半導體測試業於設備供應商之多準則評選

Multiple Criteria Evaluation of Testing Machine Suppliers for Taiwan Semiconductor Industry

吳筱恬(Hsiao-Tien Wu)

指導教授 ： 胡宜中

中原大學/商學院/企業管理研究所/碩士(2017年)

https://doi.org/10.6840/cycu201700233

摘要

台灣半導體產業在政府政策的扶植下持續發展，現在電子產業為台灣最重要的工業。創造就業人口達20萬人，更佔台灣GDP約13%。近年來隨著電子產品不斷研發創新，各應用產品重視裝置與用戶間的互動性及感知溝通能力，正因為如此IC測試設備之地位重要性與日俱增，半導體產業當中，機器設備是創造營收的主要工具。過去多數測試廠對於設備的採購多以價格為主要考量，而設備商為了端出測試廠滿意的價格會簡化設備的規格或配置次等零件以節省成本，造成設備品質不穩等問題發生，不利於高階產品對品質的高度要求。因此單次價格不能做為評估設備供應商的唯一標準，這也是測試業所思考的問題。如果能建立一公正專業之評估方式將有助於產業發展，設備商也可以從中瞭解市場的需求避免被替換，以發展長期合作關係。設備供應商該如何提升競爭力以得到測試廠的青睞，並讓台灣測試產業持續居於世界領先的地位，使雙方穩定成長達到一個雙贏的局面，便成為一個重要的議題。

本研究透過德爾菲法，建構適用於半導體測試業於設備供應商評選之研究架構，運用決策實驗室法與網路分析程序法，針對工程與採購人員之問卷填答結果進行分析，研究結果顯示，影響半導體測試設備供應商評選之關鍵準則包含「技術能力」、「產品品質穩定度」、「產品良率」、「新技術開發能力」、「交期及產能配合度」、「相互間協調」、「商業關係」。其中「產品品質穩定度」、「產品良率」、「交期及產能配合度」、 「相互間協調」、「商業關係」是現階段極需改善的指標。

因果關係顯示，設備供應商欲有效改善關鍵因素之績效以提升競爭優勢，可考慮由「技術能力」或「新技術開發能力」著手，兩者屬傾向影響其他準則的「因」，因此可做為改善所有關鍵準則績效的起始點，但若以實際作業面來看，由於技術能力為新技術開發能力之基礎，故以「技術能力」做為著手改善的起始點較為合理。本研究就此建議供應商應就「產品品質穩定度」、「產品良率」、「新技術開發能力」進行改善，以強化供應商的技術能力。

由於設備供應商評選屬於灰色系統問題，且考量測試廠對不同的設備供應商在品質或技術上可能存在不同的喜好，本研究進一步就三家市場上具代表性的廠商，結合決策實驗室法、網路分析程序法以及灰關聯分析的資料分析模式評估半導體測試業對於三家設備供應商之整體偏好，排名在前的廠商由於大部分的關鍵因素均屬於持續保持，因此較同業更具競爭優勢。

關鍵字

半導體測試設備、供應商評選、多準則決策、決策實驗室法、網路分析程序法、灰關聯分析

並列摘要

Under the government policy supporting, Taiwan semiconductor now plays a leading role in industrial sphere. Taiwan semiconductor creates more than 200 thousand employed populations, and makes the outstanding contribution of growth of national GDP around 13%. In recent years, final testing of semiconductor became more and more important due to the continuous development of electronic products innovation, the application of product emphasis on the interaction between devices and users. Most of the final testing factory evaluates the equipment take the price as the main consideration. The machine supplier will use low level part to make cost down but this will affect machine performance. To establish a fair professional assessment will improve the industry development. The vendors will make internal adjustments base on customer’s feedback to improve competitiveness and avoid being eliminated by the market. Final test factory has been thinking about the problem of the supplier evaluation. The machine supplier plays an important role recently because of the IC package refreshment. How to improve the competitiveness of equipment suppliers to get the trust from the final testing factory and how to make win-win situation is a significant issue.

This study was conducted by Delphi method, to establish the multiple criteria evaluation of test machine for Taiwan semiconductor industry. After access the top management of Taiwan semiconductor testing industry exercising Analytic Network Process (ANP) based on Decision Making Trial and Evaluation Laboratory (DMATEL) to analyze the data of final test machine supplier selection. The study result shows that the key criteria to affect the selection of semiconductor test equipment suppliers include “Technical skills”, “Product quality stability”, “Product yield rate”, “New technology development capability”, “Lead time and capacity matching”, “Coordination” and “Business relationship”. Among them, the indexes needed to improve are “Product quality stability”, “Product yield rate”, “Lead time and capacity matching”, “Coordination” and “Business relationship”

Causal schema showed that we could improve efficiently the above issue from “Technical skills” and “New technology development capability”. Both of them tended to be the reason to influence others. Consider the truly situation, “Technical skills” is the basic of “New technology development capability”, it was quite reasonable to improve “Technical skills” as an initial point. This research suggests that machine supplier must implement the management on “Product quality stability”, “Product yield rate” and “New technology development capability” to strengthen the technical skills.

Since equipment supplier selection is a gray system problem, and considering that final testing industry may have special preference on performance and technology depending on various machine suppliers, this study selected three representative companies in the market, base on “Decision Making Trial and Evaluation Laboratory (DMATEL),” “Analytic Network Process (ANP)” and “Grey Relation Analysis (GRA)” to evaluate the overall preferences of the semiconductor test industry for three equipment suppliers. Analysising the advantages of the leading manufacturer.

並列關鍵字

semiconductor final testing equipment,supplier selection,multiple criteria evaluation, DEMATEL, analytic network process, grey relational analysis.

參考文獻

胡宜中、邱永亮、蔣鵬，2017年，結合決策實驗室法與網路程序分析法評估烘焙師傅於 國際競賽獲獎之關鍵因素。觀光休閒學報， 23卷1期，頁101-127。

陳弘仁，2011年，微機電系統的市場發展現況及未來趨勢，工研院南分院微系統中心。

陳銘崑、周明芳，2016年，IC 設計業供應商評選之研究：2016第19屆科際整合管理研討會，頁234-247。

邵亭芬，2009年，台灣地區半導體後段設備供應商評選決策之研究，國立交通大學，管理學院碩士論文。

葉宗道，2002年，從SOC看國內半導體產業發展，矽導國家計畫。