晶圓代工廠黃光區機台調校決策模式之構建 -以M公司為例

Abstract

半導體廠的黃光區，同功能的機台，因型號不同，使這些機台可做的加工內容亦有差異。然而某些機台經由調校，可更改其加工內容。當可加工某些產品的機台不足時，是否要調校一些機台以增加產出量是本論文的研究主題。本論文建立一機台調校決策機制，在給定產出產品組合下，計算出黃光區各製程階段之最大產出量，並定義最小產出量之製程階段為瓶頸階段，接著找出瓶頸階段之瓶頸機台，且定義可支援瓶頸機台之機台，本論文構建機台調校數學模式，進行調校決策。經由M公司的案例探討，藉由該機台調校決策模式，可增加黃光區整體最大產出量。

This study investigated the machine adjustment decision in the photo section of a semiconductor fabrication. Under a given product mix and the machine configuration with their functions, we construct a mathematical machine adjustment model to determine whether to adjust some machines functions to support the bottle-neck machines, or to remain their original functions for maximizing outputs. First we define the bottleneck stages and the bottleneck machines. Then with the mathematical machine adjustment model, we can determine which machines should be adjusted to support those bottle-neck machines. Finally we demonstrate our approach with a case provided by M company. We find that the outputs are improved by the machine adjustment decision.