UdZ / Issue 02.23

UdZ 02.23 / 21 CSOT (China Star Optoelectronics Technology), also called Shenzhen Huaxing Photoelectric Technology, is the leading LED supplier of the TCL Group and the second largest producer of LCD displays worldwide. They use Artificial Intelligence (AI) to automate defect inspection processes to maintain their competitive edge. However, the labeling of datasets for training using current AI based methods is time-consuming and labor-intensive, requiring up to 1,500 hours training time in total for one typical manufacturing factory containing multiple production lines. To address this challenge, the Hong Kong Industrial Artificial Intelligence and Robotics Centre (FLAIR) – jointly initiated by RWTH Aachen Campus and Hong Kong Productivity Council – develops new AI based segmentation and classification techniques. With this approach, image labeling is reduced from 20 seconds to less than 1 second and more than 1 ,400 working hours can be saved per factory, providing great potential quality improvement for production management and estimated savings of more than EUR 84,000. > Artificial Intelligence Helps Quality Managers to Reduce Inspection Time and Manual Labor Novel Auto Labeling Method From FLAIR and CCI

RkJQdWJsaXNoZXIy NzcyMw==