Machines raise their voice when they need help: Bosch’s Digital Twin – IAPM takes human–machine collaboration thanks to Microsoft Azure to the next level

The challenge: Not too early and not too late—performing maintenance at just the appropriate time

When Prahallad, Partner Customer Solutions at Bosch, talks about innovation, the excitement in his voice is audible. The point is to reduce complexity and make people’s jobs easier. And, as he says, it’s about “the most important machines anywhere things are produced.” When producing energy, it’s turbines in the wind-power generators that keep everything running; in factories in the automotive industry, it’s electrical systems. It’s not uncommon for these kinds of rotating machines to have an operational life of 10 to 20 years or more. That’s a long time, during which they undergo constant servicing. “To avoid risking a breakdown, parts are frequently replaced as a precaution at regular intervals, even if it isn’t yet necessary. That’s not very resource efficient,” says Prahallad, referring not only to the cost of components and their manufacturer, but also to the time maintenance experts invest that could be better spent elsewhere.

This observation spawned an idea that was promoted directly by Bosch’s own start-up incubator, Grow. After all, Bosch itself operates hundreds of rotating machines in its own plants. “We wanted to make it possible to see things that previously couldn’t be seen: when parts need to be replaced and what a machine’s condition is. In other words, to see in real time how the machine is actually doing”, says Prahallad. “From this, it was then possible to predict when maintenance truly makes sense.” Achieving this requires many different kinds of data, such as data that the machine itself provides through sensors, and data that comes from other sources, such as historical machine data and standard data sets from the manufacturer. 

The team wanted to create true added value, and beyond that, to reduce complexity and stress. “People don’t accept new solutions unless those solutions make their personal day-to-day work easier”, says Prahallad. “That would require additional automation and interaction with the machine. A red light when something’s wrong isn’t going to cut it.” So the team developed a Digital Twin - IAPM in the cloud that not only can be used to gather, analyze, and make predictions regarding machine-generated data, but also facilitates communication with the machine.

The solution: Focus on reliability centered and total productive maintenance for more sustainability in production

“We started with the most important kind of machines—the rotating ones,” says Prahallad. “Because if you master those, you also master all the others, and that makes it easier to scale.” But that’s not the only consideration. The market also offers a lot of potential, especially in Europe, as Bhuvan Shetty, Sales & Global Strategic Partnerships at Bosch, explains: “Around 53 percent of these machines are manufactured in DACH region. The biggest companies in the area of transportation, energy, process and discrete manufacturing industries are based here. Quality standards are extremely high.” In short, if the Bosch team can gain its customers’ confidence in its solution here, it can do this all over the world.

And that works. In part, that’s because Bosch has Digital Twins running in their own plants. “We know from our colleagues who work directly on these systems that the stress level rises when, for example, their machines are subjected to unusual loads. Our Digital Twin enables them to make data-driven decisions to avoid breakdowns from the outset. As a digital assistant, it really lightens the day-to-day workload.”

All relevant machine data is made available through the Azure IoT Hub, which is connected with Power Apps. Any time when a symptom is detected, the Power App triggers an alert. “Current solutions do that only when there are already initial signs of wear. That was generally too late for us”, Bhuvan Shetty says. That’s why Bosch’s solution takes not just the sensor data into account, but the data situation as a whole, and identifies the symptoms of impending wear.“ This enables parts to be replaced at the right time—not too soon and especially not too late, but always before something breaks down.” 

If an alarm has been triggered, the Digital Twin sends a message through Microsoft Teams on the console right on the equipment and also to the handheld devices of the caretakers, indicating just what the issue is. Depending on the complexity of the alarm, a ticket is also created through Dynamics 365 and is sent to the proper experts, who can then service the machine. Inspections can be performed remotely or on site. If assistance is needed from experts based at a different location, they can be brought in virtually using HoloLens 2 and Remote Assist. This is how the ideal predictive maintenance succeeds.

The data from the Digital Twin thus directly improves sustainability. After all, parts are replaced only when it’s really necessary. But that’s not all: “When we know the breakdown scenarios and know, for example, when machines tend to overheat, then we also know how we can operate them most efficiently”, Prahallad says. “Efficient machines need less energy and thus consumption of required raw materials gets optimized, hence conserving energy. This is a key area of focus particularly for manufacturing companies.” 

Bosch’s solution thus simplifies complex interrelationships, automates processes, boosts efficiency, and helps save resources. And most importantly, it makes people’s jobs easier. And that’s precisely why the solution won the Microsoft Intelligent Manufacturing Award 2021 in the Add Value! category.