Towards the end of 2021, PSMC established the "AI Task Team Organization" to strengthen wafer manufacturing by combining innovative technologies such as AI artificial intelligence, big data, expert systems and the Internet of Things, and applied it to fab scheduling, personnel and machine productivity, process and machine control, quality improvement, etc. to build a smart factory environment. The existing production quality system mainly uses FDC/ED card-controlled machine to monitor abnormalities and a measurement machine to ensure product quality, aiming at strengthening these functions by introducing AI artificial intelligence technology related applications. In the production process, we introduced AI technology into the product before, during, and after production to strengthen the verification of the quality and reliability of related wafer production, and at the same time, to comprehensively improve the use of production energy, increase efficiency, reduce carbon emissions, and contribute to sustainable environmental development.

PSMC continues to devote itself to the development of smart factories, aiming to establish the 3Ps of machines and products, namely Prediction, Prevention, and Protection, and flexibly use AI-related technologies to predict in advance, to detect the abnormal production machine and product quality in real time, avoid abnormal product defect rate and reduce the risk of production line loss. Since 2022, through the cooperation of departments, the smart factory has gradually developed from prototype, implemented applications, and expanded its scale to practice goals.

PSMC grows with its employees to achieve innovative production management through automation, data and artificial intelligence. In order to provide customers with flexible and complex Open Foundry Services, we are committed to improving customer delivery standards and planning to introduce Production Artificial Intelligence (ProAI) intelligent production management system in order to grasp the dynamic production capacity of the factory and improve the visibility of shipment forecast for more than one quarter. At the same time, the valuable production experience and parameters of the factory can be incorporated into the ProAI system to build an accumulative platform for the manufacturing experience, and to promote a consistent production and sales coordination and communication platform for the manufacturing, business, and production planning departments to master the production situation, and to have a stable, accurate, and long-term planning mechanism for production delivery and production cycle control. The control ability of production delivery and cycle time can be greatly improved.

ProAI can effectively plan the production capacity of thousands of machines in the factory, provide simulation of the production status of thousands of products required by hundreds of customers, and optimize the production process of millions of moves per day; through ProAI, we can effectively predict the production information of each product, each station, daily output, the number of products in process, and the stacking of machines in the factory, and improve the visibility of production status to more than 90 days. In case of major abnormalities such as earthquakes and power outages, we can quickly provide customers with the results of the affected production lines; we can effectively support timely feedback to customers' supply chains and contribute to their overall supply chain control level.

In 2022, ProAI will apply AI to production management, and provide education and training as well as a user-friendly interface for production planning and manufacturing unit employees to guide them to adapt to the changes in work patterns brought about by intelligent processes and modes, and to make good use of AI to significantly improve productivity and solve production problems in order to create greater work value.