United Manufacturing Hub ·

Plant-centric Infrastructure and Unified Namespace: An Interview with Bodo Körber

Jeremy, the co-founder and CTO of United Manufacturing Hub, sits down with Bodo Körber from Bain & Company, discussing plant-centric infrastructure, Unified Namespace, and the future of manufacturing.

Plant-centric Infrastructure and Unified Namespace: An Interview with Bodo Körber

The landscape of manufacturing is undergoing significant changes, with the integration of advanced digital technologies reshaping operations across the globe. As part of our ongoing exploration into these changes, we recently had the opportunity to sit down with Bodo Körber, a partner at Bain & Company, for an in-depth discussion about the future of manufacturing.

In our conversation, Bodo brought his extensive experience to the table, sharing his insights on plant-centric infrastructure, unified namespace, and the convergence of IT and OT in the modern factory.

What is plant-centric infrastructure? How does the Unified Namespace play a role in industrial applications? How are IT and OT merging in the new-age factory? These are just some of the questions we delved into in our conversation.

Highlights of Our Conversation:

  • Bodo's Journey: We began with a look at Bodo's journey in the manufacturing sector, from his early work in Rheinmetall and McKinsey to his current position at Bain & Company.
  • Plant-centric Infrastructure: We explored the concept of plant-centric infrastructure, a layered approach to decision-making and data computation in industrial operations.
  • Unified Namespace: We discussed the impact and application of the Unified Namespace in the world of manufacturing, a crucial element in overcoming the heterogeneity of industrial environments.
  • IT and OT Convergence: We touched on the increasing convergence of IT and OT, a significant shift impacting the organization and operation of modern factories.
  • The Factory of the Future: Lastly, we shared our thoughts on the factory of the future, driven by Industry 4.0 technologies and IIoT, which are set to transform manufacturing.

Interested in these topics? Take a moment to watch the interview and hear Bodo's insightful thoughts on these matters. Transcript of the full interview can be found below.

We'd love to hear your take on these topics! Comment below and let's keep the conversation going.

Want to Discuss With an Expert?

We understand the complexities of IT/OT architecture and we're here to help.

Contact Us

The full transcript

Jeremy (00:00): Hi, this is Jeremy, co-founder and CTO of the United Manufacturing Hub. We're here today at the offices of Bain & Company in Düsseldorf, Germany. And with me is Bodo Körber, partner at Bain & Company.

Bodo (00:14): Thank you very much, Jeremy, for making this happen after nearly one week being with Bain. I'm really looking forward to our discussion about what's going on in the industry and technology.

Jeremy (00:28): To get us started, what do you like to do in your spare time when you're not working?

Bodo (00:36): Well, in my spare time, I enjoy driving my motorcycles. I actually have three of them. One is a sporty Kawasaki Z900IS, styled after the 1973 Z1900. Then for cruising, I have a Harley-Davidson Fat Bob. And for long-distance journeys, like going to the North Cape, I have a Kawasaki Versys, which is a long-distance bike. How I decide which bike to take? Sometimes I plan it and other times I decide spontaneously when I'm in front of my garage.

Jeremy (01:20): That's interesting. Now let's shift towards your professional journey. Your background is quite impressive. You started at Unilever and were recently at McKinsey. What have been the most important experiences in your professional journey? I, along with our readers and listeners, would love to hear your stories.

Bodo (01:39): Well, there are a few stories I could share. One significant experience was my time at Rheinmetall Germany, in the automotive branch. I was a member of the management team of Rheinmetall Information Systems and was responsible for all IT and OT services for Rheinmetall Automotive. This included ERP, PLM, CAD services, EDI, and all OT services in the plants. My task was essentially to digitize the entire group end-to-end, covering all group entities from R&D to aftermarket sales and services.

We implemented several first-of-a-kind solutions, such as the direct integration between SAP, PLM, and the SO systems, Catea. This was necessary to track all changes that construction people made on the models and maintain consistent cost calculations. Additionally, we implemented SAP SCM, also known as the Advanced Planet Optimizer, at Pierburg.

We made group-wide implementations with SAP and other solutions, including MES and line monitoring systems. This experience fuelled my passion for IT and OT. We carried out vertical integration for many solutions in the manufacturing sites of Rheinmetall Automotive, which was my starting point.

