Episode 10: The Next Lane: Marel and the Digital Product Highway with Roger Kabo

Come join Share PLM for another podcast episode with Roger Kabo, the Domain Architect for Marel, working on the border of business and IT in international environments. Roger has extensive PLM knowledge and experience and has experience in IT management, project management, analytical, and consulting skills.  

In this episode, we are going to talk about: 

⚉ Marel (Company Overview) 
⚉ Roger's PLM Roles and Responsibilities 
⚉ Engineering to Order (ETO) vs. Configure to Order (CTO) 
⚉ Integration of Services with PLM in Marel 
⚉ PLM Vision and Strategy 
⚉ PLM System and Future Roadmap 
⚉ Pragmatic Approach to Model-Based PLM 
⚉ Collaboration and Knowledge Sharing in the PLM Community 


MENTIONED IN THIS EPISODE:

• Marel -  https://marel.com/ 

CONNECT WITH ROGER:

• Linkedin: https://www.linkedin.com/in/roger-kabo-5ba6576/ 

CONNECT WITH SHARE PLM:

• Website: https://shareplm.com/ 

Join us every month to listen to fascinating interviews, where we cover a wide array of topics, from actionable tips, to personal experiences, to strategies that you can implement into your PLM strategy. 

If you have an interesting story to share and want to join the conversation, contact us and let's chat. We can't wait to hear from you!

Transcript

[00:00:12] HELENA:
Hello. Welcome everybody to Share PLM podcast. My name is Helena Gutierrez and today we also have with us Jos Voskuil, our co host for the podcast and Roger Kabo. Welcome Roger. Nice to meet you.

[00:00:26] ROGER:
Yeah. Thank you. And thank you for having me here.

[00:00:29] HELENA:
Thank you, Roger. Could you please introduce yourself?

[00:00:33] ROGER:
Sure. My name is Roger Kabo. I'm working with Marel and I'm in the position of a Domain Architect for the product area and essentially that is circling around PLM and the PLM use in the business. And as such, I'm responsible for defining the PLM landscape and the PLM roadmap. And giving that to the, let's say to the people that then have to realize that roadmap.

[00:00:59] JOS: 
Roger we know each other because we are both Dutch and for not everyone in the world who knows Marel, but he tells you also something about Marel as a company.

[00:01:09] ROGER:
We are machine builder for the food industry and we are active in different industries. So we are active in poultry and fish and meat and also in the retail and food ready market and also in plant based proteins. We are a global company, running about 9, 000 people active in multiple countries, and we are building machinery for all those different industries and with our customers we are creating complete food plants so to process all of these products I just mentioned.

[00:01:41] HELENA:
Truly global company. How long have you been with them?

[00:01:45] ROGER:
I've started working with Marel in 2002. So I'm, I'm with the company for quite a long time already. And I started out in the PLM department which at that time was a three people department. The company was much smaller at that point. And from there I had several different roles and always let's say between business and IT. So mostly reporting into the IT area, sometimes also reporting into the business area, but always making the match between, let's say the business needs and the realization on the IT side.

[00:02:23] HELENA:
Right. And what is your background, Roger? Are you, have you always been in PLM, or did you start as, as you…

[00:02:30] ROGER:
I've been at PLM for a long time, but not always. I was trained as a mechanical engineer and, but actually never worked as a mechanical engineer. My first serious job was with PTC. Also a PLM provider, of course, but in the days that I was with PTC, they were for mostly cat provider. So I've been working there for around and about 10 years. And then from PTC, I moved into, uh, into Marel.

[00:02:56] JOS:
Roger, we know each other for many years and we have been having discussions in the Netherlands and also in PLM. Of course, when we look at Marel, it's a typical machine manufacturer and they have this challenge that would you say it's a typical ETO type of company, engineering to order? 

