And does it mainly concern packaging – is that right?

Tobias

We ensure complete batch traceability across all our production processes. But it’s especially requested by customers in the packaging printing sector.

So, for example, a printing company comes to you and says, “We used to do this with paper notes or Excel – today we want a digital data set from you.” Something like that?

Tobias

Right. When we sell ink to a customer, there’s a batch number on the can. With that number, we can trace exactly which raw materials were used – each with its own batch number. And the intermediate products used also have their own batch numbers, which are traceable as well. So even if the end product is made up of several intermediate products, which themselves consist of other intermediates, it’s still possible to trace everything back to the original raw material batch.

Got it. And all of this can now be implemented using IoT technologies – which is what we’re talking about today. Earlier you mentioned something interesting: the quality varies depending on the raw materials used. Could you explain that in more detail? How do you detect such variations using data?

Tobias

Yes, of course. When we receive raw materials, and as soon as they’re entered into our system, we perform an incoming goods inspection. As part of that, we measure specific quality characteristics – for example, viscosity, meaning how fluid the component is. With natural products, there are inherent fluctuations in quality.
This incoming inspection gives us the ability to see: What properties does the raw material bring? If necessary, we then have to adjust our formulation to ensure that the quality standards for our customers are still met – for example, in terms of the target viscosity.

Got it, thanks. And before you started working with Innomat – how did you handle this in the past, before digitization?

Tobias

I’d say it was twofold. Some things simply weren’t handled at all in the past – either because the market or our customers didn’t require it, or because it just wasn’t necessary. In those cases, digitalization brings a clear improvement.
Other things were handled, but with a lot of manual effort. As you suspected earlier: in some cases we worked with handwritten notes. For example, the batch number was written by hand on the production order. And if there was a quality issue and you needed to trace it back, you first had to dig through the paperwork.

You are active in chemical production – were there any special requirements from the industry that your customers placed on you? I’m thinking of GMP or other standards, for example. Were there customers who said: “You have to meet these requirements if you want to work with us in the long term” – to put it bluntly?

Tobias

Yes, we do have customers with those kinds of requirements – mainly in the packaging printing sector. Some specifically request complete batch traceability, even if they don’t actually need the data in detail. In a way, they’re testing us – for instance, by expecting us to provide all relevant values and batch numbers within 24 hours.

[10:53] Solutions, offerings and services – A look at the technologies used

And now you’ve started this project together. What exactly did you implement? This question goes to you, Marco: How did you support Epple in this project, and what exactly did you work on together?

Marco

This project covered several aspects. First, it involved the legacy systems that had already been in use for over 15 years. These systems were upgraded to the latest state of the art.
As part of the project, another major step was added: Epple’s transition to a new ERP system. In this context, of course, the interfaces also had to be reconnected.
It was decided to integrate both manual and semi-automatic weighing into the new system. That means: the workstations where materials are weighed – sometimes automatically, sometimes manually – were connected to our Kalisto system.
We read the weighing data directly from the scales. Employees can pull up the order at their station and weigh the components – some automatically via our control system, others manually – and all weighing data is logged directly in Kalisto. They also enter the batch numbers of the manually weighed components so that these can be sent directly back to the ERP system after weighing. This ensures that batch traceability in the ERP is available within seconds.

Very interesting! So at Epple, you actually implemented several use cases at once. If I understand correctly, this involves quality fluctuations in raw materials, viscosity, material origin – especially when dealing with natural products. And this specific project focused on batch data at the machine level. So it’s one of several ongoing efforts, right?

Tobias

Yes, it’s not that easy to separate. The starting point for the joint project with Marco was actually our ERP system switch at the beginning of the year. We didn’t want to spend unnecessary time and money integrating outdated systems that no longer met our current requirements.

Okay, so it developed over time – not a single project with a defined beginning and end, but more of a broader initiative with multiple components and outcomes?

Tobias

Right.

Got it. And what was the point at which you said, “This project will be successful if X and Y are achieved”? Was it specifically about solving batch traceability? Or what were your success criteria – something you’ve already reached or are still working towards?

Tobias

Our ERP system sends the production order to the Kalisto system. This includes all relevant workstations, instructions, and the components with their respective target quantities to be weighed.
A typical sequence would be: weigh the first component, then five additional components, followed by mixing, and then more components. That’s the data flow we transfer to Kalisto. Maybe Marco can jump in here and explain what happens on his side.

Marco

Gladly. We receive the orders from the ERP system and first store them temporarily. The benefit of that is that production can continue even if the connection to the ERP is interrupted, since we can access the locally cached data.
The employee then selects the order they want to process. It is displayed on the screen – including the current scale values. The system shows whether a component can be dosed automatically or manually.
For an automatic component, one click is enough to start the dosing. For a manual component, the employee takes the correct raw material and weighs it.
Especially in the context of batch traceability, there’s an important feature: if one container isn’t enough, the employee can perform what we call a batch split – meaning they inform the system that part of the component was weighed using one batch, and the rest using another. This allows cross-batch operations to be documented cleanly.
Once the entire order has been weighed, it is confirmed, stored locally, and simultaneously reported back to the ERP system.
A key success factor here was user friendliness. The system needed to be designed so that employees could be trained quickly and operate it intuitively – ideally completing an entire order with just a few clicks.

