Longtime innovator SKF unveils the factory of the future on Azure hybrid cloud

Giving factories ‘cloud-controlled, locally executed’ power

Longtime innovator SKF sees an opportunity to drive the manufacturing industry forward with the digitalization and modernization of factories. With a network of 17,000 distributors spread across roughly 130 countries and regions, the company began planning its data-driven manufacturing vision five years ago. Aging factory hardware was due for an upgrade. Many manufacturing systems ran on legacy software. Strict requirements for data latency, resiliency, and sovereignty posed a compliance challenge. In addition, over time, local factory sites had adopted a variety of stopgap solutions that met temporary needs but became increasingly complex to manage globally.

According to Sven Vollbehr, Head of Digital Manufacturing at SKF Group, the industry is still defining exactly what digital manufacturing means, and SKF is taking the lead. “It’s ultimately about empowerment of the digital workforce and giving them the tools to have better analytics, better visibility into performance, and the ability to use that for quicker decision making.”

As part of its Future Factory initiative, SKF began planning what Vollbehr describes as a “cloud-controlled, locally executed” environment. The idea is to give factories a standardized IT/OT platform capable of accelerating digital innovation. Speed, reliability, and low cost are critical as is the need to support vital production systems. But as Vollbehr points out, “There is not an off-the-shelf product that we can apply and just be done with it.” After an evaluation the marketplace, SKF turned to Microsoft for help.

For years, SKF has used Microsoft Windows Hyper-V virtualization technology and Microsoft SQL Server databases. “When we look at Microsoft, we can see that it’s becoming a complete strategy for edge computing and managing your data and even security—across the cloud boundaries—from a single pane of glass,” Vollbehr observes. “This is the appeal.”

Using data on the edge to close the manufacturing loop

SKF started the Future Factory initiative two years ago and tasked it with putting into practice the vision of a reliable, lean, and digital value chain to accelerate world-class manufacturing. The team of cloud specialists set out to redefine the infrastructure layer at factories with the adoption of Azure Stack HCI. This hyperconverged infrastructure (HCI) operating system is delivered as an Azure service and provides an on-ramp to cloud services.

According to Mats Iremark, Lead Solution Architect for Hybrid Cloud at SKF Group, “The responsibility of Future Factory is to set standards and develop new ways of working in the near term and in the future. We came up with a design for a hybrid cloud platform which is located in Azure and has a footprint in the factories.”

Azure Stack HCI runs the services that SKF needs to gather data and draw conclusions from that data, Iremark says. “We can't do that with the old platform. We need a new one that we can manage with the same convenience as we manage the cloud.”

On the factory floor, machine and sensor data are rapidly collected and streamed into Azure SQL Edge, a database and streaming engine optimized for Industrial IoT (IIoT) and edge deployments. Data is then routed to Azure Arc-enabled SQL Managed Instance, an Azure database service that runs on-premises at SKF factories. It works with or without a connection to Azure—a key benefit in securing uninterrupted operations at SKF factories even where connectivity is spotty.

“SQL is at the very heart of everything we do,” Vollbehr notes. On the factory floor, Azure SQL Edge and Azure Arc–enabled SQL Managed Instance create an efficient closed loop that keeps data-processing power where it’s needed. SKF can store and process manufacturing data at the source, providing real-time access to information used, for example, in machine learning and AI models. This approach creates a virtuous cycle of improvement. As Vollbehr sees it, “The closer you get to that loop, the more you start finding other value opportunities.”

SKF’s hybrid cloud platform provides the cost-effective benefits associated with Azure, including elastic scale, cloud-style billing model, and evergreen software. For example, Azure Arc data services receive updates on a frequent basis, including servicing patches and new features similar to the experience in Azure. Instead of the typical end-of-support moment associated with on-premises database software, Azure SQL data services stay up to date. “That alone gave us both speed as well as good cost savings compared to our previous approach,” Vollbehr adds.

Scaling a hybrid cloud platform across 100 sites

Together, Azure Stack HCI and Azure Arc technologies solve a critical issue for SKF’s Future Factory organization. How do you scale a hybrid cloud platform across approximately 100 sites?

“The only way we can solve this problem is by using technologies that Azure Arc provides,” Iremark states. “The ability to do automation and to say, ‘I want this applied to 100 sites’—that is provided by technologies like Azure Arc.”

Azure Arc offers a unified management view of SKF hybrid cloud resources. Teams can organize and govern the company’s Hyper-V servers, Azure Stack HCI infrastructure, Azure Arc-enabled data services, and more from one place.

“It's a key for me,” Iremark says. “With the same effort to manage the compute platform at one site, we can now manage the compute platform at 100 sites. It would not be possible without the Azure Arc technologies and the GitOps workflow.”

SKF developers follow infrastructure as code (IaC) practices that automate infrastructure deployment and ensure the quality and consistency of builds. Developers create code and share work in Azure DevOps Services, and then use GitOps to automate best practices, like continuous integration (CI) and continuous deployment (CD).

At factory sites with Azure Stack HCI, apps run in clusters hosted on Azure Kubernetes Service (AKS), enabled by Azure Arc. The integration of GitOps lets teams quickly scale out and manage operations across all the AKS clusters.

This level of automation is a game-changer for SKF. “We can make changes at scale, while the factory keeps the application running, uninterrupted,” Vollbehr explains. “For example, if we want to scale resources for a SQL Managed Instance in a factory in Italy, we can go into Azure DevOps and change the number of CPU cores that should be available, commit the changes, and watch the update take place right away.”

For details about the architecture, see the blog post, Digital transformation at SKF through data-driven manufacturing approach using Azure Arc–enabled SQL.

Looking to the future

For SKF, hybrid is the path forward. SKF estimates that over a five-year period, its hybrid cloud platform alone will lower the total hardware-related costs by roughly 40 percent and OT-related downtime of its machinery by approximately 30 percent. Vollbehr considers the new platform “one of those remarkable additions into our technology portfolio.”

The company’s strong partnership with Microsoft also made a difference. “Our strategic relationship with Microsoft has been very collaborative and productive,” Vollbehr notes. “Microsoft puts itself in our shoes, meeting our needs, and delivering the digital solutions we want.”

Looking into the future, SKF Group will explore solutions that incorporate Azure Machine Learning among other technologies, to optimize further the manufacturing engineering processes.