Predictive maintenance is a sensor-based form of maintenance. The process uses big data analyses to predict system failures and malfunctions, enabling maintenance work to be planned more efficiently. As an Industry 4.0 application, predictive maintenance can significantly reduce costs and downtimes compared to conventional maintenance strategies.
1. how does predictive maintenance work?
1.1 Collecting data
The term predictive maintenance can be translated as predictive maintenance or Predictive maintenance be translated. To enable this "looking ahead", relevant data must first be determined by sensors in the machine. To implement a Predictive maintenance-A wide variety of data and therefore also a wide variety of sensors can be used in the application. Examples of this are
- – Electrical sensors
- – Vibration sensors
- – Temperature sensors
- – Pressure sensors
- – Humidity sensors
- – ...and many more
This variety reflects the complexity of the issue. Unfortunately, there is no one-size-fits-all solution for implementing projects. It needs a capable teamwhich creates customized solutions for the area of application. Completely different concepts are required for wind turbines than for machine tools.
This is why companies shy away from them, Predictive maintenance comprehensively. Most market participants are aware of the importance of the field, but have no experience in its implementation.
1.2 Analyzing data
The counterpart to data collection is data analysis. By evaluating the measurements from the running process, problems can be detected at an early stage. Service technicians are then dispatched before a breakdown occurs.
This requires large amounts of data, which can only be analyzed using approaches from the Big Data-environment. Even before the sensors are installed, it should of course be clarified which data indicates an imminent machine failure so that everything runs smoothly during your predictive maintenance implementation.
Implementation of predictive maintenance
As many companies do not know how to get started in the area of predictive maintenance, we have created an 8-week plan for implementing an initial pilot project in this area. More info?
AI and predictive maintenance
Artificial intelligence can also be used to analyze data and maintain systems more effectively. Machine learning offers great opportunities for the future here. More info?
2. advantages of predictive maintenance
In a Roland Berger study 79% of those surveyed stated that they see the benefits of predictive maintenance mainly in a Increased performance of production technology. This is primarily due to the higher Availability and the longer Service life of the systems. The improvement of the Process quality is a decisive factor.
21% percent, on the other hand, expected above all a Reduce your coststhrough lower expenses for spare parts and repairs. Even if some work and investment is required initially, the implementation is definitely worthwhile in the long term. Predictive maintenance has advantages over conventional maintenance strategies:
2.1 Predictive maintenance vs. preventive maintenance
The concept of preventive maintenance is actually familiar to anyone who has ever ridden an elevator. There is usually a TÜV seal with the date of the last inspection next to the button. Specialist personnel check the functionality and safety of the elevator at specified intervals. Through Predictive maintenance technicians in this example could save themselves a few missions.
Comprehensive data analyses can be used to determine which systems require maintenance and which do not yet require an inspection at the specified interval. In comparison with the preventive maintenance can predictively maintained systems therefore increase safety and at the same time avoid costs for unnecessary maintenance work. The running time and service life are also significantly improved.
2.2 Predictive maintenance vs. reactive strategy
The reactive strategy usually takes place in less safety-relevant areas. In fact, this is not really a maintenance strategy. This is because only repairs and analyses are carried out in the event of a malfunction. The disadvantages are obvious.
Running the system without maintenance until it fails can lead to far greater damage in the event of a malfunction. In addition, breakdowns occur more frequently. Supporters of this strategy often forget to factor the lower resale value of the systems into their profit calculations. A poorly maintained checkbook has no less of a negative impact in the industry than in the private car market. Those who save on maintenance often end up paying more.
Predictive maintenance has not yet really arrived in many companies. The ServiceLobby has already explained the extent to which this technology can The future of service will influence the future.
3. examples of predictive maintenance applications
Possible areas of application of Predictive maintenance are diverse. Implementation is also easier in some sectors than in others. The complexity of the systems varies. However, the individual steps of the conversion process are always the same. In some markets, however, the players are more reluctant to adopt new maintenance technologies.
- Wind energy industryVibration analyses provide suitable measurement data to raise the alarm in good time if a wind turbine is no longer running smoothly. This is comparatively easy to accomplish with today's analysis and sensor technology. The industry is therefore already well positioned and can be regarded as a benchmark industry against which others must be measured.
- rail industry: A system developed by Siemens System detects and documents cavities in the ground based on the acceleration behavior of a train. Even a slight subsidence of the track bed causes unevenness in the course of the rails, which means that the power of the drive can no longer be transmitted evenly. Severe subsidence, on the other hand, can become a safety risk. Predictive maintenance helps to better coordinate maintenance work on tracks.
- Automotive industry: In the automotive environment, engine and drive component failures are being predicted more and more accurately.
- Mechanical engineering: In mechanical engineering, we find a wide variety of different system concepts. Predictive maintenance solutions can therefore also take a wide variety of forms. There are almost unlimited opportunities to get involved here. However, a little more work is required in the development and planning of projects, as machines typically combine a large number of different technology concepts. In addition, the quantities are significantly lower than in the automotive sector, for example. This makes the development of dedicated monitoring concepts considerably more expensive, as they have to be allocated to a much smaller number of units.