[00:03:14] ROGER:
I would say we are not typical ETO. In our product development we strive to develop machines for the market, not for an individual customer. So you can look at our business processes as we have a primary value stream going from the market to the customer to the installation and the aftercare.

In the install base, and then we have a product development, which is delivering standard machines into that primary, uh, value stream and those machines are configurable. So we generate configurable machine and we try as much as possible to sell and deliver configurable machines. Obviously when you design a complete production plant, food production plant, you will have to make compromises, always. There is internal transport between the different machines. You have to imagine that a food processing plant consists of maybe 40 or more machines. And the price is product is moving in between the machines.

So the internal transport is mostly more ETO than it is CTO. And also our newer developments says so, O series, and very new developments, they tend to be more ETO, but we always transform them into a, into a CTO type of machine. There is also a bit of a difference between the different divisions that we have.

Poultry is very much configured to order, uh, other divisions maybe are a bit more in the mix of ETO and CTO.

[00:04:41] JOS:
Okay, because the question I'm asked is, is that many of those special machine manufacturers, especially if it's a low volume, they always suffer later from services. Because the purpose is to deliver the machine as fast as possible. And then services is the backend. How is the connection with services arranged in Marel?

[00:05:01] ROGER:
Yeah, pretty tight, I would say. In our PLM environment what we do is we work with product coding structures that enables us to, to deliver a spare part packages straight out of the product structures that are being configured for the customer. And also since many years we do our install base management inside of the PLM ecosystem.

So that means that every customer specific machine that we deliver ends up in our install base environment open to the nut and bolt level, which makes it quite transparent what components, mechanical, electrical, or electronical are with which customer.

So the maintenance at the customer side becomes done a lot more transparent. Also you have to think that much of our machinery has a pretty long lifetime at the customer, at the plant. So tens of years, 20 years, 30 years, maybe even. And in that time span, the development of our products is not having a standstill.

So we have new versions of machine with, with additional capabilities with better yields or, or better processing speeds, and we deliver the opportunity to our customers to upgrade from a, let's say version one machine to a version two machine or three machine. And also there are installed base management plays an important role because then we can specifically see from which version to what version customers can upgrade and we can predefine upgrade packages for our customers.

[00:06:34] HELENA:
Interesting. One basic question, Roger, do you mind managing PLM individual products, or are you managing really the whole plant? As well customer specific instance of the plant in the PLM system?

[00:06:50] ROGER:
We are managing both, but in different ways and different levels of granularity, I would say. So obviously each machine that is being developed for the market that you overcomplete machine. And as well as the configured customer specific machines out of that, they are managed up to the lowest level. And then when we have customer orders, uh, each of these machines are labeled with that specific customer order. And what we also manage, uh, is the plant layout says, so the relation between the different machines the internal transport, et cetera, also that is managed.

[00:07:25] HELENA:
Got it. And how about the install base? I think you mentioned the install base. Is it also part of your PLM ecosystem?

[00:07:33] ROGER:
It is. Yeah. I think that's all an old development already. I think that's like 10 or 15 year old. I don't remember exactly in what year we brought that life, but that is for a considerable time integral part of the PLM ecosystem. And obviously it's, it's not like, like an engineering bill of material. It's an as maintained bill of material that we done, uh keep the different versions.

[00:07:58] HELENA:
And who is the customer who uses that install base? Is it more the engineering or the product development, or do you really have your service guys also looking into?

[00:08:07] ROGER:
It is really service. So engineering uses that to a very limited extent. Uh, so in defining these upgrade packages, engineering is involved and they need to be involved in install base, really, but the primary customer, the primary user of that is the service department.

So we have people in the back office using data. So when customers call for spare parts, for example, but also our field service engineers, which are on the road for the majority of their time, they have access to the installed base to see exactly what is for that specific customer, the situation.

[00:08:46] JOS:
I think we have now a good understanding of, of the company and the process is, uh Roger, and if we go back to, I think, the time we worked together, it was 2018 on the 2022 vision. And, uh, I have some questions on, on, on the vision. Where does this concept of vision, uh, start in, in Marel? What was the reason to start building a vision?

