Demand Association Analysis of a Vender Managed Inventory System

[[abstract]] 隨著科技進步，網際網路快速發展，企業藉由資訊技術架構電子商務（E-Commerce）作業平台，有效的連結了供應商和顧客，使得市場預測更為精準，客戶回應時間縮短，提昇了供應鏈管理（SCM）的效益。企業為了提昇營業績效，在降低營運成本的同時，要求反應速度加快、提高作業精確性，則必需要讓相關供應鏈資訊快速分享使能達到，例如：有效消費者回應（ECR）、即時生產作業（JIT）的目的，其中供應商管理存貨（VMI）及協作規劃、預測與補貨（CPFR）即是在SCM經營模式下發展出來的重要支援系統。 汽車及電子產品產業在供應鏈管理的發展相當成熟，這些領導企業熱衷於供應鏈的優點而且陸續極力推展精進下，不乏對於供應鏈管理執行後，對領導企業各方面的績效及評價研究。然而，少有對於供應鏈中的供應商成員，打入供應鏈後對維持經營的優劣勢分析，或者對供應商本身企業的績效影響做深入的探討。 執行供應鏈運作的企業，在推行VMI的典型模式下，隨著末端消費者需求以及存貨狀況的推估，定期或不定期修訂計劃，並提供約一季的用料預估需求給供應商成員。各個供應商配合出貨計劃即時提供適量料源，協助客戶在無預先備足料件下，能正常而不間斷地持續生產活動，進而降低終端產品供貨的前置時間。然而預測隨時改變，供應商每次得到的預估需求總是被修訂，這對預先生產備貨的供應商而言隱藏著高度的不確定性風險。 本研究針對供應商管理存貨方面做一探討，並以個案中的扣件供應商為例，將其預估需求以資料探勘的關聯法則做進一步分析，以分析結果架構出客戶的虛擬用料清單表。期此方法可提供給供應商應用在對於需求變化的了解及比較，在投產備料的同時，能夠進一步掌握不斷修訂的預測，並藉以思考如何降低企業所承擔的存貨及營運風險，得在供應鏈體系中尋求更多的存在利基。[[abstract]]Some firms have successfully improved their supply chain performance by implementing an approach known as Vendor Managed Inventory (VMI) With VMI the vendor specifies delivery quantities sent to customers through the distribution channel using data supported by information technologies for example EDI Vendor Managed Inventory Just in Time (JIT) Distribution and Efficient Consumer Response (ECR) all refer to similar concepts but applied to different industries For example the grocery and apparel industries tend to use ECR whereas the automobile industry tends to use VMI and JIT distribution VMI reduces stock-outs and reduces inventory in the supply chain It can be made to work but the problem is not just one of logistics VMI often encounters resistance from the sales force and distributors At issue are roles and skills trust and power shifts For a VMI system to work the concerns of distributors and sales force must be addressed This thesis addressed a case study about how to transform the sales role into one of marketing In this case study the VMI approaches for instance rolling forecasting of a fastener manufacturer are described A data mining approach for analyzing demand associations from customer rolling forecasting based orders is proposed The discovered demand associations among items were further utilized to build a virtual BOM for supporting the sales to create marketing strategies to reduce risk of uncertainty from rolling forecasts and may get more niches in the supply chain syste