集束型顯示器製程設備屬於多腔體的大型機電整合系統，整體效能的複雜性比生產一個單一製程產品的設備來的不易分析，除了考慮各元件本身的可靠度之外，還必須要知道各元件間的彼此物理特性以及關聯性，況且其造價隨著各世代的大小演進而極其珍貴，各廠商無不希望能在有限的資源下，達到製程設備的最佳產出率，在實際生產產品時，每一段時間每個製程腔體就必須要做潔淨一次，以保持每個鍍膜腔體的膜值以及電性的正確性，但多腔體設備若在鍍膜期間同時做腔體潔淨，則會導致新的待鍍膜玻璃在真空機械手臂上等待，造成產出率的嚴重下降。
本研究主要設計改進一個符合國際顯示器設備標準的製程設備模組控制器，主要目的就是改善集束型顯示器製程設備的生產效能，經由分析客戶實際的玻璃鍍膜數據，找出腔體利用率不佳的瓶頸，藉由重新設計設備系統中的路由器模組，用動態調整玻璃已鍍膜數目的方式，實際驗證於實際設備中，實驗結果顯示增加實際機台的產出率改善度達到53.4%，著實達到良好的改善效果。

A cluster tool of display equipment belong to a large assemble system, the overall effectiveness are not easy to analysis than the production of a single process equipment. In addition to considering the elements reliability between each other, and we also need to know the components of physical meaning and relations. Moreover, its costs become more and more expensive with the size of each generation; the manufacturers are all hoping to achieve the best throughput of the process equipment with limited resources. In the real production of products, every period of time, each process chamber must be done once clean, in order to keep the coating cavity and electric conduction of the film. When the multi-chamber do the chamber clean together, it will cause the new glass to wait on the vacuum robot, causing a serious decline in throughput.
This study was primarily designed to improve the display equipment in accordance with international standards of process controller module. The main purpose is to improve the cluster tool of display equipment efficiency through the analysis of actual customer production data. By the analysis of the bottleneck of chamber body, and re-design the router module of the equipment, to improve by doing the chamber clean recipe with automatic dynamic adjustment of the glass clean number, then verify in real equipment. In the final with the new method, we can reach the really good improvement of 53.4% of throughput.

[1]林嘉彬, “Advanced CE Display PECVD Course” , Applied Material製程講義, 民101年4月25日。
[2]莊書銘, 薄膜液晶顯示器組立廠之生產排程規劃系統之設計, 國立交通大學機械工程研究所碩士論文, 民96年。
[3]鄭隆傑, 300mm 型多腔式製程設備之建模及自動化排程, 國立交通大學機械工程研究所碩士論文, 民90年。
[4]張謙煙, 集束型半導體製程設備控制器之物件導向設計與形式分析, 國立交通大學機械工程研究所碩士論文,民89年。
[5]胡志豪,叢聚式半導體製程設備工作排程與效能分析工具之設計,國立臺灣大學,民87年。
[6]Michael Pinedo (2002) , “The role of scheduling, Scheduling theory, algorithms, and systems”, ISBN0130281387, Prentice Hall, c2002, pp.1-15.
[7]Jeffery Herrmann (1999), Master’s Thesis, Operational models for evaluating the impact of process changes on cluster tool performance.
[8]J. Moyne, R. Telfeyan, A. Hurwitz, and J. Taylor, “A Process-independent
run-to-run controller and its application to chemical-mechanical planarization”, IEEE Transactions on Semiconductor Manufacturing, Vol. 10, pp.194-200, Nov. 1995.
[9]M. Hankinson, T. Vincent, K. B. Irani, P. P. Khargonekar, “Integrated real-time
and run-to-run control of etch depth in reactive ion etching”, IEEE Transactions on
Semiconductor Manufacturing, Vol. 10, pp.121-130, Feb. 1997.
[10]Manh-Quan T. Nguyen and Jeffrey W. Herrmann (2000), Sequencing wafer handler moves to improve the performance of hybrid cluster tools, Dept. of Mechanical Engineering and Institute for Systems Research, University of Maryland.
[11]陳文豪,集束型半導體製程設備之可靠度分析, 國立交通大學機械工程研究所碩士論文,民83年。
[12]AKT PECVD TFT MCC Operations, Version:8，2013/6/30.
[13]陳嘉宏,集束型晶圓輸送膜組之控制器與遠端監控系統實作,國立中山大學機械工程研究所碩士論文,民89 年
[14]Communications Environment HSMS/SECS II for CTMC” SEMI document #2318, July 20. 1994.
[15]AKT ICMCC Operations Manual, Version: 2，2015/5/1.
[16]Takako Takehara. (2005) AKT CVD, PVD systems TPT Timing Diagram, AKT System throughput design meeting.
[17]https://msdn.microsoft.com/zh-tw/library/kx37x362.aspx
[18]Jeffrey Richter, 葛子昂,“Windows應用程式開發經典”,悅之文化股份有限公司,台北市,2009/11/18。
[19]蔡文龍,“Visual C# 2013程式設計經典”,碁峰股份有限公司,台北市,2014/3/3。
[20]MCC Training Handbook, Version: 0.98, Jan. 2004/5/31.