晶圓製造廠緊急訂單承接模式之構建

Abstract

面對市場競爭激烈, 卻無法預期客戶的下單模式而又必須提供最佳服 務以掌握客源的雙重考量之下, 半導體之晶圓製造廠經常需面臨緊急訂單 所衍生出來的規劃管理, 以及決策判斷指標的訂定問題. 本文之研究目的 即在建立一緊急訂單的評估模式, 希望以一套合理的程序來協助承接與否 的決策, 因而提高生產系統在客戶滿意方面的達成程度. 在產能固定, 系統接近滿載, 緊急訂單的插入必犧牲既有訂單的完工時間. 因此, 本文 首先以訂單分類模式, 決定插入訂單所屬的優先等級, 並推估各訂單未完 成迴圈的流動時間, 以及調整後的投料與預計完工時點. 其次根據新的預 計完工時點以及預先設立之一般訂單與緊急訂單限制門檻, 做超出系統限 制容許量的判斷, 以保障系統受干擾程度不致過份擴張. 再歸結至企業賺 取利潤的目標, 應用限制理論, 以有效產出, 存貨與營運費用三者來決定 插單在經濟層面上的表現, 以期評估成效能兼收製造與財務雙方面的管理 效果. 最後, 利用在製品之延遲度與權重值兩項因子,推算系統在規劃週 期所呈現的平均緊迫程度, 將[ 延遲事件 ]做相對且具比較性質的量化. 驗證結果顯示, 本文在實際執行上有其可行性與方便性. 在規劃週期中, 能對緊急插單事件做適切的因應, 配合系統狀態來調整既有訂單之投料與 預計完工時間; 為求插單不致過度擴張既有訂單的生產週期, 由管理者依 據工廠特性, 彈性調整既有訂單之交貨日可被延遲之比例的高低, 藉以緊 縮或放寬系統容許的範圍, 維持預期的生產現場穩定度, 並預知交貨延遲 比例. 此外, 本文具體呈現生產製造與經濟財務間的互動, 並可預知對提 昇利潤有正面作用的緊急訂單索價範圍. 最後, 達到凸顯緊急插單對系統 干擾的程度之目的, 建立承接緊急訂單時完整全面的參考資訊. Because of the difficulty of accurate order forecasting and the need ofimproving customer service level, the management and planning problems andthe reference providing for decision making induced by rush orders usuallybother the wafer fabrication factories. The objective of this research is to designa rush- order acceptance model to increase the customer satisfaction for theproduction system. This research begins with the order classification model to decide thepriority of the inserted order. Since taking rush orders will increase the flowtime of other existing orders under the full-loaded condition, the remaining flowtime for each released order needs to be verified so as to adjust the materialrelease and completion times, the tardiness situations of normal orders are checked to assure that the disturbance from rush orders is allowed. In order to achieve both manufacturing and financial management effect, thetheory of constraint is applied to evaluate the economic performance of the order insertion by the three factors - "throughput", "inventory", and "operationexpense". Also, the system's average tightness during the planning period isestimated according to the lateness and weight of each work- in-process in thequeue line. The average tightness is calculated before and after accepting therush order so as to be compared. This model can be executed easily. With a simulation model, it can re-estimateorders' release and completion time to deal with the changes caused by insertion of rush orders. To achieve the commitment of orders' due dates, the manager can observethe percentage of existing orders which are delayed after accepting rush orders, thenjustify the allowance tardy range, and the stability of the shop floor. Besides, thismodel shows the interaction between production and financial management. It can be usedto estimate the pricing range of rush orders to avoid loosing profit. The information andinfluence about accepting rush orders thus are fully provided.