Increasing demands on the availability and efficiency of production facilities lead to a greater need for constant information about their condition. These are obtained by condition monitoring systems (CMS). The sensor-based monitoring of the current status enables the early detection of damage and wear and tear, which is used to optimize maintenance programs. In industrial applications in particular, however, the range of applications and operating environments of the components to be monitored is very large. A clear example are the widely used roller bearings in drive technology. The smallest designs can be found in medical technology, while diameters of several meters are reached in wind turbines and cranes, and the loads to be absorbed, speeds and operating times also vary from a few hundred to a few hundred thousand hours. Clear adjustments to the CMS are therefore necessary for each individual case. First of all, this concerns the instrumentation with various types of sensors, for vibration, ultrasound, temperatures or measurement of impurities in lubricants. It turns out that different measured variables are suitable for generating meaningful damage characteristics. This results in adjustment work on the algorithms used for signal analysis, feature extraction and classification. Particular potential lies in the permanent instrumentation with heterogeneous, networked sensors in the sense of the IoT (Internet of Things) in connection with a multivariate data analysis. In the SKALISENS project, a scalable sensor network is being developed and tested, which has the ability to integrate different sensors and their data streams, while being easily adaptable to different problems, e.g. by adding more sensors.

The Fraunhofer IZFP contributes its expertise in ultrasonic analysis and a flexibly programmable sensor platform, which enables the acquisition of various other measured variables. Fraunhofer IIS-EAS has many years of experience in industrial data analysis using machine learning methods and contributes the DeepInsights data platform. Both institutes are established partners of the industry in questions of condition monitoring and condition diagnosis.


Main benefits for the user and advantages:

• Increasing the accuracy of the data analysis by aggregating heterogeneous measurement data

• Increased robustness of the system, e.g. against temporary failures through mutual validation of sensors

• Rapid adjustment and configuration of the system for applications for which a system based on only one sensor quantity (e.g. vibration) does not provide satisfactory results.

• IoT-based system concept enables integration into IT infrastructures and easy scaling by adding additional sensors.

Main features


  • Data analysis based on heterogeneous datasets


  • Easy addition of different sensors
  • High scalability


  • Rapid adaptation and configuration of the system for different applications

Please contact us!

For the extension of our research project, we are looking for you! You are in search of a suitable scalable sensor network? Do you want to enter the digital age together with us? Then write to us!