| Summary: | 碩士 === 國立清華大學 === 工業工程與工程管理學系 === 93 === Recently the semiconductor industry is facing dramatic changes owing to global competition and technology reaching post-Moore’s law. The investment of capacity expansion usually takes huge amount, which is over 3 Billion US$ for advanced 300mm fab. Therefore, the completeness and effectiveness of fab capacity planning model will absolutely affect the response speed and decision quality of top management. Many parameters need to be inputted into the capacity planning model during the construction stage, for example, the definition of tool group, tool available time, the efficiency of operator, and ideal throughput for each production recipe. The most critical parameter among them is WPH (Wafer Per Hour) and the recipe WPH is the most concerned index for manufacturing group.
This thesis aims to construct an adequate “Tool WPH Monitor and Management System” for semiconductor wafer fab capacity planning model. We analyze the records of lot’s process from begin to end in manufacturing execution system and the process time of every step collected from the semiconductor equipment communication standard. Then, we propose using the statistical process control method to define the baseline and upper/lower control limit of WPH control chart. This abnormality management system will help the manager of production line to diagnose the abnormal situation to adopt the containment action and remove the root cause. The updated recipe WPH will be feedback to capacity planning model so as to increase its accuracy. We validate this approach in a fab in Taiwan.
|
|---|
