在半導體產業中，資本支出相較於一般傳統製造業者及科技業者都較為龐大，而 其產品生命週期短暫、技術汰換率過高等高波動率的產業特性，也使得各大廠商不得 不被客戶要求要能夠以低成本、技術創新及能夠快速的進入市場。 本研究以應用 X 因子理論(X-Factor theory)及利特爾法則(Little′s law)找出半導體晶 圓廠中產能瓶頸區域的研究方法，將半導體晶圓廠預設十個加工工作站，假設當半導 體市場需求資訊給定的情況下，如何利用最具效率的方法找出晶圓廠區中的瓶頸工作 區域需要進行優化改善，以利在最短的時間內有效地改善在製品存貨量過多、製造時 間過於攏長、製造成本過高等問題，使生產及資本的效益最佳化。 本研究之研究模型主要利用西門子(Siemens)公司所提供的 Tecnomatix Plant Simulation 13 進行廠區模擬產生出相關的生產績效資料，根據其起始解得出的數值， 利用 SPSS 軟體進行迴歸分析找出其中影響廠區運作效率的關鍵因子，藉以初步判定出 影響廠區整體效益的關鍵瓶頸區域落於何處，最後再以驗證解與起始解比較，以驗證 此解為可行解，以利爾後找出最適之產能配置。;The semi-conductor industry has a capital expenditure relatively higher than that of traditional manufacturers and technology industry. With industry characteristics of short product life cycles and high technical replacement rate, major companies are forced by clients to quickly enter the market with low cost and innovative techniques. This study applies the research methods of X-Factory theory and Little’s law to find out the bottleneck area of production capacity in a semi-conductor company. With a default of ten processing work station in the semi-conductor company, if the information of demand in the semi-conductor market is fixed, how to adopt the most efficient method to find out that the bottleneck area of the semi-conductor company is in need of improvement, so as to effectively solve problems such as excessive inventory, long production time, and high production cost in the shortest possible time, therefore, maximizing the benefits of production and capital. This research model of this study is mainly the use of Tecnomatix Plant Simulation 13 provided by Siemen to simulate relevant production performance data of the company. Based on initial numbers and by using the SPSS software for regression analysis, key factors in affecting the operating efficiency of the company are found. Initial determination is made to understand where the key bottleneck area that affects the overall benefit of the company is, and eventually make comparisons between the verified solution and starting solution so as to verify that such solution is feasible, which in turn helps find the best production capacity distribution in the future.
