Condition-based maintenance saves costs and time, as spare parts are not replaced on the basis of time intervals or operating hours, but on the basis of actual wear and tear. The basic technology portfolio for condition-based maintenance has hardly changed in recent years. Essentially, maintenance can make use of the following technologies: oil analysis, vibration analysis (simple, primarily based on threshold values), vibration analysis (complex, in the sense of vibration analysis), ultrasound (passive, by recording sound waves in the frequency range above 20 kHz) and infrared/thermography.

Against the background of cost-efficient maintenance, a good cost-benefit ratio is particularly important. The benefit can primarily be derived from the depth of results delivered, while the cost is driven by investment in sensor technology and the ongoing cost of evaluations. If the individual technologies are presented according to these criteria, the following picture emerges:

Ultrasound stands out from other technologies due to its unique (physical) properties, which could make it the technology of choice.

1. Ultrasonic technology will experience the greatest progress through artificial intelligence due to its physical properties.
2. Ultrasound has a very broad field of application that goes far beyond the monitoring of bearings and rotating components.
3. Ultrasound is a technology with a very good cost/benefit ratio and is easy to use.
4. Ultrasound does not require permanent sensors.

The reasons for the excellent cost/benefit ratio of ultrasound lie primarily in the simple and non-invasive measurement option. Systems such as motors, pumps or gearboxes can be measured completely non-invasively during normal operation and areas of application can be covered in which, for example, the common vibration technology fails (slowly rotating components, high vibration environment).

As the introduction of a new technology is always associated with expense, it should also be ensured that the technology in question is future-proof. As digitalization does not stop at maintenance and the use of artificial intelligence and machine learning is becoming increasingly important, a crucial question for the future-proofing of ultrasound is: Why will ultrasound experience the greatest progress in combination with artificial intelligence?

Due to its physical properties, ultrasound can only perform "mode changes" to a very small extent. This means that two motors in close proximity do not negatively influence the measurement. Ultrasound does not "migrate" from the housing of one motor into the air and then penetrate the housing of the next motor.

Factories generate little ultrasonic noise, so the data is very "pure" and free of background noise. Isolated sources of airborne ultrasound (leaks) do not mix with the structure-borne noise generated by the equipment.

Ultrasound has a high information density and a few seconds of measurement time on a system are sufficient for evaluation.

These properties make it possible to learn from data sources from different factories and systems. Similar motors from different factories and of different ages can be combined to form a "virtual" motor. Due to the expected progress through the combination of ultrasound with artificial intelligence, the overview shown at the beginning will also change in favor of ultrasound.