Taipei, November 7, 2022 / PRNewswire / – The main features of current and future technologies, from smartphones to self-driving cars, medical test equipment to security monitoring, are optical cameras. It is important that the optical components are properly designed and efficient, which can be achieved with the use of artificial intelligence (AI). Most of these components are manufactured in Taiwan. In Taichung, where most manufacturers have a base, annual revenue has exceeded NT $ 100 billion. Despite the established industry, there is the question of whether the optical industry in Taiwan There are challenges as it progresses.
Through its AI software Of Taiwan The Office of Industrial Development (IDB) is set to identify these challenges and to see how AI can help and enhance its competitiveness. Of Taiwan Manufacturer. The project involved visiting and interviewing 20 companies in the Taiwan Optical Association and the participation of experts and consultants to identify problems and provide AI solutions.
The most pressing issues that lead to increased production costs and the difficulty of customization for manufacturers occur during production quality control (PQC). Problems arise from the result of mistakes that come from minor mistakes and the inability to focus clearly. To overcome these challenges, AI engine vision is proposed as the best solution if sufficient data about disability can be collected. The use of the human eye for disability screening, which some companies have done, increases the cost of labor and the ongoing need for change due to the decline of one person’s eyes over time. Recently, it has become more difficult to recruit for such positions, which affects the company’s ability to maintain demand. A successful case from an AI application involves an optical company that uses AI defect detection technology and has a disability detection rate of over 60%. This causes more increase NT $ 18 million In production capacity.
Another case from AI software has improved another company’s control process by switching its imaging approach from geometric optics to differential optics. With differential AI, the image reproduces with high resolution and high precision to create clearer images where flaws can be easily detected.
After helping 10 companies implement AI technology in their production lines, AI has introduced the following to other companies interested in doing the same:
Improve automation technology first, then collect big data about the experience.
Build infrastructure for machine vision with adequate optical sensors, integrated with the regulation and operation of mechanical control parameters.
Improving the digital environment to connect the entire production line.
Expand intelligent manufacturing from AI control for PQC to all production lines for the entire optical industry.
AI software believes that a broader share of the blueprint of how AI is being effectively adopted in manufacturing can serve as a useful and productive tool for technology promotion.
Source: Chung-Hua Institute for Economic Research