Industrial Automation

A monitoring system that can hear production errors

In industrial production, the testing of machines and products by means of acoustic signals still takes a niche role. Researchers at Fraunhofer IDMT have developed a cognitive system that can hear erroneous sounds more objectively than human hearing. The technology was proved in initial practical tests, in which it detected up to 99 percent of the errors.

© MEV Verlag
© Fraunhofer IDMT
The Fraunhofer IDMT offers procedures for the end-of-line inspection of car parts, such as motors for seats, for the sake of automated quality analysis by means of airborne sound measurement.

In industrial production, it is crucial that the machines work and that the product does not have any defects. The production process is therefore continuously monitored by humans, but also by more and more sensors, cameras, software and hardware. In most cases, machine-based automated testing relies on visual or physical criteria. Only people also use their ears naturally: if something sounds unusual, a person switches the machine off for safety. The problem, however, is that everyone perceives noises somewhat differently.

More objective than human hearing

Researchers at the Fraunhofer Institute for Digital Media Technology IDMT would now like to integrate the intelligence of listening into the industrial condition control of machines and automated test systems for products. In developing cognitive systems that use acoustic signals to detect errors exactly, they are combining intelligent acoustic measuring technology and signal analysis, machine learning, as well as data-safe, flexible data storage. Once they have been fed with many data records and trained, cognitive systems can hear more objectively than a human can.

Assigning sounds unequivocally

The scientists identify possible sources of noises and analyze their causes, create a noise model of the environment, and focus their microphones there. From the total signal, the system calculates out background sounds. This is then repeatedly compared with laboratory-pure reference noise. With the help of artificial neural networks, the scientists are gradually developing algorithms that are able to detect noises occurring from errors. The cleaner the acoustic signal is, the better the cognitive system recognizes deviations. The technology is so sensitive that it also displays nuances in error intensity and manages complex tasks. In modern car seats, for example, a large number of individual motors are installed, which the driver can use to adjust his or her seat individually. The design of the motors is not the same, their noises are different, and they are installed in different places. The system devised by the Ilmenau-based researchers has dealt with this challenge with aplomb: in a pilot project with an automotive supplier, it was able to perfectly detect all of the error sources the system was trained for.

Flexible, secure data storage in the cloud

The Fraunhofer researchers are able to ensure the data security of the collected acoustic signals through user authorizations as well as rights and identity management. One example is the decoupling of real and virtual identities in order not to violate user rights when evaluating the data by different persons. Machines and test systems are usually installed in the production line. The researchers store their acoustic data records in a secure cloud.

Last modified: