Digital Manufacturing Platform: ZF plants are now more efficient thanks to the smart factory solution based on Microsoft Azure

The challenge: Making better use of the efficiency benefits of digital transformation

At ZF’s 188 production sites, the various departments work closely together. ZF employees keep production running as smoothly as possible in the company’s four technology domains—Vehicle Motion Control, Integrated Safety, Automated Driving, and Electric Mobility. To do so, they must adhere to clearly defined workflows. For the 400-strong production team at ZF’s Diepholz pilot plant in northwestern Germany, such workflows used to include going to the machine control consoles to manually collate the number of parts produced and any machine downtimes during a given shift and entering that data into a central PC. On average, this took up around five minutes of everybody’s time at the end of every shift. Tooling—in other words, setting up a machine for a particular processing step—also requires a certain degree of documentation work. It was a similar story for the maintenance team: work orders and notifications from the SAP system had to be printed out. Loaded with these printouts, employees would go over to the malfunctioning machine, for instance to connect a diagnostics device to find the source of the error. Once identified, it was a case of fixing the problem and then reporting the outcome. Printing out the information, attending to the machine, and feeding the new data into SAP would occupy a member of the maintenance team for around 15 minutes. After all, having to physically move from place to place, and the formalities that go along with that, eat up time that would be better spent on more beneficial tasks.

At first, each ZF plant came up with its own IT solution for this kind of challenge—a consequence of the Group’s rapid growth. “From a Group perspective, it was inefficient. It clashed with the whole ethos of production, which is to continuously improve efficiency,” says Joachim Sprinz, Program Manager IIoT – Digital Manufacturing Platform at ZF. “Digital transformation allows us to generate more value from our data, which in turn raises the efficiency and performance of our plants. At the same time, this transformation ensures company-wide connectivity along the entire value chain.” What was needed was a solution that connected data from different sources and brought together the full potential of digital transformation on a single platform. It was crucial that this standard solution could be expanded to cover all plants to create data transparency and lean processes both throughout the company and at each individual plant. This was the root idea that grew into the DMP.

The solution: A platform in Microsoft Azure for data-driven efficiency at all plants

“The first step toward raising efficiency across all plants is to achieve a synergy of people and data,” Sprinz says. “This is the only way to accelerate our production’s digital transformation and establish a global digital production network.” Microsoft Azure provides the basis for all of this. “Our Group follows a cloud strategy built on Azure. Since our partnership with Microsoft was already well established and we were familiar with its products, Azure was the most sensible choice as the basis for our DMP,” he adds.

Today, ZF collects production machine data with Azure IoT connectivity components, stores it in Azure Data Lake, and then uses Azure Data Explorer to prepare and provide it for analysis. User management is through Azure Active Directory: users open DMP web applications in the front end to view the KPIs presented on easily understandable Power BI dashboards. The DMP is being upgraded all the time and parts if it are already being used at 18 plants, including Diepholz.

The Diepholz plant is among those striving to raise machine efficiency, improve its management of material stocks and flows, and make better use of resources. Today, employees there are aided by the applications Worker Cockpit, OEE Analyzer, and Condition Monitoring. Worker Cockpit helps them log production losses during a given shift, almost in real time. Machines are equipped with a display that allows employees to see at any time which order is currently being filled, where losses in quality have occurred, and where a machine was idle. Machines automatically document when and why they malfunction and send this information to the DMP. “Smart algorithms speed up the process, making the information more transparent and available sooner,” Schulze says. “This also makes the tooling process more comprehensible, for instance. The status and overall process time are displayed immediately.” Next, experts and planners evaluate the Worker Cockpit data using OEE Analyzer; the results provide the basis for optimization. “This is how, for example, we see if a machine isn’t running as well as it could, and we can intervene directly. Once optimized, the machine runs more efficiently and performs better. But before all that, we have to be able to see the problem, which is precisely what our Azure-based DMP allows us to do,” Schulze says.

Preventing production downtimes and optimizing inventory

Using Condition Monitoring, the Diepholz maintenance team can now identify and rectify problems with the machines faster than before. If the data generated by a machine exceeds a particular threshold, a push notification is automatically sent to the team. This notification contains all the information required to tell at a glance which machine and which parameters are affected. Experienced staff can then decide which tools or spare parts they will need for the repair before they have even left their desk. “This saves them unnecessary trips between their desk, the spare parts store, and the machine,” Schulze says. “Impending downtimes are also detected in advance. Let’s say there’s a machine fitted with two drilling tools and the data analysis shows that one is running at a significantly higher temperature than the other. We look over the details and order any spares in good time. This allows us to prevent unscheduled downtimes that, in the current supply situation, could potentially last several weeks.” In the future, ZP plans to introduce automatic anomaly detection to supplement its employees’ expertise.

“The DMP program has had a profound impact on how our company works. Business and IT collaborate more closely, we’ve set up expert forums, and we’re sharing expertise across all levels,” Sprinz says. “We’ve established the processes, mindset, and structures we need to flexibly scale the technology and how we manage it.” After all, the idea was for the DMP to be more than merely a Group reporting platform, but rather an IIoT platform that doubled as a central data and application hub for all 188 ZF production locations around the world—a single point of information for flexible, efficient management of all factories. This also paves the way for achieving yet more optimization, eliminating tasks that don’t add value, and driving forward company-wide connectivity in the future.