© istockphoto.com/NatalyaBurova

Predictive Maintenance – A Game Changer empowered by Industry 4.0

Steinbeis experts develop decision-making tool for analyzing the economic efficiency of predictive maintenance

The tools of smart factories (Industry 4.0) offer companies a number of ways to improve competitiveness and, at the same time, secure their future viability. The interoperability of production systems can be improved by introducing state-of-the-art information and communication technology, while sensor technology, cloud services, and artificial intelligence foster innovative production and customer processes. The potential areas of application include operational maintenance, a key element of which is predictive maintenance. The question is: When does it pay to invest in such tools, especially for SMEs? This was the question examined by the Saarbrücken Institute for Controlling Innovations (saar#cinnovaton, a Steinbeis Innovation Center).

Predictive maintenance advances condition-based maintenance, which determines maintenance requirements by the physical condition of objects. This condition is inspected at regular intervals by machine operators or checked as part of continual monitoring processes. Predictive maintenance extends this principle by predicting degradation rates based on an independent assessment of machine data. This intends an optimal timing and selection of maintenance actions.

The challenges faced in successfully introducing predictive maintenance include identifying the right sensor technology to monitor equipment condition and gather data, structuring acquired data, and using suitable methods to analyze data. It is important to capture sensitive data on machine condition and machine usage, and this information needs to be monitored in real time. For example, to detect deviating machine behaviour, measurements need to be taken of ambient conditions (temperature, air pressure, or relative humidity), speed ranges, sound, the temperatures of individual components, or vibrations. This allows for early predictions of potential maintenance requirements.

A number of innovative analysis methods can be used to evaluate gathered data, especially artificial intelligence. In addition to traditional predictive models, which apply the laws of physics to determine degradation behavior, methods based on algorithms, such as machine learning, can be used to detect patterns in data sets and predict required maintenance. Such algorithms have to be trained using historical data, which requires detailed documentation on, for example, the timing and type of maintenance carried out. Using artificial intelligence offers a variety of advantages, e. g. a holistic supervision of machine behavior, an early detection of potential failures or the prediction of optimal maintenance times. 

Using predictive maintenance to cut costs

For many companies, maintenance is a key cost driver. Implementing a predictive maintenance strategy can significantly reduce costs, ensure machines and systems are kept fully functional and avoid unplanned machine downtimes. Predictive maintenance is typically much more cost-effective compared to preventive maintenance, i.e. regularly replacing wear parts, since the service life of machinery and plant is extended optimally utilizing the components’ potential of degradation. Furthermore, downtimes can be minimized by adapting maintenance whenever needed to pending orders.

EWIK – a useful decision-making tool for SMEs

SMEs are increasingly forced to invest in smart factory technology (Industry 4.0), and adopting a predictive maintenance strategy promises a variety of benefits. However, SMEs often feel overwhelmed by decisions regarding investments in the technology required to enable predictive maintenance.

There are a number of reasons for this, including the benefits are often difficult to quantify, or the fact that SMEs generally lack sufficient planning resources. This often results in SMEs failing to make the necessary investments in predictive maintenance solutions, which consequently jeopardizes competitiveness. It would therefore be highly beneficial to SMEs to have a tool that makes it easier to assess the benefits of an investment in Industry 4.0 solutions, also so they can receive recommendations on which measures are worthwhile taking under which circumstances.

This is precisely the goal of the EWIK research initiative (EWIK is a German acronym for “decision-making tool for analysing the economic viability of investments in predictive maintenance at SMEs in Saarland”). Due to its cross-company significance, the project is funded by the Saarland Ministry for Economic Affairs, Labour, Energy and Transport and is being implemented by an interdisciplinary team of economic and engineering experts at the Saarbrücken Institute for Controlling Innovations (saar#cinnovaton, a Steinbeis Innovation Center).

The project team is currently conducting an empirical evaluation of the maintenance challenges faced by SMEs. It is also working alongside SMEs to come up with different approaches for assessing investments in predictive maintenance instruments. This will not only make it possible to understand the current status when it comes to introducing predictive, sensor-based maintenance, it will also help identify the need for decision-making support surrounding investment measures and pinpoint potentially conflicting goals. Whether a specific application involves processing machines or conveyor vehicles – just to name two examples – the underlying decision-making issues are similar. As a result, a tool that is standardized, yet individually adaptable when it comes to defining parameters, can be extremely useful. Working alongside SMEs also ensures that profitability evaluations are sound in methodological terms and user-friendly.


Prof. Dr. Alexander Baumeister (author)
Steinbeis Entrepreneur
Steinbeis Innovation Center Saarbrücken Institute for Controlling Innovations (Saarbrücken)

Andreas Nagel (author)
Project Assistant
Steinbeis Innovation Center Saarbrücken Institute for Controlling Innovations (Saarbrücken)