“Do you hear this humming sound? That sounds strange, doesn’t it?” – People working in an industrial plant and operating moving or rotating machinery might be familiar with that situation. The human ear is very sensitive, which is why certain (unusual) sounds and noise prompts us to draw certain conclusions and make certain decisions – such as switching off a motor before it may cause damage, because it doesn’t sound as it should. But there is one problem with relying on human hearing only: everyone hears the same sound differently, and comparing what one person hears to what another person hears is difficult. Particularly if we speak about fully automated production processes, technology based solutions need to be in place in order to ensure flawless production and highest product quality. 

Given the limitations of the human ear and eye, Fraunhofer IDMT’s new business area “Industrial Media Applications” offers quality assessment solutions featuring audiovisual sensors (i.e., cameras and microphones) and intelligent signal processing algorithms. The new quality inspection method to be presented at CeBIT combines analysis of airborne sound with signal analysis and machine learning, allowing the assessment of the quality of products and the condition of machinery on the basis of acoustic pattern recognition. The new method can be applied in various contexts. In production facilities, the method can be used for condition monitoring of machinery. “As soon as irregularities are observed, machine operators are informed and may take preventive action before any damage can occur”, explains Judith Liebetrau, project manager at Fraunhofer IDMT. 

Automatic quality control via air borne sound analysis
Another area of application is quality assessment of components to be integrated with other products. For GÖPEL electronic GmbH, a Thuringian engineering company, Fraunhofer IDMT is currently developing a system for assessing the quality of electrical motors integrated in car seats. At GÖPEL, these and other components to be installed in cars used to be tested with traditional methods measuring structure-borne sound. “If tests need to be made with motors that are already installed inside the seats, this method is problematic, as it is not possible to capture the sound waves directly”, says Judith Liebetrau. Frank Pauli, department manager and in charge of end-of-line testing at GÖPEL, is thrilled when he speaks about the possibilities offered by the new Fraunhofer method: “We are currently discussing with our partners from Fraunhofer how we can integrate measuring instruments combining airborne sound measurement and automated quality assessment in our production process. This would clearly help to optimize end-of-line testing at GÖPEL”.

To allow secure processing, transfer, and storage of measured data, Fraunhofer IDMT offers additional validation and encryption technologies, ensuring that data is not corrupted (intentionally or unintentionally) and that sensitive data may be accessed by authorized people and machines only.

If you are interested in novel ways of quality assessment, please visit us at the Fraunhofer booth B36, hall 6, and get to know Fraunhofer‘s integrated inspection method.