從半導體設備工程系統到工程鏈管理系統的建置與能力模型之研究

Abstract

論文題目：從半導體設備工程系統到工程鏈管理系統的建置與能力模型之研究 作者：陳文耀 指導教授：翁崇雄 博士 本研究採用整合的能力成熟度模型（CMMI）的觀念，對先進製程控制（APC）系統、異常偵測與歸類（FDC）系統、設備工程系統（EES）、電子化製造系統（e-Manufacturing）及工程鏈管理系統（ECMS）提出一個共通的能力模型，這個能力模型可作為衡量這些相關系統建置的指標。根據此能力模型本研究又進一步分別定義出這些系統的架構，亦即相關的系統維度（Dimension）及能力階層（Capability Level）。半導體業界可據此衡量其相關系統建置的程度及擬定系統未來的開發、努力方向。 為建立量化的衡量指標，本研究進一步針對FDC系統定義出「FDC失誤」及「FDC蛋形圖」，據此再推導出FDC系統的量化衡量指標。半導體業界可據此量化的指標來衡量其FDC系統所達到的能力階層及績效。本研究亦進一步提出FDC系統的持續改善循環，根據相關的步驟半導體業界可依此一步一步改善其FDC系統的能力階層及績效。 在從設備工程系統（EES）的建置進展到電子化製造系統（e-Manufacturing）的建置，良率管理系統與設備工程系統的整合是其中的關鍵。因為這兩個系統的整合將可使半導體業者從中得以推知良率的變動情形與製程及設備參數變異的關聯性。有沒有這項能力在半導體業界將會是項關鍵性的差異。 根據相關的文獻探討得知，資訊系統的建置必須與公司的經營策略、技術、組織、及管理方向一致，否則將難以成功。因此，半導體業者不論是在建置設備工程系統，或是在建置電子化製造系統，都必須將這些因素調整成一致，否則無法成功。本研究更進一步推導出工程鏈管理系統的建置，因為牽涉到半導體業界跨公司的整合，將使其建置的困難度遠遠超出公司內設備工程系統的建置與電子化製造系統的建置。 在本研究的最後，針對如何從設備工程系統的建置，進展到電子化製造系統的建置，再進一步進展到工程鏈管理系統的建置，提出了「螺旋式」的建置方式的建議。也就是說，半導體業者在建置設備工程系統、電子化製造系統、及工程鏈管理系統時，除了可以依據各別系統的維度，一階一階的分頭推展之外，這三個系統的建置也可以依據「螺旋式」的建置方式「循序並進」。雖然，從設備工程系統的建置，到電子化製造系統的建置，再進展到工程鏈管理系統的建置，是一趟漫漫的旅程，但最終半導體業界將會協力達成，以實現半導體業界的設計協同作業與工程協同作業的需求。

This research, referred to the concept of CMMI, has defined a simple and generic methodology, Capability Model, for the IC makers to measure the progress from APC, FDC, EES, e-Manufacturing to ECMS systems implementation. Based on the Capability Model, this research has further defined the Framework, Dimensions versus Capability Levels, for APC, FDC, EES, e-Manufacturing and ECMS, respectively. These Capability Models and Frameworks will facilitate the industry to derive the future development directions and approaches for APC, FDC, EES, e-Manufacturing and ECMS implementation. In this research, the “FDC Errors” and “FDC system Egg Diagram” were defined to measure FDC system performance and implementation capability on a quantifiable basis. In addition, a FDC system continuous improvement cycle was further derived, which clearly defined the approaches for FDC system performance improvement step-by-step on a quantifiable basis. From EES to e-Manufacturing, the integration of YMS and EES will enable IC makers to discover the know-how of product yield variation versus process and equipment parameters deviation. This capability will be a decisive differentiation among IC makers. According to the related works study, information system implementation should align with company’s strategy, organization, technology and management approaches. Thus, either EES or e-Manufacturing implementation, the IC makers should align with these factors first. Otherwise, these systems implementation won’t succeed thoroughly. Moreover, in this research, we derived why ECMS implementation will be much more difficult than EES and e-manufacturing implementation. That is because ECMS implementation involves with cross companies’ collaboration. Before ECMS can be successfully implemented, the related companies’ strategy, organization, management, technology and information systems should be aligned first. And, it will be a difficult challenge. At the end of this research, a spiral implementation approach was proposed for EES, e-Manufacturing and ECMS implementation, i.e. these systems can be level-by-level and concurrently implemented. It is a long way journey, the Odyssey, for the IC makers to migrate from EES, to e-Manufacturing, then to ECMS. However, following the spiral implementation approach, step-by-step and concurrently, the IC makers will realize the Design collaboration and Engineering collaboration among the industry some day in the future.