[00:09:09] ROGER:
Yeah. Good question. So maybe I need to take a small step further back in time. So Marel is, is an old time PLM user. I think we started to use, let's say in its rudimentary shape. We started using PLM in the nineties somewhere at that time, it was really, uh, engineering based, but in an organic way, it grew quite quickly outside of the engineering environment into manufacturing supply chain service.

And we did not really have a well documented sound and shared vision at that time on what PLM was supposed to be for Marel. With the progression of time and as it grew bigger and bigger also our PLM system grew older and older. So we are using a system that was then conceived in the, in the nineties.

And we are still using that today. So that's one development and that system is getting out of service in a few years time. And that is one development. The other development is that Marel is a company that grows to a natural growth in in one way but by acquisitions in another way.

So over time, we acquired quite a few companies, which all carried their own PLM or PDM systems. And we are now in a landscape where we have one big PLM system, but several smaller PLM systems alongside, and that makes it difficult to freely share product information across the different divisions and locations.

So that combined with the end of life of our biggest PLM system  triggered us back in 2018 to rethink our PLM strategy, because one thing what you could do is to replace one for one the out of service PLM system with another system But that would be very short sighted because it's not easy to let's say refocus and really think about your strategy if you cannot change your system.

So it's a seldom opportunity that you have in such a situation. So we took the decision not to do this one for one replacement, but to start from scratch, think about what PLM should do for the company. And that starts with the company goals and company vision in the first place. So at the time, and you were part of that we took that from scratch and we've built a vision a bit naively calling that PLM vision 2022 within the back of our minds. 

So I are by 2022, we will be ready  with all of this four years time, but that was incredibly naive. So we delivered I think in the beginning of 2019 that vision, we presented it to our senior management, they went along with that and that was a good game. 

So we defined what PLM should be. We also defined the I say the overall architecture of a future PLM landscape. We did not at that point define the strategy how to realize the vision. And the work that went in, let's say starting 2019.

And then that took a few years before we really had that completely plotted out, uh, how we would do that. And yeah, now we're at the point that we sharpened that vision that we had done. We have a strategy put to it and we are now realizing let's say this architecture that we defined back in, uh, in 2019, that still stands.

[00:12:37] JOS:
Yeah, I think it's interesting to talk with you Roger, because I consider Marel on the PLM side quite advanced, you did real PLM already, not only engineering tool in the early part of this century, and also you are doing because of the merchant acquisitions already master data management in the company as a project.

I mean, very few companies do master data management to align the data and to get more ready for a data-driven approach. I mean, that was quite futuristic. And I think also with the plan now, I think it's a roadmap that you see also other companies are slowly discovering. Can you tell us more about your future roadmap, how it looks like?

[00:13:22] ROGER:
If you take a look at our vision and if you take a look at the overall architecture, what we want to do is we want to break down the borders between all the divisions and, and all the functions that we have in the company. And the PLM ecosystem that we are building should be a sustainable ecosystem, sustainable in the sense that it should be capable of living through all the company changes that we go through.

And, you know we see the company changing all the time. It also needs to be authoritative, meaning that and I learned that from one of the previous podcasts. I think it's not about the source of truth. It's about the source of change. So if you want to know where to get the information it should be the PLM site in the end.

And it should be a scalable system that people are happy to work in. And then, you know, we defined a number of areas like product Portfolio like service like operational efficiency that it all should contain. Translating that with the architecture into a roadmap we defined a two-layer architecture.

So we are not going with one big PLM system. We are actually implementing three different applications in our ecosystem. And that is because uh, we want our landscape to be composable. So, instead of one moloch doing it all we look at compartmenting the different parts in our architecture.