Another major project was implementing SAP ERP and SCM at BMW in component and vehicle manufacturing plants. We ran the entire process chain, from taking over the configured cars to supply to line and production systems in the plants.

Jeremy (04:23): So, you practically oversaw the entire enterprise architecture for these companies, covering the entire manufacturing cycle with PLM, as well as the supply chain. Is there any area of enterprise architecture you haven't touched yet?

Bodo (04:36): I can't think of any. We pretty much covered everything. A notable achievement was establishing the first private cloud in Germany for Rheinmetall's defense and automotive parts in 2002. Back then, it was termed as outsourcing but today it would be called a cloud.

Years later, while at McKinsey, we shaped the industrial cloud and digital production platform based on technical and business requirements. We also built the partner ecosystem for the platform, including suppliers and other contributors. Simultaneously, we worked with production and logistics to identify and validate more than 100 use cases. These efforts contributed to reducing manufacturing and supply chain costs. As this was a cloud platform, it was one of the first industrial cloud projects globally.

The project started in 2018, with AWS selected as the IT cloud platform partner and Siemens as the integration partner. We set up the entire governance, development processes, and rollout logic for the plans in terms of which use cases should go first and in which plants. We also assessed which plants were mature enough to connect to the cloud.

We carried out an analysis of IT and OT within the plants, focusing on aspects such as manufacturing cells that could be connected to the cloud, even from a security perspective. We looked at PLCs, clock times, data flows, and security specifics - everything.

Jeremy (07:05): And how much is it integrated with the topic of connected cars?

Bodo (07:10): That's a different topic right there for the VW is finding a different cloud.

Jeremy (07:14): Okay, okay so you basically build up everything for them and architected for everything for the manufacturing, production, logistics?

Bodo (07:21): Exactly

Jeremy (07:23): What brought you to Bain and Company in your current role? What are your responsibilities there?

Bodo (07:30): I'm leading a unit called Advanced Digital Technologies for Industrials. We aim to provide an unbiased view on the architectures and business processes of our clients, which spans across areas like engineering product design, manufacturing operations, supply chain, and product services for manufacturers.

We also focus on companies with a large installed base in the field, seeking solutions for remote operation.

Jeremy (08:09): You recently posted an article of Bain on LinkedIn called Factory of the Future: How Industry 4.0 and AI Can Transform Manufacturing.. It ties in well with my next question, which is: how do you see the factory of the future?

Bodo (08:33): My colleagues, Jörg, Thomas, Frank, and Christian, and I have been busy translating current observations from ongoing or recently completed Bain projects into this report. The factory of the future is already embedded into the supply chain. We aim to be more flexible, with line-less mobile assembly systems replacing fixed lines, and the increased use of mobile and stationary robots.

There are technologies that can help improve quality, such as AI-powered visual inspection solutions.

Jeremy (09:39): You often talk about a plant-centric infrastructure. Is this what you're referring to?

Bodo (09:46): From the architecture standpoint, it's crucial to understand what's happening at the machine level and what can be computed and decided on the edge, where data is gathered and the PLC decides what to do. That's what I mean by edge-centric.

Plant-centric refers to what can be analyzed and computed at the plant level. With a tiered approach, we can decide which services to use from the cloud and what data to send to the cloud for computation.

One size does not fit all. At Bain, in my unit, we focus on crafting the ideal architecture for each client.

It's not a vendor-driven architecture. Each vendor wants to position their architecture in terms of footprint, subscription, fees, and so on. We see the need for an unbiased perspective to develop robust, scalable, and economically reasonable architectures.

In my automotive background, and at Rheinmetall, we had to consider each investment. For instance, we developed a line monitoring system that could detect air inclusions during the molding of motor blocks.

It was an effort that cost around 14,000 euros. We had to meet these requirements without any luxury, just delivering what was needed. That's a typical automotive manufacturing perspective.

We aim to provide what's needed to optimize the process and product, not overwhelm the people. This down-to-earth perspective is what we aim to bring to our clients.

Jeremy (12:16): How do you balance creating a value-add and focusing on a use case, while building on a scalable architecture? These seem like different workstreams.

