| 摘要: | 電子商務的成功因素有許多，其中之一就是完善的物流服務。全球電商龍頭亞馬遜深知此道理，在2014年開始啟用Kiva機器人來改善揀貨作業。Kiva揀貨系統顛覆傳統的揀貨方式，在傳統的揀貨系統，貨架是固定的，揀貨人員必須行至貨架進行揀貨；但是在Kiva揀貨系統，貨架則是可以移動的，Kiva機器人載貨架（亞馬遜稱此貨架為Pod）至揀貨人員面前進行揀貨。自從亞馬遜使用Kiva揀貨系統後，大陸也有公司開發出類似的系統，並成功地被網路電商所使用。如同製造業的「工業4.0智慧自動化製造」，利用Kiva機器人於揀貨自動化將是物流業邁向「物流業4.0智慧自動化物流」的重要里程碑。Kiva揀貨系統的成功不是僅仰賴硬體的創新，同樣重要的是其系統內的作業管理、決策與控制問題的解決，這些問題會影響系統的表現。本研究的目的是希望瞭解類Kiva揀貨系統會有哪些作業管理、決策與控制問題；然後研究這些問題，並提出解決方法；最後，進行模擬測試來比較所發展的方法。本研究規劃以三年的時間來進行。第一年計畫已完成，第二年計畫目前進行中，此次申請的是第三年計畫。第一年的問題有「訂單選取問題」與「訂單分配（至揀貨工作站）問題」。第二年則有「Pod分配之揀貨工作站挑選問題」、「Pod分配問題」以及「Pod分配之品項分配問題」。接續前兩年的研究，第三年計畫將針對「Kiva派車問題」、「補貨條件問題」、「補貨方法問題」以及「Pod等候暫存區位置指派問題」來進行探討。「Kiva派車問題」的目的是要找一台最適合的Kiva機器人來派送它去載Pod並送至揀貨工作站。「補貨條件問題」的目的是要找出最好的Pod補貨條件。「補貨方法問題」的目的在找出最好的Pod補貨方法。最後，「Pod等候暫存區位置指派問題」的目的是當Pod要回去Pod等候暫存區時，如何從Pod等候暫存區找出一個空格子位置給Pod。這四種問題對系統的揀貨表現是有影響的，也因此，針對這些問題發展優良有效的方法是有其必要性的。本研究希望研究成果有應用價值，因此在方法發展上會重視方法於實際環境的有效性與可行性。雖然目前國內廠商尚未有使用類Kiva揀貨系統的產品，但是本研究相信類Kiva揀貨系統將會是物流倉庫智慧自動化的未來主流之一。本研究相信本計畫成果將能對國內廠商，未來在使用或是開發類Kiva揀貨系統上有所助益，也因此希望能有機會來順利完成本研究所規劃的三年期計畫。 ;Research shows e-commerce has many success factors and good logistics service is one of them. Amazon, the global e-commerce leader, knows this well. In 2014, it unveiled its eighth-generation logistics center which used Kiva robots to improve order-picking operations. Kiva picking systems drastically change the traditional picking method. In the traditional picking system, shelves are fixed and pickers approach shelves for picking. However, in a Kiva picking system, shelves (also known as pods) are removable. Kiva robots carry shelves to pickers for picking. Similar to the “Industry 4.0’s intelligent and automated manufacturing” in the manufacturing industry, using Kiva robots for picking operations will be an important milestone for the logistics industry toward the “Logistics 4.0’s intelligent and automated logistics center.” The success of Kiva systems is not solely due to the innovation of hardware devices, equally important are the solutions to operations management, decision-making and control problems within the system, as the resolution of these problems greatly affect the system’s performance. The purpose of this project is to study them and propose solutions for them; and conduct simulation experiments to test them. This project plans to study the operations management, decision-making and control problems for a Kiva-like picking system and proposed solutions for them in three years. The first year of the project has been completed. The second year of the project has been subsidized and is currently in progress. This application is for the third year. The first year's research problems are “order selection” and “order assignment (to picking workstations).” The second year’s problems are the “workstation-selection problem of pod assignment”, “pod assignment problem” and “item-allocation problem of pod assignment.” The third year’s problems include “Kiva dispatching problem”, “replenishment condition problem”, “replenishment method problem” and the “Pod waiting location assignment problem.” The purpose of the “Kiva dispatching problem” is to find out which Kiva should be dispatched to carry a Pod. The purpose of the “replenishment condition problem” is to find the best replenishment condition. The purpose of the “replenishment method problem” is to find the best replenishment method. And, the purpose of the “Pod waiting location assignment problem” is to find the location that a Pod can stay and wait for its next task. These four problems can affect the performance of the Kiva system. Therefore, it is necessary to develop excellent and effective methods for them. This project hopes the research results will have substantial application value, thus when developing methods, this project will pay attention to the feasibility and effectiveness of methods in the actual environment. Although, at present, domestic companies have not yet used or developed any Kiva-like picking systems, we believe Kiva-like picking systems will be one of the future mainstreams of intelligent and automated logistics centers. Therefore, we believe the results of this project will benefit domestic companies in their future adoption or development of Kiva-like picking systems and hope to have the opportunity to successfully complete this three-year research project. |
