en.Wedoany.com Reported - Deeply Inc., an industrial acoustic AI solution developer, announced on the 11th that its "Listen AI" system has achieved a 99.87% accuracy rate in detecting connector engagement sounds on the production lines of a global automaker, validating the field applicability and reliability of acoustic AI-based quality inspection technology. Building on this foundation, the company plans to participate in "Automate 2026," North America's largest robotics and automation exhibition, held from June 22 to 25, 2026, in Chicago, to actively expand into the North American automotive manufacturer and parts supplier market.
As global automakers accelerate their transition to software-defined vehicles and electrification systems, the density of electronic components such as in-vehicle infotainment, ADAS, and high-voltage batteries has significantly increased. During assembly, verifying the perfect engagement of numerous invisible wire harness connectors has become a core task in manufacturing quality. On traditional automotive assembly lines, workers rely solely on hearing or touch to judge the faint "click" sound when connectors are inserted, but this method is susceptible to worker fatigue and yields highly subjective results. Additionally, inspection histories cannot be recorded as data, posing potential risks for future recalls or warranty costs.
Deeply's industrial acoustic AI solution, "Listen AI Industrial," has moved beyond the pilot stage and is ready for immediate deployment across various manufacturing production lines. The system is already operational on mass production lines at global automaker Company H's plants in South Korea and Mexico, achieving a 99.87% accuracy rate in connector engagement detection and meeting the stringent inspection standards of large enterprises. Compared to traditional manual listening methods, this solution demonstrates higher detection precision, clearly identifying connector engagement sounds even amidst the noisy manufacturing environment. Beyond Company H, the system has completed proof-of-concept (PoC) projects at multiple global companies for battery assembly lines, motor defect detection, and robotic automation assembly lines, and is transitioning to production deployment phases.
Noise levels inside automotive manufacturing plants typically exceed 85 decibels, originating from impact wrenches, pneumatic tools, metal friction, and conveyor belts. With over eight years of focus on manufacturing acoustic data, Deeply has built a proprietary foundation model based on a database of over 10 million process event data points and more than 2.1 million hours of real factory noise. By applying noise reduction algorithms that precisely eliminate complex background noise, the system can capture faint "semi-engaged" or soft connection sounds that are difficult for the human ear to detect. Its frequency band analysis technology distinguishes waveform differences between the primary lock (3-16 kHz) and secondary lock (7-16 kHz), clearly identifying the state before a connector is fully locked, fundamentally preventing intermittent electrical signal errors or failures in critical components like airbags after vehicles leave the factory due to incomplete engagement.
Listen AI operates accurately from the first day of hardware installation, requiring no weeks of data collection or training. It completely eliminates the need to modify existing manufacturing line equipment, tools, or layouts, which is another reason global OEMs choose Deeply. The system supports over 20 industrial sensor variants, including fixed directional microphones, wristband microphones, and array microphones for high-noise areas, enabling flexible operation on dynamic production lines where worker movement paths change. Inspection results (OK/NG) can be linked in real-time to factory MES or PLC systems, capitalizing on traceable data.
Deeply CEO Suji Lee stated that connector engagement sound detection was previously considered difficult to fully automate, but the acoustic AI solution has proven that this issue can be resolved on global production lines. With inquiries for collaboration coming from multiple global automakers and parts suppliers, including those in North America, the company will use Automate 2026 as a springboard to fully enter the complex manufacturing environment of the North American market, leading the AI transformation on the factory floor. Deeply's booth is located at #33046 in the North Hall, where it plans to conduct live demonstrations of the Listen AI connector engagement sound diagnosis solution and initiate collaborations with North American automotive OEMs and Tier 1 parts suppliers. The exhibition runs for four days starting June 22.
Acoustic AI solution company Deeply achieves "machine hearing" through its proprietary non-verbal AI sound analysis technology, managing safety in industrial settings and daily life. Its industrial AI sound analysis solution, "Listen AI Industrial," is used to automate product quality inspections that rely on human hearing, distinguishing subtle differences in component sounds to enable noise detection and component engagement sound detection for industrial parts such as actuators, motors, and gears. It is currently being supplied to H Group (an automaker), Hyosung Electric, and Korea Railroad Corporation. The public safety AI sound analysis solution, "Listen AI Safety," ensures safety in visual blind spots such as restrooms, changing rooms, and blind areas by identifying sound signals like screams, loud voices, crying, and calls for help. This solution has been deployed at the Government Complex Sejong Gymnasium, Naejangsan National Park, Incheon Transit Corporation, and Gangwon Land, and has been exported to global markets including the Singapore Ministry of Home Affairs, Thailand, and Vietnam.
This article is compiled by Wedoany. All AI citations must indicate the source as "Wedoany". If there is any infringement or other issues, please notify us promptly, and we will modify or delete it accordingly. Email: news@wedoany.com
