隨著快商務 (Quick Commerce) 時代來臨,為了提升配送速度,有越來越多的物流業者會在更靠近消費者的區域建置倉庫,其中占地300至500坪左右並設置在都會區的次微型速配中心 (Sub-Micro Fulfillment Center, SMFC) 是一種適合的方案。 在傳統上,評估揀貨效率常以「距離」作為優化程度的指標。然而在實際作業上,揀貨波段完成時間的精確預估更為重要。有鑒於此,本研究以實務應用為基礎,透過系統模擬探討SMFC可行的揀貨模式。首先運用位階順序分群演算法(Rank Order Clustering, ROC) 將訂單批次處理,再針對揀貨順序進行最短路徑規劃,達到縮短移動距離以提高揀貨效率。接著透過FlexSim模擬軟體建立「以人為基礎」的多單同揀模型以及「以自動化設備」為基礎的無人搬運車 (Automated Guided Vehicle, AGV) 揀貨模型進行實驗。 為了使不同的揀貨模式能夠具備一致性,並在考量個別特性的情況下進行比較,本研究以系統模擬為基礎,提出了異質性物流方案評估架構 (Heterogeneous Logistics Alternatives Evaluation Framework, HLAEF),深入探討SMFC採用不同揀貨模式之特性與作業績效。 經本研究模擬實驗後,可以得知影響揀貨效率的主要因素以及SMFC最佳揀貨人數與AGV數量配置,並在本研究所提出之方案評估架構下,了解SMFC採用不同模式的揀貨能力與作業瓶頸。最後,本研究也對SMFC採用不同揀貨模式提出建議,並提供未來可持續研究之方向(例如分析不同揀貨模式之成本與報酬率)。
With the rise of Quick Commerce, more and more logistics companies build warehouses in metropolitan area. In this context, Sub-Micro Fulfillment Center (SMFC) which is about 10677 to 17795 square feet is a suitable solution. Traditionally, "distance" is an indicator of the degree of optimization to assess picking efficiency. However, the accuracy of estimation for picking makespan in wave picking is important. In view of this, this study discusses a model that is based on practical application for picking in SMFC by simulation. First, this study employs Ranked Order Clustering (ROC) for order batching, then find the shortest routes by picking sequence to enhance picking efficiency. Second, building two models to simulation by FlexSim. One is a "human-based" model for pick-to-box and the other is "automated equipment-based" for picking by AGV. In order to have consistency in the different picking modes and to compare them in unique characteristics of individual solutions. Based on the simulation, this study proposes the Heterogeneous Logistics Alternatives Evaluation Framework (HLAEF) to investigate the characteristics and performance of different picking modes by SMFC. After the simulation experiment, the main factors affecting the picking efficiency and the optimal number of pickers and AGVs for SMFC can be known. The solution framework proposed in this study can provide the picking capacity and bottleneck of SMFC with different modes. Finally, this study provides directions for future sustainable research (e.g., cost-benefit analysis of SMFC using different modes).