/Courtesy of Diplee

Audio artificial intelligence (AI) solution corporations "Deeply" said on the 10th that it supplied the industrial AI sound analysis solution "Listen AI" to Hyosung Electric, a global motor manufacturing corporation.

Deeply is a specialist audio and noise analysis corporation that identifies and manages safety through sounds in industrial sites and daily life using nonverbal AI acoustic analysis technology.

Hyosung Electric is an automobile motor manufacturing corporation and has introduced the Listen AI solution across the entire line for producing automotive electric actuators. An electric actuator is a key component that receives an electrical signal and converts it into actual rotational or linear motion to control various vehicle functions.

With Listen AI, AI learns and analyzes subtle differences in operating sound generated when actuators or motors run to distinguish good products from defective ones.

Whereas conventional factories directly operated large-scale noise inspection lines and had people listen to each part by ear to make determinations, now the system has learned hundreds of thousands of sound data points to improve inspection accuracy.

Listen AI learned 300,000 motor operating sound data points and demonstrated more than 99% validity within one month of mass-production line testing. Because the solution can be deployed on the production line, quality can be inspected as soon as products are made.

A Hyosung Electric official said, "After applying an industrial AI sound analysis solution across the entire actuator production line, we also plan to apply the AI solution to the new plant currently under construction," adding, "As an operating strategy of a Smart Factory that makes the entire manufacturing process intelligent, we will achieve fast production and high completeness."

Lee Su-ji, CEO of Deeply, said, "The Listen AI solution identifies quality differences in parts that cannot be distinguished by eye, improving product safety," and noted, "The industrial AI sound analysis technology will become even more useful going forward."

※ This article has been translated by AI. Share your feedback here.