台灣便利配配送流程之脆弱度分析

Abstract

網路科技日新月異，電子商務已普及於人們的生活中。消費者也逐漸改變了購物型態，改採線上購物的方式。雖然電子商務的科技發展愈來愈快速、安全，然而若沒有實體物流配送，仍無法將貨品送至顧客手中。由於實體配送與消費者的購物習性有關，消費者重視配送的效率、取貨的安全以及可得性，因此對於物流資訊化的要求也愈來愈高。因此，電子商務與物流的發展可說是相輔相成，缺一不可。 由於台灣便利商店高度聚集、24小時營運以及低成本的特性，減少了時空的限制，使電子商務可提供更多種便利的取貨服務，衍伸出一種新的店到店配送的服務。「線上購物，超商取貨」的方式已成為台灣電子商務中即受矚目的物流暨金流運作方式。台灣便利配(ezShip)即為其一。此服務是於2005年開始由全家、萊爾富以及OK等便利商店合資的便利達康(CVS.com)所提供的跨企業店配物流。由於其對收件商品的要求，主要客群著眼於消費者對消費者之間的貨物配送。由於便利達康承諾顧客「今天寄，後天早上到」，因此在這極短的時間內，複雜的流程中的任一環節發生問題，皆有可能導致貨物無法準時送達，以致失去顧客的信賴。 現今各行各業皆非常重視風險管理的議題，尤其是脆弱度的概念更是廣泛地引起各領域研究和探討。脆弱度是表示一災害發生之後，對於一系統所造成的影響，若影響大，即代表脆弱度高，反之則低。透過脆弱度的了解，有關單位可增加資源的投入在流程中最為脆弱的地方，將資源做最有效的配置。本研究主要目的在了解台灣便利配之配送流程，並欲利用故障樹轉換貝氏網路的方式探究不同實體及介面的結合中易發生脆弱的地方。首先定義最不希望發生之事項為「顧客無法在第三天拿到貨」，透過專家訪談獲取系統資訊，建構故障樹，並藉由專家經驗給定基本事件失誤率。接者，再轉化成貝氏網路，計算頂端事件失誤率，並利用機率推論探討流程中最脆弱的節點。 根據推論結果發現，「技術人員操作失誤」、「POS系統異常」、「物流資訊面排程問題」、「系統故障或異常」、「暫時閉店」為系統內最脆弱之部分。雖然這些失誤發生率較低，但對不希望發生之事件敏感度卻很高，因此重要性不容忽視。而「技術人員未按時更新」、「物流中心人員交接班問題」、「店員發生失誤」、「店內人員交接班問題」等也是系統內最為脆弱之部分。這些失誤不僅有較高之發生率，且對不希望發生之事件敏感度也高。由此分析結果，可使提供管理者重新思考決策方向及分配重要之資源予最脆弱之環節及有效降低高失誤頻率之事件發生次數。

Because of great development in electronic commerce, online shopping has become more and more popular. Many marketing experts believe that websites are the most important retail channel. Online trade is not only business to customer (B2C), but also customer to customer (C2C). That is, people can buy or sell their own property through e-commerce systems like e-auctions. No matter which kind of trade is taking place, strong support from an efficient logistics system is necessary to send goods to customers efficiently and safely. Convenience stores in Taiwan have integrated e-commerce systems into their logistics systems to develop a new retail delivery model in which consumers shop in an online store and pick up their goods in a convenience store. This model has had remarkable success. ezShip is the one of the retail delivery systems served by CVS.com. CVS.com is a joint venture fun by three convenience store chains: FamilyMart, OK, Hi-Life. This delivery system is complicated and risky. Vulnerability is a concept in risk analysis. If CVS.com managers know which parts of their business are vulnerable, they can take action in advance and create resources to deal with those vulnerabilities. Resources can then be distributed appropriately to avoid risk. The objective of this study is to discuss the vulnerabilities of the ezShip delivery process using quantitative methodologies, fault tree analysis (FTA), and Bayesian network (BN). Firstly, we make interviewed experts to collect information about the ezShip delivery system. Secondly, we define the top event as: buyers do not receive their goods on the third day” as the most undesired event, and then mapped and measured a fault tree for this system. We then converted it into a BN. Finally, we evaluated the BN and did both predictive and diagnostic analyses to uncover vulnerabilities in the ezShip process. The results of the BN were reasonable. It was found that technical personnel operational errors, POS system errors, system scheduling problems, unusual errors in, or breakdown of, the information system, and temporarily closed CS were the most vulnerable parts of the ezShip delivery system that had low failure probabilities. On the other hand, late e-map information updates, shift changes problems (DC), clerk errors, and shift changes problems were the most vulnerable parts that had high failure probabilities. Using our results, managers could reallocate resources and formulate strategies to strengthen the vulnerable parts of the system and reduce the frequency of the events that have higher failure probabilities.