Existing QA procedures cannot solve all the problems that arise, such as unexpected downtimes, continuous monitoring of the machine or the evaluation of existing machine data. There is also a lack of experienced specialists who know their machines so well that they can recognize errors directly from the noise and react to them. Using acoustic measurement technology and AI algorithms, we mimic the hearing of experienced machine operators and thus provide a reliable, automated system that uses the noises occurring in production to indicate irregularities in machines or processes.
The principle demonstrator shows different-sounding electric motors whose operating noises are recorded by a microphone and analyzed using AI algorithms. Visitors to the trade fair can see the analysis results on a tablet.
The intelligent noise analysis is used, for example, for process monitoring during welding, where it is analyzed during the process to determine whether the weld seam is successful. If the acoustic AI detects deviations or even a specific defect such as pores or blowholes, the process can be stopped and parameters reset. This type of near-real-time quality inspection of weld seams can be used, for example, in the production of battery or fuel cell stacks and contributes to process optimization.
We look forward to seeing you at stand D-01 in hall 4.
Your contacts at the stand:
Mareike Helbig | Dr. Katharina Anding
Fraunhofer IDMT
