必要工具儲放與取送服務的物流方略

Abstract

本研究啟發自一所大型醫院實體病歷物流服務的效率改善，主要的硏究目的則是針對所有具類似問題特徵的物流服務系統，使能以有限的資源作創新的策略規劃，來縮短物流運作負荷與客戶等候。研究過程分別以倉儲規劃以及運輸設計作為策略再造的兩個切入點：(1)倉儲規劃方面，針對具隨機性回訪需求特徵的必要工具提出一種優質的「時近型儲位指派策略」以及「串級式儲區組態配置」，儲位指派與儲區配置的目標分別為縮減撿貨的巡行距離以及倉儲成本。首先描述反覆性需求的模型，其次是新策略與傳統序號型儲位指派策略的存取績效分析，接著提出一個串級式儲區配置的最佳化數學規劃模型，最後則進行一家醫院客戶回診的個案研究；(2)運輸設計方面，針對雙目標多週期的取送貨途程提出一種結合「跨週期的中心轉運」與「彈性化的週期設定」的運作策略，其目標為同時縮減快遞服務的工作負荷量以及客戶需求的遞送時間。首先是新策略與現行週期途程策略的績效分析，接著以醫院的實況模擬資料來進行個案研究，最後則探討戰術面的彈性運用以及其最佳化數學規劃模型。整篇論文針對倉儲規劃以及運輸設計所提出的各項創新策略與作法，皆經過平均值分析方法的推導以及模擬方法的驗證，可證明其績效確實優於實務上的現行作法。本研究也釐清欲有效運用這些新策略時所必須跨越的門檻條件。此成果可延用到其它產業領域具有類似問題特徵的物流服務。

This research is motivated by a project on improving hospital logistics. The purpose is to reduce workload and customer waiting under limited resource with innovative strategy for all systems of similar characteristics. Strategy redesign on both storage and transportation is used to approach this logistics problem. (1) For storage, a recency-based storage assignment policy and a cascaded warehouse configuration for essential accessory of recurrent demand are proposed. The objectives are to reduce traverse and operation cost respectively. A model of recurrent demand is first built. Efficiency of the new policy and current ID-based policy is then analyzed. After that, a programming model for optimal configuration of cascaded warehouses is presented. Finally, a case study of hospital visits is performed. (2) For transportation, it proposes an operation design of hub transshipment across time periods and flexible period setting for dual-objective multi-period pickup and delivery routing problem. The objectives are to reduce workload and delivery time. The new policy and existing periodical routing policy are first compared. A case study is then performed by using field data of the hospital. Finally, flexibility in application and its programming models are presented. Being verified with mean-value analysis and simulation, all proposed innovative methods in this dissertation are outdoing the current ones. Thresholds for applying these new policies are also analytically derived. The new policies can be extendedly applied to other logistics fields of similar pattern.