Summary: | 碩士 === 中華大學 === 企業管理學系 === 105 === The semiconductor wafer testing process can be seen as a post-step for the entire semiconductor manufacturing supply chain. However, there is no a way to predict the wafer test shipments accurately due to its complex test conditions and production requirements, as well as too many unpredictable intertwined interactions. Also, the production scheduling is the same problem. This causes the factory cost more, and get the revenue it deserves inefficiently. In such circumstances, creating an accurate and rapid prediction mechanism or system program to measure and build an effective scheduling in the production line. It not only can provide wafer testing plant a better use of the machine Rate, high production efficiency, and high customer rate, but also reduce the waiting time of the goods to be tested , and keep the number of the goods to be tested low. Eventually, the company's revenue and capacity utilization opportunity will be enhanced.
In view of this, the single-machine scheduling, the queuing theory which are based on several literatures of many scholars will be used in this paper. Also, the weighting concept will show on this paper. Hope to construct a systematic production model for the semiconductor wafer test company. In this way, the model constructed on this basis will not only simulate the production process of the product, but also provide a rapid and performance evaluation. Thus, it can build our shipment delivery forecast engine. In addition, we also use dispatch method from two points of view, the machine and the product . For each machine, it can find a best batch order to test. For the goods, it can select the most suitable machine for testing. This research is based on using simulation way evaluates dispatch method to create a schedule system.
|