服務性零組件之最後訂購時點與數量之研究-以汽車產業為例

Abstract

對許多提供售後服務的產品公司而言，滿足顧客對於零組件的需求是非常重要的。尤其是汽車產業，在車輛產出後，由於無法直接在世界各地提供製造後服務，因此通常會透過當地代理商利用其原先所建立的通路來提供售後服務。由於用以維修及更換的服務性零組件服務期間往往比車輛生產期間還長，在車輛停止生產後，零組件所需供應服務的期間通常還需持續一段時間。但此時多數零組件的生命週期已步入衰退期，零組件供應商為考量經濟生產批量的成本因素下，並不會持續生產至代理商的服務期間結束。此將造成代理商面臨上游供應商即將停產某些服務性零組件的情況，因而必須針對即將停產的服務性零組件進行最後一次訂購以滿足剩餘服務年限內所有的需求。代理商必須在成本控管與服務水準之間取得平衡，決定最佳的最後訂購時點與數量。 因此，本研究希望能夠提供一個有效的決策模式，協助代理商處理最後訂購問題。本研究首先利用負指數迴歸模型，建立一套有效的衰退期零組件需求預測模式，並以此預測模式的需求預測值建構出一個量化決策模型用以決策最後訂購時點與數量。此模型將正常訂購成本、急單成本、缺貨成本及存貨持有成本納入模型目標函數，以總成本最低為目標，並發展特殊求解模式，以求得該模型之最佳解，進而得到最佳的最後訂購時點與數量。最後，即以實際需求資料比較個案T公司和本研究所提出最後訂購方法之優劣，並對個案廠商提供相關建議。以期幫助面臨最後訂購問題的廠商，不僅能以更具成本效益的訂貨方式進行訂貨決策，而且可以有效的滿足顧客在剩餘服務期限的零組件需求。

For many companies which provide after-sales service, to meet customer demand for service parts is very important. Especially in the automotive industry, international automobile companies usually require local agents to provide after-sales service through their channels, because for the upstream firms, it’s difficult to directly provide after-sales service to satisfy individual local market. But the periods for maintenance and replacement of the service parts are usually much longer than vehicle’s production periods, the service parts requirement will need to stock for a certain period of time even after the vehicle manufacturer stops producing the car. The period after the vehicle stops production is called the end-of-life service period. In this period, service parts suppliers will not continue to produce until the end of this period, because scale economies of quantity remains mainly concerned. Therefore, the agents need to decide the best final order timing and quantity to satisfy the service levels and cost restrictions. In light of these concerns, the aim of this research is to provide a useful decision-making model to help the agents to make the right decision about final orders. Firstly, this study will use the negative exponential model to obtain a better demand forecast, and construct a cost model to determine the optimal final order timing and quantity by using the demand forecast. Inside this cost model, the regular order cost, rush order costs, lost sales costs and inventory holding costs are taken into account. Then, A special algorithm is develop for the model to obtain the optimal solution. Finally, the real data form company T were incorporated to compare the decision-making model with current practice of company T. The result shows our model reduces cost efficiently and effectively meets customer needs.