So I think I previously mentioned our primary value stream, uh, from the market to the customer. That has a quite different characteristic than the new product development process and the idea-to-market process. 

So we took the approach that the primary value stream needs to be supported by a different PLM system than the idea to market process. So that primary value stream that is managing a released data. So whatevder layout engineers, mechanical engineers, develop and release and sub in that upper, uh, layer. And we call that, uh, the digital product highway.

[00:15:29] ROGER:
And we call it like that because it's, uh, it's intended to connect all the different locations, divisions, global functions, etc. including maybe customers even.

[00:15:41] JOS:
Okay. So it's really a digital platform, connect all the stakeholders as much as possible real time.

[00:15:47] ROGER:
And that's, uh, we started to implement first and you just mentioned most data management. And that was actually the first approach to implementing that. So the most data management part in that here is that we started to implement to manage our overall, let's say machine structure or let's call it a commercial product structure not with both in another level, but on a commercial level, these are the standard machines that we have on offering.

And these are the modules that are part of these standard machines. And then to that we connect several kinds of sales information that flow through different systems. So we used to be in a situation where, for example standard sales text or sales prices they were managed in the ERP systems.

We have many different ERP systems. When people needed to change, for example one standard sales text, they needed to change that  in tens of different locations. And that was all manual work. So that was our first approach to implementing that digital product highway. And that was for a good reason because we could show immediate success.

Uh, we could also take the approach away from the let's say traditional engineering side of things. And following that we implemented a spare part management on the additional product highway layer, meaning that we consolidate the spare parts that are created in different PLM systems into the digital product highway, making it much more easier.

For service people, for manufacturing people to communicate with the customer and to deliver spare parts to the customer. In the end, going from there on the dish product highway layer there will be other steps that we take to finalize that one on the bottom layer at the PDM layer we just started implementing that part. So that is a very early stages that we are in.

[00:17:48] HELENA:
I had a question related that, uh, I love the name Digital Product Highway. I think it's a great name. When you mentioned the construction of the vision, you mentioned a three year timeframe and you said that it was a bit naive. Now with the experience that you have, what do you think it's a good timeframe? When companies are thinking about their visions and roadmaps, what should they think of?

[00:18:09] ROGER:
The time frame depends a lot on the situation that you're in with the program. In the beginning year, so let's say 2019, uh, 2020, uh, but also a bit later in time. We did a big effort, to sell the program. So to people deeply involved in PLM, it's pretty obvious if your PLM system is going end of life that is basically a business continuity risk how simple it seems for PLM people to, to realize that, uh, it is equally difficult to convey that message to, uh, to senior management sometimes. 

[00:18:46] JOS:
To hook in on this, uh, the good point was you had a, what we call a burning platform and you had a need to change. Everyone was aware, but I think also what was, uh, important, the education part to get the people along what is possible to do the future.

[00:19:05] ROGER:
Yes, so in detail, so just saying, well, you know, the old system is going to die. We need to implement a new one. That's not going to do the trick. It also needs to have benefits and in that division helped quite a lot. So when you talk to different people, some people care about the end of life of dual system, but many people also care about the possibilities that they want to realize that the current systems don't offer us.

So think of IOT related applications or think about other areas where we can't do things today. Simple things that that current systems have  like a split between different types of bombs, engineering bomb, production bomb, et cetera. We could do not do with our current system.

Our manufacturing and supply chain areas are screaming for these kinds of possibilities. And so that helps in selling a division and selling the program. Going back to the timeline. So that is one thing that you have to overcome, so you have to sell it. And then there is also other, other types of factors.

So in our situation, obviously the end of life deadline is, is given us a timeline. So we don't have too much freedom in defining that timeline. And we are currently looking at a very aggressive timeline of realizing the major part of realizing our ecosystem in, in the next upcoming three years, which is quite short.

So if you take a look in the PLM arena a usual throughput time would be at minimum five years. So we really need to make an effort to shorten timelines as, as much as possible while implementing new processes as well because that's the other thing we don't want to replicate the old system. That means we also don't want to replicate the old processes.