Bodo (12:34): The balance is in thinking of the ideal architecture right from the beginning, considering not just one use case, but enabling a set of use cases. The architecture is agnostic to the use cases that can be built on it.

We also look for the most distinctive partners in different areas.

I'm not talking about the size of companies, but about unique control points that can solve things in favor of and better for the benefit of our clients. We are also looking at architectures that can be scaled across a plant network, given that the manufacturing world is often completely heterogeneous due to M&A or local sourcing decisions.

No plant is similar to the next one in terms of the software and hardware. We have to find approaches and solutions that extend, rather than rip and replace, and also serve our clients with intermediate architectures.

Even if we design a target architecture, often our clients do not have time. They'd like to see a payback within six months.

Should they wait for the ideal target state regardless of the rollout needs from the ERP, SM, etc.? No, we have the assets, approaches, and methodologies to define architectures for the intermediate states, upon which we can build and evolve.

Jeremy (14:55): So, as a final question on the topic of plant-centric infrastructure, how do you strike the right balance between architecture and use cases? Maybe just in a few sentences.

Bodo (15:07): For any kind of use case, there are requirements from the technology. This means considering what kind of data sources and applications should be addressed, or if there's a need for a completely new application. We also see composite applications where the use case needs data from multiple systems, essentially composing a new one and creating new results.

It must be clear in our analysis that all these requirements are being assessed, and this will lead to the architectural perspective.

Jeremy (15:44): Alright, so we've discussed plant-centric infrastructure and how to strike the right balance between architecture and use case development. Now, what would be an example of such an architecture?

Bodo (16:00): Let's talk about one enabler that has grown in importance in recent years – an event-driven architecture, specifically the publish and subscribe technology.

When you look at the traditional ISA95 layers, the applications were only allowed to send data from one layer to the next one. This could be from the ERP to the MES, and then down to the device level.

This is being overcome with the technology set that we have, and a Unified Namespace, for example, also helps to overcome heterogeneous architectures.

Given that 80% of the industrial space is made up of brownfield environments and only 20% greenfield, a key question for any CEO or CIO is how to deal with their numerous heterogeneous plants.

The answer is to start with what you have and build on this with a rapid extent approach, leveraging technologies like the unified namespace. Over time, you can build your target architecture on these technologies to support your use cases.

Jeremy (17:41): I think most of our listeners already know the concept of a unified namespace. They would be delighted to see how you explain this to a CEO. But thinking about this, it's not just a technical thing. What are other important levers toward digital transformation? Because it's not only about technology. I would even say technology is the easiest. What's the hard part with organization and skills?

Bodo (18:14): One example is the convergence of IT and OT. The people from IT and those from the shop floor also need to work together.

Here's a practical example. The shop floor IT usually reports to the CIO, and then you have automation technology and plant maintenance reporting to the CEO or plant director at the plant level.

We need a joint ITIL process where any kind of incident and problem management will be handled jointly. This requires plant maintenance, automation technology, and shop floor IT working together.

We need to identify what the root cause is. Is it the network? The PLC? An application? The database? This calls for the best methodologies that we have already implemented in business IT, including ITIL and other methodologies and technologies.

So we need to blend the workforce together, and they should work on a unified process in order to support the plant in the best way, reducing downtime and ensuring that everybody knows what needs to be done.

Also, with IT-OT convergence, there needs to be alignment on technologies and security. It's not just about OT security or IT security anymore. If you connect the networks, they need to be aligned on how to handle security.

In terms of IT-OT convergence with ERP and the MES systems in the plant, there's always a discussion about the handover. That is, how far you go in the ERP in terms of the capabilities and functionalities that you're using, where the handover in the plant occurs, where the plant system takes over, and when you go back into the ERP.

This can also vary based on the type of product you are producing in a plant, the manufacturing type, and whether it's make-to-stock or engineer-to-order. That could lead to different kinds of architectures that you need.

Bodo (21:01): Jeremy, you are also doing really highly innovative and inspiring things. So, what is your vision to help companies scale their architectures, leveraging the Unified Manufacturing approach and technologies?

Jeremy (21:15): Yes, in terms of technologies, we try to be conservative. I know it sounds a bit strange coming from a startup, but I believe this is what the industry needs at the moment.

