How Volkswagen is Transforming its Global Manufacturing on the Cloud

Nihar Patel: When we started, we had 120 manufacturing locations churning out close to 11 million vehicles a year, and over time, each of those plants had created its own IT structure and built on it. And when that happens, you gain plant-level autonomy, but you lose the IT infrastructure consistency that allows you to work across all your plants. We saw new digital and cloud technology and realized if we were going to make a change to drive efficiency, a bold move was necessary.

We quickly realized we’d need new partners to help us, and AWS is a key one. Our initial focus was our production systems, so we instituted a carrier architecture called the DPP, our digital production platform. Then we had to bring across the rest of the company. Logistics, supply chain, and purchasing would all require the same sort of attention. But we had to prove the case with the production platform, and that’s where Frank got involved.

AWS: Doug, is this common in other companies, or is it unique to Volkswagen?

Doug Bellin: I think many big companies struggle to build once and scale to many locations. Plants are not typically built with technology in mind, and most existing employees are not technology experts. So, you have to hire for a new skillset or take people away from their day job, and costs start to get a little crazy. That’s what the partnership is really bringing — the ability to build a use case and scale it going forward.

That’s something AWS brings to many customers. Our marketplace is not just big companies, but small companies too, and there are thousands of solutions already built. It could be a simple OEE (overall equipment effectiveness) dashboard. It could be a predictive maintenance offering. It could be many different things, but now a company can go in, quickly find their solution, implement it, and keep moving on with the continuous improvement capabilities that every plant wants.

AWS: Where do you start with this? How did you decide which use cases and mission-based projects to start on right away?

Nihar Patel: We had a list of things we knew needed fixing and solutions that worked but needed scaling. The starting point was the platform. AWS lets you create a platform that you can build on and scale across the board.

Frank Goeller: The first questions were, what is the purpose? And what is the target we want to achieve? Without those, people will stick with the status quo, and things won’t change much. So first, describe why we need to change the whole setup of our IT system and infrastructure—not only with the IT people but also with production experts because they have to understand how it will affect existing processes. For example, with traditional quality-assurance processes, masses of people check cars, parameters, and so on. Industrial computer-vision systems that automatically check these things change how people work, and it must be clear in the beginning that this change will come. Also, consider which factories are willing to pilot the new ideas and which will see big performance improvements. Then take it step by step. Make a roadmap showing which people and which factories will be engaged and what the steps will be so that if something goes wrong, the whole system won’t break down.

AWS: Good idea. How did you build a rallying cry of advocates from various teams who would fight for this change and work through it even when things weren’t going well?

Frank Goeller: We’re still doing that. The three most important words are communicate, communicate, communicate. We have to show the impact of our change. In the beginning, people said we were spending so much money and using so many resources, and only a tiny impact was visible on the finance side. And that’s true. In the beginning, the apples you get from the tree are not that big, but they will grow, and then the appetite will also get bigger. The flywheel turns quicker. You have to show the pilots and the impact of the change, and then people will begin to ask for it.

When that happens, you’re on your way. But if you have to push people to let you do a pilot, then you are still in the incubation phase. At the moment, we’re between those phases; we have some factories and brands that are really behind it and others that are waiting to see how it develops. But I think if we talk again in a year, we’ll be celebrating a big success.

AWS: Can you share some examples of when something didn’t go as planned, but you were able to course-correct and learn from it?

Frank Goeller: We had a hard time, in the beginning, establishing the AWS training due to legal and administrative blockages, and it was painful because the training was available and there was good content we could deliver to our people. But now we’re fully on track, with more than 500 people training and the number growing every day.

AWS: It’s great that you focused on educating and training the workforce before bringing in the new technology. How did you choose the training?

Frank Goeller: AWS was a big help because they know the content and the technical elements, so we used existing elements from AWS combined with new content from our side. Talking about connectivity and how to link our machine world to the virtual world was the first step in the training. We have to develop more content because the number of applications will grow, and therefore the need for training and capability-building is growing.

AWS: Nihar, can you provide some examples of capabilities Volkswagen is focused on building with the workforce and how the company wants to grow its employees?

Nihar Patel: Sure. One of the most critical capabilities is software skills because we want to get more involved in the software world, and there’s no better partner to learn from than AWS. We’ve set up joint operations in Berlin and other locations to think through the industrial, coding, and solution requirements and then work together to build the expertise. We’ve invested in more education and training around software and created some additional partnerships for that. And we’re leaning into our own organization because we have a significant number of partner companies. One is MHP, a subsidiary of Porsche that does consulting, and we’re now muscling them up to see how they could provide more DevOps skill sets. So, things that aren’t part of our normal operations but that we know exist in the family; we’re building them through joint work with AWS and other partnerships because we recognize this is a great opportunity. But we have to move at speed; if we take the usual path, we’ll be trailing significantly.

