Condition monitoring of cutting processes

Detection of incorrect activation of valves
In a dust extraction system, the incorrect switching of valves is to be detected. Due to the strong room acoustics and the reverberation under the valve cover, it is not possible to locate the running time.
Requirements
- Detection of incorrect activation of 280 valves
- Faulty activation of a valve should be detected by localizing the cleaning noise
- Measurement with 16 microphones on the system
Solution
- Improved machine learning approach enables significant reduction in hardware costs
- From over 70 statistical features per cleaning sound and microphone, the five most meaningful are selected for analysis.
Contact Person
Erwin Schimbäck
Business Area Manager Sensors & CommunicationFurther reference projects
Early fault detection for all injection molding machines
Early fault detection for all injection molding machines Reduction of machine downtimes By detecting anomalies in injection molding machines at…
Fault detection in a dedusting system
Condition monitoring of cutting processes Detection of incorrect activation of valves In a dust extraction system, the incorrect switching of…
Condition monitoring of cutting processes
Condition monitoring of cutting processes Quality assurance in fiber production Lenzing produces a wide variety of fibers from the raw…