Madeleine

Very interesting! You’ve mentioned the product Kalisto IoT several times – maybe let’s talk about that a bit more. Marco, I’ve known your company Innomat for a while now, and Kalisto IoT is – let’s say – a modular software platform that you specifically developed to implement custom IoT solutions.
There are a few unique features. Could you describe what exactly was implemented in this project? Was Kalisto IoT used in its standard form, or were additional components involved? What did you contribute – and what came from Epple?

Marco

With Kalisto, we’ve developed a version tailored to the paint and coatings industry – specifically for managing dosing-related orders. While Kalisto isn’t limited to this industry, it’s particularly well-suited for mapping out the order-handling process here. Typically, we can also work with complete formulations, but that wasn’t necessary in Epple’s case.
What was “special” at Epple: we implemented a custom ERP interface. That’s actually standard in nearly all our projects – no two ERP systems are alike. So we always develop a tailored connection for the data exchange.
With our standard product, such as the weighing software, there were a few minor customizations for Epple. But that’s all within the usual scope of what we do to meet customer requirements as precisely as possible.
Another key aspect at Epple was the connection to existing hardware. We had to develop a specific interface to connect with their existing weighing systems and retrieve the relevant scale values for the workstation terminals.

Did you have to directly access the control system of the dosing units for that? Or what exactly was the challenge with the existing hardware?

Marco

There was no problem with the dosing units – the core software comes from us, and the required interfaces are already part of our standard offering.
The challenge was with the manual and semi-automated workstations. These used systems from two different manufacturers – one provided the weighing equipment, the other parts of the dosing technology. Here, we had to develop specific interfaces and implement them individually to ensure full integration into our system.

I see. And how many workstations are we talking about in total? Did you retrofit any hardware at those stations – or how exactly did you implement this?

Marco

In total, we’re talking about six manual workstations. Five of them are semi-automated – meaning certain components can be dosed automatically. One station is fully manual.
In addition, there are two fully automated systems: one smaller unit and a larger one with four dosing stations that handle the dosing processes completely automatically.

Okay, great. Tobias, if we take a look at this from a practical perspective: your employees work directly at these stations and machines. What exactly has changed or improved for them? You had described several challenges earlier – how has the new system impacted your team?

Tobias

One of the biggest changes is that everything is now much more standardized. Previously, we had several systems in use – and depending on the workstation, the interface looked different. That meant employees who were very familiar with one setup couldn’t necessarily work at another because they didn’t know the respective system.
Now – as Marco already mentioned – the user interface is designed to be as simple as possible. We also had a few small customizations implemented just for us. One example: the system won’t allow you to continue unless a batch number has been entered. Without it, no further weighing steps can be completed.
Another major advantage is the flexibility we now have. Marco has already mentioned it: We have very different working stations. Some customers order large quantities – several tons of one ink – while others might only need a single kilo. That means the requirements for the weighing system vary widely.
Thanks to the new system, we can now respond flexibly. We use a unified weighing solution that can be accessed from any workstation. That allows us, for example, to start weighing at one station and continue at another. Depending on how the production is set up, I can dose five components automatically at one scale and the remaining ten at another.
In the end, everything runs through one centralized system with a standardized protocol that feeds directly back into the ERP.
That’s a huge improvement for our employees – a unified, easy-to-use system that lets them work directly on the weighing order.

That’s really interesting – and a major advantage from a business perspective as well, especially if the previous process required a lot of workarounds.
What you just said about the batch number also shows how much time you can save. Did you actually calculate a return on investment? After all, you’re investing in new technology. Or would you say: we know where our pain points were – and the benefits are obvious, even without detailed calculations?

Tobias

Exactly, we didn’t create a detailed ROI calculation. But like you said: there are plenty of points where we can logically say the investment was worth it. Reducing errors alone, or just avoiding a single failed production batch, might already justify the effort.
On top of that, this investment was necessary due to our ERP migration. Without an interface, we couldn’t have continued operating with the existing equipment. We deliberately chose to phase out isolated solutions and move toward a centralized, integrated system. And in the end, this meant that we had to program fewer individual interfaces.

Yes, that makes total sense. Marco, would you say this was a typical project for you? Kalisto IoT comes in different variants, and your clients vary a lot – not all of them are from the ink industry. Would you consider this a standard case or something more customized?

Marco

That’s a good question – what is a “typical” project, really? In terms of process, this project was pretty similar to many others we’ve done. But each project comes with its own particularities.
Here, for example, the unique challenge was the interface with systems from other manufacturers – like we talked about earlier. That’s not entirely unusual, but it’s always specific to each project.
You need to coordinate with the external vendors and figure out how to establish the connection – and that always brings a certain level of uncertainty. But with our approach, we usually find a solution. I think Tobias can confirm – we managed all the coordination with the third-party providers of the semi-automated dosing systems behind the scenes. Epple didn’t have to deal with those technical details at all.

