Fuzzy Available-to-promise modeling of Make-To-Order manufacturing with material supply uncertainty

[[abstract]]在台灣經濟體系中，高科技產業扮演一個重要的腳色。而競爭日益激烈的今 天，使得製造商面臨到更嚴苛的客戶服務需求，如：到期日允諾與更短的訂單前 置時間。而可允諾量(Available-to-promise; ATP)則能幫助製造商快速地回應客 戶端的訂單需求，及協助製造商並確保準時達交(On-Time Delivery)。然而，在 執行ATP 的同時，有很多的不確定的因素可能導致對顧客允諾的交期與數量造成 影響。例如：不確定地物料供給數量可能導致過高的庫存或生產過程的中斷。少 數的研究亦探討在物料供給不確定性下對ATP 系統造成的影響結果，在本研究中， 透過模糊理論(Fuzzy Theory)建模來處理物料供給的不確定性，並建立一個應用 在接單式生產(Make-To-Order; MTO)環境中具有多廠區的TFT-LCD 產業，其製造 商在產能與物料限制下以求取最大化利潤為目標的滾動式模糊ATP 數學規劃模型 (Fuzzy Mathematic Programming)。這模糊ATP 模型的機制主要包含有兩大階段： 第一階段，將顧客訂單最適地配置到各廠區生產。第二階段，當第一步驟所配置 的結果，因可用產能及物料限制無法滿足客戶到期日要求，而需再搜尋所有廠區 中可用之產能及物料來重新配置需求訂單。結果發現，對製造商而言，當實際地 物料供給量低於預估物料供給量，此預估物料供給量是製造商在允諾客戶訂單時 所評估能承擔的風險機會，此模糊ATP 能降低已允諾確未能準時達交的機會。此 外，製造商亦能透過模糊物料供給量來給予顧客一個模糊交期日範圍，此模糊交 期則能夠幫助顧客設計其市場的配送計畫。[[abstract]]High-tech industries have played a significant role in Taiwan’s economy. As the competition becomes even more intense than before, manufacturers are facing strict customer service requirements on due date promising and shorter lead time. Available-to-promise (ATP) can help manufacturers respond quickly requests from customers, and assist manufacturers in assuring on time delivery. However, in implementing ATP, many uncertain factors could jeopardize the quality of promised delivery time and quantity to customers. For example, uncertain material supply could result in either high inventory level or holding up of production. Few research stated the issue of uncertain part supply affecting outcomes from ATP systems. We implemented fuzzy modeling to tackle uncertain part supply, and modeled the proposed fuzzy-ATP as a rolling fuzzy programming problem to maximize profits of manufacturers with capacity and material constraints under multi-site and Make-To-Order (MTO) TFT-LCD industry environment. The fuzzy-ATP mechanism includes mainly two stages: the first is to optimize customer orders allocated to multi-site plants, and the second is to re-allocate orders that cannot be allocated in the first stage by searching among all plants with available capacity and material to meet customer due dates. Results found that for manufacturers, fuzzy-ATP can reduce the chance of undelivered promised customer orders when actual material supply is below estimated material supply that the manufacturers can evaluate their own chance of risk in promising customer orders. Besides, manufacturers can provide customers with a fuzzy delivery date based on fuzzy material supply, which can help customers in devising their delivery plans to markets