[00:20:52] HELENA:
What is your experience, Joss? I want to hear from you. What is a good time frame?

[00:20:57] JOS:
I don't believe in timeframes. I believe in the journey. And the journey never stops and you address every time the most important business streams you want to address. Of course, here we have the, also the concept of, uh, we have to migrate data because the system becomes obsolete, which some other companies might not have.

Because if you remember also the story and you were referred to it, uh, Roger from Yusuf, uh, in the podcast, he also mentioned that everything should start from the business value that you want to achieve and then in an agile manner, uh, short the verbals, uh, build solutions and grow with that. So no proof of concepts, but minimum viable products that you enhance.

But for that, you need to have a vision, at least when everything needs to fit in that vision. And that's why I started also with the discussion of the vision. Roger, I think in Marel, the vision is quite stable and that's the power where you can work towards.

[00:21:58] ROGER:
Yeah. That, that's the beauty in data. So what we defined in 2019, it still stands. And that is a good thing. So maybe on one level down there, the details might differ. A new terminology might be there but in principle, the division still stands.

And that helps you because having a vision helps you in in you in defining the road map, and it helps you in taking a lot of day to day decisions as well, because you have this end point in mind, and you start with the end in mind.

[00:22:28] HELENA:
What has been the most challenging thing since you started? 

[00:22:32] ROGER:
What has been the most challenging thing? I think explaining PLM to the outside of the PLM specialists, that, that is quite a thing. We had numerous sessions with management on, uh, on all the different aspects and from a management perspective let's say the end goal is clear but the difficulties of the need, they are not always well understood that.

So I have one, one colleague once who described it, people want the cherry on the cake, but they don't want to build the cake that lies underneath that cherry. Right. So…

[00:23:05] HELENA:
I love that, that's a good one.

[00:23:09] ROGER:
So if you want to do all kinds of super advanced, super nice stuff that you read in the suppliers and in the software deliver companies’ language you need to have the basis for us and building them that basis that is where the effort goes. And that's, that's the cake that you need to put the cherry upon.

[00:23:30] JOS:
And do you have special programs for that, for, uh, educating people, aligning people, uh, spreading the news?

[00:23:38] ROGER:
Not as much as I would like to. So we are building quite a bit in the organizational change management roles. So that is an area that we are professionalizing in, I would say the last few years we could do more, but then I'm not the most patient person around, I think, so I also realize that building up such a capability and the organizational change management capability takes time to do.

So we are doing a lot on it, but not in the, in the most efficient way yet, I believe. The good thing is, is that in the program, the key players in the program all realize very well that that is very hard needed to have to do good change management with organizational change management comes training and education, of course.

[00:24:26] HELENA:
Why do you think that the stakeholders understand the importance of change management? Is it because of previous experiences or…

[00:24:34] ROGER:
Yes, I think so. I think, you know, some of the choices that we made in our applications that we chose and in the vision that we chose they mean that we are making big changes in the company. Not only on a, on a system level but on a process level.

[00:24:52] ROGER:
And that touches the roles of people. What do people do on a day to day level and how are they organized on a day to day level? And luckily people understand that that is a hard job to do. Maybe good to mention also before this program I think what would it be 10 years ago?

We started another program to implement the old PLM system in different new locations. And in that program we learned a lot about change management as well. So I think many people carry that experience, uh, with them to, uh, to this program also.

[00:25:26] HELENA:
What would you recommend someone who is starting a journey similar to yours who needs to replace an old platform and needs to sell it to management? What would be your top three recommendations?

[00:25:39] ROGER:
Yeah, there would be a few, I think. So I think the first one and maybe the most important one is to start with the end in mind has to start with your vision and then think about how you can realize that vision. I think that is one important element. I think also an important element is that you have to, to sell your vision and you have to sell your program and in, in doing that, and I think that's It's maybe an open door to mention.

