As industry 4.0 is developing, so does the demand for a reliable maintenance. Unscheduled breakdowns increase operating costs due to repairs and production losses. But scheduled maintenance implies taking the risk of replacing expensive parts that are still fully operational, while neglecting other parts in spite of their failure. Condition monitoring systems are the solution to optimize your maintenance plan and save costs. However, there are still major drawbacks to their use:
- Undetected breakdowns
- Lack of expertise in signal processing
- Time-consuming analysis
- System-dependent models
- Difficult to use Furthermore, data volume is exponentially increasing due to the development of IIOT systems, but its automatic analysis remains a challenge.
Technology
The software is created thanks to thirty years of expertise in signal processing applied to preventive maintenance. Relying on an expert-level automatic signal processing, it is able to detect on each signal every frequency representing of each part health. By tracking the evolution of features associated to each frequency, the software detects every abnormal evolution thanks to machine learning algorithms based only on the current dataset. Operating on a server, it can monitor each part of a plant remotely and analyze large datasets. Its user-friendly interface ensures comprehensive diagnostics to operators and enhances the efficiency of analysts by providing every calculation detail.
Advantages
- Expert-level signal processing
- Automatic analysis of large datasets
- Remote online diagnosis
- Reduction of false alarms
- User-friendly interface
- No need for historical datasets
State of progress
Validated on a wind turbine, test benches and paper manufacture, we are ready to be used on any plants requiring preventive monitoring in order to test its performances. A machine learning solution is soon to be tested for an automatization of the diagnostic.
Applications
- Energy production
- Manufacturing
- Oil & gas
- Transport
- Industrial machinery