I’m asking because I know you handle a wide range of projects – some are classic on-premise setups, others are cloud-based. From what I understand, you work with both SAP and Microsoft Azure – and your solutions are also compatible with Azure. Maybe someone listening is planning a similar project – even if the specific requirements are different. So I just wanted to highlight: you don’t just handle projects like this one, but offer a broad range of services, including Azure-based setups. In this case, though, I believe the project was run using a local ERP system – is that right?

Marco

That’s correct – Kalisto is running entirely on-premise in this case. That was a specific request from the customer: if something goes wrong, everything should still function locally.
We’re using the customer’s existing infrastructure – we didn’t provide any additional IT hardware. The system runs virtually, which makes it easy to manage backups and quick restores.
As for the ERP: Epple is using Microsoft Dynamics 365 Business Central in the cloud. The integration is done via a REST API, allowing us to exchange data between Kalisto and the cloud-based ERP system.

We’ve talked a lot about batch traceability. But I know that Innomat has a much broader scope. You’re also involved in interface management for autonomous transport systems, and I believe even in hospital and MedTech environments. Can you share a bit more about that? What other kinds of projects do you handle, and how broad is your portfolio? You’re not just focused on batch traceability, right?

Marco

That’s right – batch traceability is something we’ve been doing for more than 20 years, so we have a lot of experience in that area. But our platform can do a lot more.
Like you mentioned, we also work on interface projects involving autonomous guided vehicles (AGVs), where we respond to commands from machines or external software and send commands the AGVs accordingly.
One example: our platform acts as a gateway. So if an AGV needs to access an elevator, we can use a REST API to trigger the elevator, and then use OPC UA to inform the AGV: “The door is open – you can enter.” That way, the entire communication process is automated.
Another area is data collection and monitoring. We gather and analyze metrics like electricity and compressed air consumption. It starts with simple KPIs but can scale up to full track-and-trace systems, where entire datasets are stored and analyzed – including for batch traceability purposes.

Very exciting – sounds like a great topic for another episode! There’s definitely a lot of potential there. So let’s wrap up with one final question for today – this one’s for you, Tobias: What’s next at your company? How is the project evolving? Are there any upcoming initiatives in IoT or digitalization that you’re working on or planning? What can we look forward to?

Tobias

For us, status tracking in production is currently a particularly exciting topic. Right now, the process looks like this: A production order is created, transferred to the system – and from that point on, it’s basically a black box. You can only find out the current status through manual follow-up.
That’s why we started a project where we use scanners in production. These send information directly to our cloud-based ERP system. So, for example, when someone scans a production sheet and completes a step, it becomes visible in the system in real time. We’re currently expanding this – it’s our most important digitization project in production at the moment.
I also see Marco’s system as a potential data source. As mentioned earlier, we send work instructions to the weighing system. Initially, that was just to improve clarity – so the employees would have the same information on paper and in the system.
But I see additional potential here: if a step is completed in Kalisto, we could use that information as well – almost “for free” – and feed it back into the ERP system.

That’s great! Marco, let’s talk about the future from your perspective. I have to bring up the obvious: everyone’s talking about AI right now. Is that a topic for you as well – specifically AI based on IoT data? What are you currently working on?

Marco

Yes, AI is definitely a hot topic – and of course we’re actively exploring it too. But the key question is: how can AI be used in a meaningful way?
Especially when it comes to things like batch data, it’s all about hard facts. You can’t have an AI randomly generating numbers – it simply wouldn’t be reliable. And that’s not the point of AI anyway. The goal is to quickly gain insights and uncover correlations.
What we’re focusing on right now is monitoring. We’re also looking into low-code integrations to make the data from our Kalisto system even more accessible. One example of this is the topic of event triggering.
A feature we already have is alarm management: we can automatically send notifications based on incoming data. This allows for comprehensive monitoring with alerts.
There are many more ideas and use cases we’re working on to broaden the scope of Kalisto and offer even more value to our customers.

Fantastic! And by the way – speaking of AI and IoT data – we recently released a new podcast episode on this topic. It’s about a pilot project with OptoTech, part of the SCHUNK Group. That’s episode 163 – if you’re curious, check it out.
As always, if you’re listening and think, “Hey, I have a similar project,” or you simply want to connect – feel free to reach out. Tobias, Marco – I’ll include your LinkedIn profiles and all relevant links in the show notes.
Anything you’d like to share before we wrap up – maybe an event you’ll be attending or a quick message for the audience?

Marco

Just reach out! If you have an idea about what could be automated or integrated with an IoT solution, we’re always happy to discuss it.

Thanks again to both of you for joining. This was a really insightful conversation – a great real-world project that I’m sure many listeners will benefit from. And Kalisto is clearly a powerful system – both for today’s use cases and future applications.
Thanks again for the great insights – I’ll let you have the final word.

Tobias

Thanks so much for the invitation – I really enjoyed it!

Marco

Same here – thank you! I’m looking forward to more exciting projects – especially with you, Tobias. Let’s see what the next few years bring.

Take care, and have a great rest of the week. Bye!

Marco

Thanks a lot – bye Madeleine!

Tobias

Thank you, bye!