Predictive maintenance – a core component of Industry 4.0
This is often due to unforeseen damage or the premature replacement of parts that are actually still intact in order to prevent machine failure. All in all, these measures incur high costs, only to maintain the status quo in the end? At this point, we should therefore first clarify why technical systems fail in the first place.
The theory of the wear stock
Technical objects, according to the theory of wear stock, have a certain stock of performance they can deliver before failure occurs. Through various maintenance measures, this stock can be increased in order to extend the life of the object.

Nowadays, a plant failure can entail high downtime costs. Therefore, an ideal maintenance strategy should aim at using up the wear stock as much as possible but without causing a failure. This is where predictive maintenance comes in. Before we go into these further, a brief overview of the developments in maintenance strategies in recent years:

What is Predictive Maintenance?
Already in the case of condition-oriented, also preventive, maintenance, methods are sought to carry out a maintenance measure depending on the actual component condition. Predictive maintenance goes one step further: Instead of relying on the actual conditions of a plant, the ideal maintenance times are predicted on the basis of past values, sensor-supported condition monitoring or a combination of both. For example, data such as pressure, humidity, noise level, throughput times, etc. can be recorded. These data are continuously controlled and documented. In this way, an impending failure can be noticed at an early stage.
The new technologies of Industry 4.0 provide support here, as a successful predictive maintenance strategy could not be implemented without the possibility of collecting and evaluating large volumes of data.
What are the benefits of the predictive maintenance approach?
The predictive maintenance approach is one of the key enablers for the smart factories of the future and brings numerous benefits. Among other things, unplanned machine downtimes can be avoided, field service assignments of service staff can be optimized, and maintenance or repairs can be better planned and carried out. In addition, spare parts management can be better planned, downtimes can be reduced and, in the long run, the service life of machines can be extended.
In short, the early detection and elimination of faults can improve the availability of a machine.
But there are challenges and limitations here as well. For example, problems can occur with data analysis. In addition, not all machines can be equipped with the necessary measuring devices. With the help of retrofit measures, however, it is also possible to include older machines.
Vorbeugende vs. vorausschauende Wartung
Preventive maintenance aims to keep the machine in good condition and prevent failures. The basis is the expected time of failure, disregarding actual operating data of a machine. So, in other words, this is planned maintenance. This type of maintenance has always existed, regardless of the Internet of Things and digitalization. Predictive maintenance has emerged over time due to the development of significant ways to obtain and analyze data.
It will be exciting to see how artificial intelligence and machine learning methods can be integrated into this area in the future. Perhaps then, in the future, we will be able to completely avoid unplanned downtime while still fully consuming all resources.