本研究以半導體封裝測試廠黏晶片流程作為研究範圍,並應用六標準差(six sigma)方法進行黏晶片生產良率之改善,主要目的為強化作業中的生產資訊及建立系統辨識能力,以防範因人員作業疏失所造成的產品異常。本研究中導入失效模式與效應分析(FMEA)之評估,整理出問題改善順序,並設立階段性改善計畫,在不影響生產線生產的情況下先是使用PDCA模式作為循環改善基礎,最後再利用六標準差DMAIC的模型建立系統辨識以降低異常發生率,並據此建立系統管制圖及故障處理計畫(OCAP)作為失效重現時的管制流程。經本研究半年的努力後,生產線的異常終於得以降低,如此促進了整體產品品質的提升及生產效率的優化,最終達成客戶的期望目標。
The scope of this study is the semiconductor packaging test plant Die Bond process, also use the method of the Six Sigma to improved yield of Die Bond process, to prevent product malfunctions caused by personnel operations, the main purpose of this study is to strengthen the operation of production information and the establishment of system identification. This study uses the assessment of potential Failure Mode & Effect Analysis (FMEA), not only organize the order of improving the problem, but set up a stage improvement plan. Based on production line production is not affected, first, use the PDCA model to improve the foundation. In order to reduce the incidence of abnormalities. Finally, using the Six Sigma DMAIC model to establish system identification, and establish system control chart and out of control action plan (OCAP) as the control process. After six months of the research efforts, the abnormally production line is reduced, and elevate the overall product quality and production efficiency. Eventually we reach our customers' expectations