KUKA manages its robot fleets in real time with a software platform based on Azure IoT

Industrial robots work quickly, precisely, and almost around the clock in many fields, not just the automotive industry or electronics manufacturing. For their robot systems, automation specialist KUKA and its subsidiary Device Insight have developed the iiQoT software platform based on Microsoft Azure IoT. The platform provides important condition data in real time and enables remote monitoring of entire robot fleets. This allows efficient control of facility management, including fault monitoring and maintenance. An internal KUKA lighthouse project in the German city of Augsburg has shown that in one year of production, the platform saved 25 hours of machine downtime and administrative work.

The challenge: Achieve smooth processes for optimum output

They maneuver heavy parts into position with millimeter precision and can manipulate the smallest components with lightning speed and accuracy. They weld perfect seams, paint large surfaces evenly, and stack pallets smoothly and efficiently: KUKA robots can be found in the factories of the world’s top manufacturing companies. “Industrial robots are used in production around the clock and are an absolutely crucial component given their ability to precisely repeat movements. Our customers put very high demands on the accuracy, quality, and available uptime of our products,” says Julian Krebs, IIoT Business Development Manager at KUKA AG. The latter depends heavily on the stress they are subjected to as well as the production environment or industry. “In the automotive industry, for instance, robots are more output-driven because they need to produce consistent parts over a period of several years. In other industries, their tasks change more frequently,” explains Thomas Stammeier, CTO and Managing Director of KUKA subsidiary Device Insight. “Maximizing throughput without compromising on quality is the first goal of any optimization effort.”

“Industrial robots are high-value products that as a rule break down very rarely, but they have gears and motors and so they sometimes require maintenance,” Stammeier says. Krebs adds: “Unplanned downtime has to be avoided no matter what. Repairs will of course be necessary, but should happen only at the scheduled maintenance times.” However, disruptions in operations are seldom caused by wear and tear, but rather when changes are made to established processes. As Krebs explains, “Actively stepping in to make changes is necessary in practice. When points along the production line are changed or machines are swapped out, this may push a robot off its optimum course.” Through signal monitoring, it is possible to detect such cases and report the disruption. The robot comes to a complete stop and waits for a specialist.

The solution: iiQoT platform connects robot fleets to the cloud

“One advantage of digital transformation is being able to predict the occurrence of a fault based on data, or at least be able to detect and fix it efficiently and quickly,” Krebs says. Stammeier adds: “There’s also a better overview when a company uses a lot of robots.”

“It wasn’t too long ago that employees were using clipboards to write down which robots were where and what software they were running, and then typed that information into Excel spreadsheets. With the iiQoT platform based on Microsoft Azure IoT, customers get an immediate overview of their robot fleet, can troubleshoot faster, and make better forecasts.”

Thomas Stammeier, CTO and Managing Director, Device Insight GmbH

KUKA’s integrated platform uses the industrial internet of things (IIoT) to collect and analyze data from robot fleets. As Stammeier explains, “For example, it collects condition data, such as what the robot is currently doing and how long it’s been running, or system data like what software is installed and in what version.” He continues: “There’s also the option to receive fault notifications directly or to set up condition monitoring, such as when you want to keep an eye on the control system’s processor load.” This data can be displayed graphically in iiQoT on a dashboard organized by category.

IIoT specialist Device Insight, which KUKA acquired in 2018, played a key role in developing the platform. “We’ve been looking into IIoT topics since the founding of our company in 2003. In addition to our own in-house products and components, we rely on services from cloud providers. First and foremost among them is Microsoft, of which we are a partner,” Stammeier says. As a result, iiQoT uses products and services from the cloud environment. “A cloud solution based on Azure makes it easier for us to distribute and maintain software and integrate it into the KUKA ecosystem,” he says. “In addition, our solution is very close to what you would call a Microsoft Azure IoT reference architecture.” Because they are part of the production network, the robots are not directly connected to the internet for security reasons. Instead, a module serves as a connector on-site to collect the data and transfer it to the cloud via a secure channel.

On his laptop, Krebs calls up the overview of the 160 robots owned by KUKA subsidiaries. He selects “Operating Mode” across the fleet to show which modes the robots are currently running in. “Now I can dive deeper and deeper into the asset,” he says as he refines the view with the search criteria in the drop-down menu—moving from an entire production site to a factory floor to a cell of five robots to a single robot. Other parameters, like, a robot’s program speed, can also be selected. “For example, if I’m managing a shift, I can look at the history for the week and see if and when a particular robot has been running below 75 percent program speed,” he says.

“We want to give our customers the ability to have their experts draw from a breadth of data that they didn’t have before,” Krebs explains. Stammeier adds: “Customers also have the option of contacting KUKA support much more quickly and directly.” The time advantage when a fault occurs is obvious.

“In the past, an employee would have to go to the factory floor with a USB stick, read out the diagnostic file, carry it back to their computer, and send it to KUKA as an e-mail attachment. With iiQoT and Microsoft Azure, this can be done in just three clicks. That’s because not only is the robot integrated into the system, but the KUKA Support ticket system is as well.”

Julian Krebs, IIoT Business Development Manager, KUKA AG

KUKA and its customers learn for the future

The platform went live in October 2021 and is connected to more than 12,000 robots in various industries. One customer from the automotive industry connected several thousand robots to the platform all at once and can view all the data at each location. This shows the scaling advantage of a cloud solution. Initial figures on the benefits of the platform in production operations was provided by an internal KUKA lighthouse project at the company’s own manufacturing facilities in Augsburg. “We were able to save 25 hours of machine downtime and administrative work in one year of production while still achieving the same end result,” Krebs says.

Many of the pilot projects underway with customers promise to deliver important data. How can maintenance intervals be adapted to production running at full capacity? What influence do environmental conditions such as the temperature in the plant have? What anomalies can be detected in the robots’ behavior and what do they tell us? If, for example, a robot consumes noticeably more energy for the same movement, what conclusions can be drawn from that about the service life of the gears? “We don’t have any ready-made solutions for such forecasts in the platform’s functional scope, at least not yet,” Stammeier says. But KUKA supports its customers in related use cases. “When we as a manufacturer provide our customers with such a solution, we are the best possible source because we know our robots. At the same time, we can further develop our products, as well as the algorithms and thus the validity of forecasts. Everyone wins,” Krebs says.

“When we as a manufacturer provide our customers with such a solution, we are the best possible source because we know our robots. But the insights we gain using Azure IoT also enable us to further develop our products, as well as the algorithms and thus the validity of forecasts.”

Julian Krebs, IIoT Business Development Manager, KUKA AG