Don't use technical language, don't use PLM language, but, but talk about company benefits and, and ensuring business continuity and, and added value as, so those are the, the terms that you need to talk. And then do it step by step and that's that's your thing. I don't try to sell a complete program that will cost multimilions which you talk about a journey you can't effectively predict the cost of a journey in the first place. Uh, you can also not exactly predict the benefits of the, of that will be the outcome of the journey.

So do it step by step, break it down in pieces and start with an element that you know, that will bring a lot of value, that will bring enthusiasm and that will trigger the next step to take again. I think those would be my, my top three.

[00:26:55] HELENA:
Those are great. Thank you very much for sharing Roger.

[00:26:59] JOS:
Yeah, they are great three points because this is the best practice for, for PLM implementations general. And now I want to dive a little bit in the future is data and model based. How much is this also part of your journey?

[00:27:13] ROGER:
Yeah. Model based. Model based is high in attention. And we do want to be our future ecosystem, PLM ecosystem to let it be model based. We also realize that translating all the legacy that we have into a model based future is not the way to go. So we did some investigation into that.

That is not the sound way to go. So what you need in our future eco PLM ecosystem is we need to be able to manage all of our legacy in the classic way. And at the same time we need to make a step to do our future development model based. I think the application that we chose on the, on the, what we call it, the PDM layer as, and then the layer that is supporting the idea to market process which shows for a platform that will be able to manage a model base and which is basically natively supporting model based.

So that gives us a good, a good starting point. And in developing our processes that we are now in the beginning phases of, we need to, uh, to take the right steps to do that.

[00:28:20] JOS:
Okay, yeah, interesting and, and indeed very pragmatic, because the legacy data lives there long.

[00:28:26] ROGER:
Pragmatic is a good word. I think, you know in a business reality you need to be pragmatic because in the end the business needs to run and that's your first priority, always.

[00:28:37] JOS:
Right.

[00:28:38] HELENA:
One last question from my side, Roger. You mentioned that, uh, you have collaborated with Jos in terms of preparing for the journey. Is there other resources, books, or some kind of, um, Blocks or, or, or something that you have used to get ready for it. Or is it mainly really experienced and a good partners?

[00:28:58] ROGER:
Good partners for sure. So there was a lot of value in, uh, in, in working with Jos in the division definition. We also worked quite a bit with some data, which also was a, was a great resource. And then you know, there is no PLM school. As far as I know it is a lot of fact finding there is loads of information available, there is white papers, there is colleagues around the globe that, uh, you can talk to. I tend to pick up quite a bit of information on, uh on seminars and the like, conferences. In the Netherlands, there is a something that is called the PLM platform, which is a get together of Dutch companies who are, uh, dealing with PLM. And we are coming together four times a year. And, it's, let's say, we do, obviously it's networking exercise as well, but it's also a topic based presentations from users to users.

And, uh, what we do as well as we also include the vendors and data. So, uh, so it's, it's basically a get together in the Dutch area of everybody that has something to do with PLM and that brings a lot of knowledge as well.

[00:30:09] HELENA:
Perfect. Thank you very much. Excellent. I think this was a very good conversation. A topic that many, many companies, um, have at some point in their life. And I think that you have shared a lot of insights and tips for others to get prepared and to go through the journey. Jos, is there any final notes from your side that we should add?

[00:30:30] JOS:
No, I think we should listen again to this podcast. It was a very nice explanation of how to do a PLM journey. And I think, uh, as I said at the beginning, Marel was always in the past, also advanced in their PLM journey compared to other companies. And as Roger said, and pragmatic.

[00:30:47] JOS:
So thank you very much again for sharing it with us.

[00:30:50] ROGER:
Yeah, thank you for your kind words. It was a pleasure.

[00:30:53] HELENA:
Thank you, Roger.