I had a nice talk with Dirk Metzler of Bosch Rexroth at the Hannover Messe press preview this month, and we discussed the company’s exhibition plans for the upcoming international trade fair.
Metzler said the company was touting its work with a customer, Kraiburg, in the rubber industry. Bosch Rexroth does predictive maintenance for Kraiburg, because it’s a critical need for the company.
“If you have machines that work all day long without any stopping, it’s better to predict downtimes, when you’ll have to replace some components. So we offer our service, ODiN, the Online Diagnostic Network,” said Metzler. “We have sensors in our components that measure temperature, vibrations, and so on. Previously, you would get some minimum and maximum values. Now these values are combined. Even if you exceed the maximum value, if it’s only one sensor, then you don’t get an alarm. But if you have several sensors faulting over the maximum, you get an alarm that you have to replace something.”
Knowing that there’s something wrong, he explained, allows the customer to replace the questionable components during the next scheduled downtime. Metzler said that when first setting up a new machine or system, Bosch Rexroth measures a health index of the machine, and this health index is then constantly monitored.
“We work with the customer to set up the initial health index. The ODiN network is intelligent, so the machine is also learning—all day long, during the complete lifetime of the machine, what’s the right decision as to whether to set an alarm or not,” he said.
What’s more, Metzler said, is that this is true not only for one machine. If a company has a whole network of machines in one plant, the learning from one machine can also influence the other machines’ operation.
“You can get the ODiN system running on your own cloud, or on your own servers at your company, or you can also use the Bosch cloud,” he said.
The system, which has been around for a year, is an exciting theme for Bosch Rexroth, said Metzler, because they are now able to show not just what’s possible, but how it’s actually working with real world customers