SKF’s AI springboard into a new business model

IT leaders at the Swedish industrial machinery manufacturer work in tandem to ensure that tech development, automation, and digitization of the value chain are in lockstep with the business.

Three years ago, Gothenburg-based SKF brought in new CEO Rickard Gustafson and over time, a new strategic framework was established to take the company from a centralized to a decentralized organization. And at the heart of that new business model journey was, and continues to be, CIO Pedro de Freitas and group head of digital engineering and innovation, Maria Stranne.

“We see the value in having the IT aspects close to the business where support is needed,” says Stranne.

For de Freitas, the continuing focus is on getting the entire company to a common business system platform, and by the end of 2025, he expects it to be introduced worldwide amid the longstanding phasing in of S/4HANA. “In the last 18 months, the work has accelerated,” he says. “We’ve gone from having 20% of our business on SAP at the end of 2022, to 77% today.” And the next important stage is when SKF moves to SAP’s cloud solution, Rise.

“It’s a huge step for us,” says de Freitas. “We leave complex license infrastructure and service levels up to SAP’s one-stop-shop in the cloud. It gives us a better opportunity to benefit from its news and latest releases in security, functionality and technical stability.”

Strategic AI

Finding the right applications of AI is also high on the agenda. “Applied AI is one of our strategic technology priorities,” Stranne says. The focus is on three areas, the first of which is about how AI can be used in various tools such as those SKF can get via SAP’s Rise but also via, for example, Microsoft. “Right now we’re testing M365 Copilot to a limited extent and gathering user input to understand where the value for the company lies,” she adds. “At the same time, it also means that our employees gain an understanding of what AI can contribute to the tasks they perform every day.”

The second area is SKF to develop its own AI models within its core business. For example, over the last couple of years, the company has been working with condition monitoring, or checking the condition of their machines during operation to see when it’s time to maintain them, and make sure the right parts are in place. And in order for employees to test new models, SKF has set up separate environments, so-called sandbox environments, together with its cloud providers, where it’s possible to experiment in a safe way. There’s also a community where interested employees can discuss and exchange experiences, or maybe test similar areas of use together.

The third area is about getting the most effective AI strategies in place in order to go forward. “What should we focus on? What major global use cases do we see that we think can have a big impact?” says Stranne. “We put a lot of emphasis on that right now.”

Guardrails for AI

She and de Freitas also work closely together to get how the use of AI at the company should take place in a safe way, so there are certain frameworks set up.

“It’s good that there’s such a high energy, but we also have to make sure there are safeguards, because SKF is a big company that has to keep risks down,” says Stranne. In addition, AI is a large data consumer but also generates a lot of data, and a good infrastructure that can handle such massive amounts is required, says de Freitas.

“To ensure we have a good understanding of everything we have in production where AI is involved in one way or another, we’ve also created an AI governance board,” says Maria Stranne.

Data in the wash

But high-quality data and access to it is required for AI to make a real difference.

“We’re far from washing all our data, but we’re halfway through validating our most important data,” says de Freitas. “As far as our supply chain is concerned, we’re almost done.” It’s also where he sees the greatest potential going forward for AI-powered analytics.

“There we keep track of all the different variables to put in, from the purchase of materials all the way to the best way to ship our products,” he says. “So when we have a handle on all this data, that’s where I see the biggest potential.

Another important thing is to build data competence within the company. “We need to make sure everyone understands what kind of data they produce and that it supports the future needs of the company,” says Stranne. “It’s something that must exist in the entire organization.”

Welcome democratization

In its 2023 annual report, SKF says that with the ambition to become a data-driven company, they invest largely in gaining actionable insights throughout the entire value chain with the help of advanced analysis. Today, employees use Microsoft Power Apps to optimize and streamline their own workflows, and it’s part of democratization that Stranne welcomes. “It’s the core of what we want to do, and it has two sides,” she says. One, she explains, is the democratization for everyone where guardrails are in place for what can be done. There, a control method has also been created so when it gets above a certain set of variables and thresholds, it’s considered a business-critical application that should be treated in a special way when it comes to scaling up.

“The second is, through this type of democratization, to be able to better utilize the power of a new generation that comes to our industry, and has lived with technology since they were young,” she says.

Digital twins and quantum technology

Alongside AI and increasingly advanced analyses, the use of digital twins internally and externally is also a priority.

“Internally, it’s linked to virtual verification and is used in our development work,” she says. “Externally, it’s aimed at our customers, and today we have fantastic simulation possibilities that we’ve built up over many years. Here we see great potential in how we can connect these to data from the real world and create even more value.”

She also points out so-called spatial computing, such as AR and VR, metaverse, and quantum technology. “All of this is, of course, very interesting for us,” she says. “And in terms of our materials research, we follow quantum computers with great interest as well. We need to look at these other technologies and understand how we can leverage them.”

The global domino effect

de Freitas also points to how quickly the technology landscape is changing, but also to how other factors outside of technology have impact.

“Two years ago, we wouldn’t have talked about a trade war, and we have important elections in the United States coming up,” he says. “It’ll mean a certain strain on our value chain in the way we buy our raw materials. We have a crisis in the Red Sea, too, where ships can’t pass, as well as legislation in the works. So those are things that push the boundaries of manufacturing and rethinking models.” And that, in turn, has effects on how the technology is set up going forward.

“We have to think about things like how we should relate to SAP single instance in the light of geopolitical changes,” de Freitas adds. “As volatility increases, we need to look at distributed instances. I’m not saying that’s the way we’re going to go, but these are the impacts geopolitical developments have on technology. And that’s also one of the reasons why the supply chain is so important.”

SUBSCRIBE TO OUR NEWSLETTER