The researchers calculate what the sensors would otherwise measure using only the engine data that is generated during normal operation. And they are teaching the drive to use this knowledge. Together with project partners, they are currently researching and testing various methods. The aim is to make production more cost-effective and flexible and to permanently monitor machines for faults or wear.

Sensors are omnipresent today. Cars, for example, contain dozens of these tiny artificial sensory organs that warn drivers when something gets too close, the cooling water is too hot or the tank is too empty. But the sensitive mini-sensors can also be defective and fail to do their job - in which case the car can be left on the side of the road. What applies to cars also applies to machines and systems of all kinds - where a faulty sensor can lead to production downtime and loss-making business.

"We are developing an important new category of sensor: the motor itself," says drive technology expert Professor Matthias Nienhaus from Saarland University. "The advantage: the engineers access measurement data that is generated during normal operation anyway. This makes the process very cost-effective because no additional sensors need to be installed. We are researching how we can elegantly extract data from the motor that we can use for control and process monitoring. To this end, we are also working with partners to further develop the design of small drives and build motors in such a way that they provide us with as much information as possible," explains Nienhaus. He specializes in miniature systems and small and micro electromagnetic drives with an output of 0.1 to several hundred watts.

Reading out the motor is sufficient

Just as a doctor draws conclusions about a patient's state of health from blood values, Nienhaus and his team use the engine data to read how the engine is doing. "We determine which engine condition is related to which measured values, which measured value changes and how, when everything is no longer running smoothly," he explains. The signals from normal operation are also particularly informative for the researchers: the more data they know about the engine, the more efficiently they can control it. From the mass of data, the researchers identify the signal patterns that are meaningful for this or that occur in the event of certain changes, such as faults or wear. They develop mathematical models for the various states of the engine as well as the degree of faults and wear.

They use these results to feed a microcontroller, the brain of the system, in which the data is evaluated: If the signals change, the controller can assign them to a specific malfunction and then react accordingly. The motors sensitized in this way can also work together in a network via a network operating system, which opens up new possibilities for maintenance, quality assurance and production: It would be conceivable, for example, for another motor to take over automatically if one fails.

To read the data from the motor, Nienhaus and his team look at how exactly the strength of the magnetic field is distributed in the motor. If, for example, current flows through the coils arranged around rotating permanent magnets in a small drive, a specific electromagnetic field is generated. The researchers record how this magnetic field changes when the motor rotates. Using this data, they can calculate the position of the rotor and draw further conclusions that can be used to control the motor very efficiently or detect faults.

As part of the project "Modular sensor systems for real-time process control and smart condition assessment" (MoSeS-Pro: see background below), in which companies such as Bosch, Festo, Sensitec, Pollmeier, CANWAY and Lenord, Bauer & Co. are involved, Nienhaus is currently testing various methods to determine which are best suited to obtaining data from the motor. The researchers are also determining which speed range provides the best data or which engine is most suitable. The Federal Ministry of Education and Research is funding this project.