六標準差結合TRIZ進行製程良率提升之研究

Study on the process improvement by Six-Sigma and TRIZ

賴偉銘(Wei-Ming Lai)

指導教授 ： 楊康宏

中原大學/電機資訊學院/工業與系統工程研究所/碩士(2016年)

https://doi.org/10.6840/cycu201600283

摘要

愛因斯坦說「我從不去想未來，它來得夠快。」在半導體封裝技術是隨著時代不斷的在進步，隨之而來的也是一連串挑戰。當客戶的要求不再只是高品質的產品而已(高的散熱功能、電磁波雜訊干擾、重量、環保廢棄物回收等需求也日益提高)不再只是輕、薄、短、小而已，更低的生產成本已經是半導體封裝科技業的基本要件。由於市場競爭者眾，只能以較小的毛利換取較大的數量，接受低獲利的現實，因此兼顧高品質及高獲利，不是非常困難就是不符實際。顯然未來的半導體封裝科技業將是一場殘酷的存亡賽局。
產品於生產製造中常常會發生很多生產異常狀況，這些異常狀況產生了大量的不良品，為了解決這些不良品進而提高產品良率，現今已不能再依靠過去的方式花更多成本去購買更好的機台或者是花更多的資源去提升技術能力，來解決生產產品時所產生的不良品，因為這只會造成生產成本不斷的提升，降低了整個公司的競爭力，尤其是對於已經成熟的半導體封裝業來說，更是一種沉重的負擔。
本研究針對這個議題去研究及探討，應用品質管理中六標準差(6 sigma)方法結合系統化創新(TRIZ)的概念，以DMAIC模式為架構，針對個案公司生產製程上所產生的產品異常所做的個案研究，以客戶需求為出發點及公司的生產成本為考量，於各階段中應用六標準差改善工具及統計品管手法，找出問題並進行分析，最後結合系統化創新(TRIZ)的方法，提供問題的多樣解答，以期能做到花費最少成本，而能同樣的得到最大改善流程方案。
經過個案研究探討及實際驗證，順利將生產所產生的異常降低，提高生產的良率，該生產異常由原本的3.84% 改善到0.02%, 符合半導體客戶需求，並且同時提供多樣解答，從多樣解答中選擇不須要花費高成本也能有相同改善結果。驗證了六標準差的品質手法結合TRIZ的成效，本個案亦可作為相關產業問題解決方案時的參考。

關鍵字

6 sigma 、TRIZ、Quality management

並列摘要

“I never think of the future – it comes soon enough”, said Albert Einstein. As the semiconductor packaging technology advances, a series of challenges arise. When customer requirements no longer just focus on the high quality of products (requirements on heat dissipation, electromagnetic interference, weight, environmental waste recycle, etc become increasingly higher) or the light, thin, short and compact design only, a low production cost is one of the basic elements for the semiconductor packaging industry. Since the market competition is severe, manufacturers have no other choice except accepting the fact of a low profit and using a smaller margin in exchange for a larger sale volume. Taking both high quality and profit into account simultaneously becomes very difficult, if not unrealistic. Obviously , the semiconductor packaging industry will be a brutal survival game in the future.
During manufacture and production, different abnormal situations of products occur frequently, and such abnormal situations produce a large quantity of defectives. To overcome the issue of the defects and improve the yield rate, manufacturers will not rely on the past method of spending more cost to buy better machines or using more resources to upgrade technical skills anymore, because such method constantly increases the production cost and lowers the competitiveness of the whole company. Especially for the mature semiconductor packaging industry, high production cost and low competitiveness are just heavy burdens.
This study focuses on the aforementioned subject and applies the 6 sigma method of quality management accompanied with the concept of systematic innovation (TRIZ) and uses the DMAIC model as the architecture to conduct case studies on product abnormalities occurred during the manufacturing process of different case companies. Customer requirement and production cost are taken into consideration, and the 6 sigma improvement tool and statistical quality control measures combined with the systematic innovation (TRIZ) method are used in different stages, and diversified solutions for problems are provided in hope of obtaining the best improvement solution and process by using the minimal cost.
Through the exploration of case studies and practical verifications, the abnormality occurred in product is lowered and the yield rate is improved, and the abnormality of production is improved from the original 3.84% to 0.02%, which in compliance with the semiconductor customer requirements and provides different solutions. The same improvement result can be obtained by selection one of the solutions without spending a high cost. The effects of the 6-sigma quality method and the TRIZ are verified, and this case study may be used as a reference or a solution for related industry.

並列關鍵字

6 sigma, TRIZ, Quality management

參考文獻

43. 馮勳烈(2001)”六標準差應用於傳統樹脂製造業之品質改善研究＂。頁 1-53。

19. 羅芳琪(2014)“以六標準差結合IDEF0分析方法進行流程改善之研究－以個案的稽核流程為例”頁1-47。

30. 邱創鈞、莊禮帆 (2009)“運用田口方法於迴轉成型塑膠發泡製程最佳化設計”頁4-11。

26. 林清池(2011)“失效模式與效應分析應用於營舍防火安全之研究析”頁9-48。

27. 楊錦洲 (2011)“服務品質”頁197-215。