Porsche develops artificial intelligence for noise detection

Porsche's digital division has developed an artificial intelligence (AI) based digital assistant called ‘Sounce’ that offers an improved ability to detect noise. This technology can be used during vehicle component tests. It can also help improve the development and production of components in the automotive industry.

The technology uses deep learning methods and can reliably and precisely detect noise, for example, during endurance tests. It can also make error documentation more precise and simplify root cause analysis. Use of this method takes the burden off development engineers, who typically have to be personally present throughout such endurance tests.

Sounce also makes it possible to reliably check the noise development of load-bearing components round the clock and under various conditions. Acoustics testing based on AI increases quality and reduces costs across a wide range of applications. "(The technology) enables us to improve the opportunities for analysis in early component tests," explains Patricia Rennert, Head of Industry Solutions at Porsche Digital.

The technology can be applied to various areas. It is conceivable, for instance, to use it in test situations where a large number of different acoustic signals make analog analysis via the human ear difficult. The application has been designed by Porsche Digital together with the development department of Porsche AG.