AWS: That makes sense. Frank, what did you learn about yourself as a leader working through this challenge? Was there anything that surprised you?

Frank Goeller: One very new experience was working with the AWS teams on press releases and a narrative for our executive board and then seeing how the narrative was discussed and improved in a constructive way.

Another lesson was that you can never communicate too much while executing such a big project. You have to talk to all the stakeholders, always be aware that information can be misunderstood, and clarify and sort out any misunderstandings. That is a big topic to foster as a leader, bringing the right people together and keeping the departments and stakeholders aligned to the joint target because we have so many distractions around us. There’s always the chance that you lose the track, and you have to take care that the track is visible, that people understand the track. If there is a bypass or a hurdle you have to go around, you have to communicate that too. Fifty percent of my time is talking to people, bringing them together, and giving them the right messages.

AWS: How did you choose the right avenue of communication? Did you choose to speak in person for specific reasons, or would you sometimes use email?

Frank Goeller: My ideal world would be to have small-group interactions so that people can face one another and have a good exchange on the topic. With coronavirus, we relied too much on email, which is the communication channel I least prefer because it can deliver a lot of misunderstanding. The spoken word, face to face, is best. I’m a big fan of small workshop sessions. The sprint approach is a very agile method; maybe one very focused day, but then it’s over. Half-hour meetings jumping between 20 topics in one day are not that effective. I think the impact is strongest when you have a focus and stay on topic.

AWS: Definitely. You’ve mentioned a lot of processes that Amazon collaborates on with partners and customers. Doug, can we explain these a little, starting with the flywheel?

Doug Bellin: Jeff Bezos came up with the flywheel concept years ago when looking at amazon.com’s value creation. He knew if he could get the best selection, he’d get more customers. That’s two cogs in the flywheel. Then if he could get more customers and a better selection, he could lower prices. Lower prices would then get more customers, more customers bring more sellers, and so on. Now, how do you attract more customers within that cog of the flywheel? Marketing, outreach, and making it easier to work with you as a company.

We’ve looked at manufacturing and even higher-level capabilities the same way: If something is better for your users, they will want to stay with you and buy more of your product, and so on.

AWS: Also, Frank mentioned press releases. Can you talk a little about that?

Engineering and production were typically siloed. We wanted to tighten production planning and capabilities, but then how does that production information come back to help with the production design? What we try to do is work backward, and one of the capabilities is building out a press release just like you would for The Wall Street Journal or The New York Times. Now, I’m not saying we’re going to send our press release to them, but if you build a document that everybody agrees on, now you have a project plan. Our issue is x, and we did y to solve it. It’s a unique way to look at solutions, offerings, and capabilities internally that a customer typically may not look at. You define who the customer is, and sometimes, like with a VW, you may have three or four different levels of customer in the same organization—say a line worker, a plant manager, and the person who’s buying the vehicle.

AWS: Let’s look five, 10 years into the future. What would be your number-one piece of advice to any industrial manufacturing company facing similar issues?

Nihar Patel: You should have a bold vision for where you want to go, but one built around some proof points. The vision we have is to create the flywheel Doug described by working with smart partners, prove the case out at VW, and quickly migrate it across the rest of the industry. I think it’s important to keep that North Star view; we know there’s a demand for a singular repository of IoT solutions, and there has to be a pain-free way of getting access to that stuff. We know as a customer, we want it, so I think it’s critical to note the customer’s mindset and match it to the bold vision. Then having significant levels of communication—internal, internal, internal before you go external, and making sure the right governances are in place. Then you have to put a bunch of smart people and smart organizations together to achieve this vision. You might go some distance solo, but if you want a journey, you’ve got to partner with people all the way through. It’s a marathon.

AWS: That’s an excellent perspective. Frank, can you share your number-one piece of advice?

Frank Goeller: I really connect with the marathon metaphor. It’s not the running shoe or the trousers or the sunglasses that make you win—it’s the team around you and you personally, what you are doing, and how you are trained. It’s the same for any transformation. It’s not the technology we are using. It’s not the innovation we can bring to the transformation. It’s the people, the team, that’s driving it. That’s a big challenge because we have loads of people, and we have to select the ones who can create high enough energy to run it for several years. To change the whole network, we need a lot of energy, which can only come from finding the right people and putting them in the right places.