近年來電子商務與供應鏈整合崛起，全球的商業模式有顯著的改變。現今的商業模式為了滿足消費者「少量、多樣、多頻率」的消費習慣，發展出現代化的物流作業。在工業4.0和物聯網的趨勢下，自動化與智慧化的物流中心能更加滿足顧客高頻率且大量的訂單。 亞馬遜的第八代物流中心，是結合各種自動化設備而成的現代化物流中心。其核心技術，Kiva系統，是現代化物流中心最有效率的揀貨策略之一。利用中央電腦對系統中的Kiva機器人下達指令，搬運貨架（pod）至揀貨工作站的揀貨員面前執行揀貨作業，以此模式節省揀貨員的旅行時間來達成提升揀貨效率之目的。 本研究針對Kiva系統中的兩個決策問題做探討，分別是訂單選取問題和揀貨工作站挑選問題，並針對兩個研究問題分別提出數個研究法則。利用仿真模擬軟體Arena建構類似Amazon物流中心的環境，從兩個研究問題的數個法則中找出績效最高的因子組合，提升揀貨中心作業效率。 ;In recent years, E-commerce and Supply chain integration emerge, and the global business model has changed significantly. In order to satisfy customer’s consumption habits, such as small amount, high diversity and high frequency, the modern logistics operations are developed. In industry 4.0 and IoT trend, automated and intelligent logistics center can satisfy customer’s high frequency and large amount of orders more. Amazon′s eighth-generation logistics center, is a modern logistics center combined with various automation equipment. Kiva system, the core technology, is one of the best solution of picking operation. The computer will command kiva robots to carry pod to workstation and execute picking operation. This new method saves traveling time of picker and improve picking efficiency. In this study, I focus on two research problems, workstation selection of order assignment and order selection, and also propose few methods of this two problems. Furthermore, I used simulation software called Arena to build the simulation environment like logistics center of amazon and try to find the best method to solve these two research problems.