They need best practices. What's new for them is, for example, what you've talked about. I can only concur. For OT and IT people that need to communicate, our approach is to adopt the best practices from both OT and IT.

Both sides present something new for the other. But in general, it's considered very basic.

To give an example on the technology level, to scale and send a large amount of data through the architecture, we are leveraging the systems used by IT giants like Google, Amazon, and Facebook, such as Kubernetes, Docker etc.

Kafka is another system. These systems are proven to work but have not been widely used in OT.

But in IT, they're just standard. This is one of the examples where we say, we just use established technologies, just in a new context.

In general, to make this scalable, we have the technology aspect, like best practices for OT and IT. From an organizational point of view, it's not only about the technology, it's also about making it usable for the people.

Let's take the example of Node-RED . It's a good example because it can be used by both IT and OT people.

It might not be the best tool for each of them, but both departments can work with it.

We also see large enterprises using such tools to scale through their factories. They know that whether they're in Brazil or Singapore, they will find people who can work with these tools and implement them.

This is what we're doing at the United Manufacturing Hub, using IT and OT best practices, combining them, and packaging methodologies and tools that are open source.

Everything that we do is open source. The entire core, the source code, is available, which helps us build a great community around it.

We are ecosystem agnostic by using vendor-neutral technologies such as MQTT. It's really easy to connect different ecosystems. It doesn't matter if you're from Siemens, Honeywell, or others. Almost all of them have some connection with MQTT.

Our third unique point is end-to-end. It's not only about data infrastructure or the organization, but also about where this application should run, how to maintain it. You need to run, for example, an edge device and have someone maintain it.

We also consider what types of tools and technologies are being used and put all of this into a central interface, so people in the factory can use these best practices without extensive training.

Jeremy (24:20): Coming to the final question, why should companies choose Bain and not try out the topic of architecture themselves or hire a system integrator? Why should they come to you, for example?

Bodo (24:34): I joined Bain with a clear mission in the advanced digital technologies business unit for industrial companies to be the advocate of the client.

I think I mentioned this already in our previous discussion. We focus on robust, scalable, and economically reasonable architectures.

We don't receive any fees from vendors. We also don't have the overhead of managing hundreds of thousands of application management people.

Right, that's one. The second one is that we will establish a partner network, and we already have a lot of distinctive partners. We will continue finding the pearls in the partnership landscape, giving us and the clients uniqueness in terms of the control points that help them build and scale these architectures.

Also, applications from the PLM side going down to install base management for machine builders, as well as in the MES space and so on.

We are building a special co-innovation ecosystem where we invite our clients and partners to work on specific challenges and solutions under NDA.

This means that everyone is open to share roadmaps and solutions. We also come to initial projects to test scalability, not just pilot.

Number three is that we will meet the clients where they are. Everyone already has a status quo of application architectures and rolled-out use cases and processes. We won't add complexity but instead accelerate the process and reduce complexity.

The fourth point is very important, which is bridging for the benefit of the client between technology, process improvement, operations, industry paradigms, and best practices.

We have a seamlessly integrated solution and experience, unlike a system integrator that asks, "What should I do?" and works with various frameworks and partners. We are aligned and have one framework that we use.

Jeremy (28:05): Why should a company not do it themselves? Well, first of all, we have the overview of the most applicable solution, methodology, and technologies.

Bodo (28:23): The other important element is the change aspect, which should not be neglected. Enabling technology is not a self-fulfilling prophecy. We also address the organizational challenges and change, ensuring value capture.

Jeremy (29:08): Thank you, Bodo, for inviting me to be here. And thank you for being on this podcast episode. If you're listening or viewing this, feel free to leave a comment under this YouTube video or wherever you find it, and share it with your senior management. If you're a senior management member watching this, feel free to contact Bodo if you have any questions. You can reach him via LinkedIn.

Read next

Node-RED meets Benthos!
benthos · Featured

Node-RED meets Benthos!

Yes, we’ve made it possible to bring Node-RED-style simplicity into Benthos pipelines, enabling your chosen LLM to handle tedious manufacturing connectivity and contextualization tasks—whether you’re prototyping or managing production-grade data flows.

Share, Engage, and Contribute!

Discover how you can share your ideas, contribute to our blog, and connect with us on other platforms.