| 摘要: | 近年來資訊科技迅速發展，再加上行動網絡的普及化，造成電子商務的盛行，因此市場需求逐漸轉變為「少量、多樣化」，市場需求的改變同時也提升了物流中心的作業難度，其中對於揀貨作業更甚明顯。根據De Koster et al.（2007）的研究指出，目前大多數的物流中心仍屬於勞力密集的產業，揀貨作業不僅相當耗費成本，更是一種屬於勞力密集的活動，在物流中心裡與揀貨作業相關的人力佔了50%以上。因此為了因應「少量、多樣化」需求的時代來臨，適時地導入自動化設備，並規劃一個合適的揀貨策略，將對物流中心的成本、產能以及效率有著決定性的影響。
基於前述原因，本研究旨在深入探討類Locusbots系統在分區揀貨倉庫中的揀貨策略，與以往Locusbots系統研究不同的是此研究的環境為類Locusbots分區揀貨倉庫。我們打破了傳統揀貨人員只能在自己的Block進行揀貨作業的限制，揀貨人員可以在屬於自己Zone內的Block互相幫忙，吾人將針對「揀貨人員的Block選取問題」進行探討。本研究對「揀貨人員的Block選取問題」提出了第一類及第二類法則，其中第一類法則為本研究所設計之揀貨人員的Block選取法則；而第二類法則則是對第一類法則的延伸，其目的為事先在其所屬Zone內挑選出較繁忙的Block進行揀貨作業，評估方式為計算揀貨Block之（Robot數量/揀貨人員數量）的比例（又稱R/P ratio），若比例大於1，則代表其為繁忙Block，挑選出繁忙Block後再使用本研究所設計之揀貨人員的第一類Block選取法則針對其進行揀貨。若無R/P ratio > 1的Block，則直接使用揀貨人員的第一類Block選取法則選擇Block進行揀貨。
最後，我們利用Arena模擬軟體進行實驗，分析「揀貨人員的Block選取問題」之第一類與第二類法則比較，並針對本研究所提出的數種法則在不同績效指標下的表現。期望藉此找出最適合的揀貨策略組合，同時為未來類似研究提供有價值的參考與貢獻。
;In recent years, the rapid development of information technology, coupled with the widespread adoption of mobile networks, has led to the prevalence of e-commerce. Consequently, market demands have shifted towards "small quantity and diversity," increasing the complexity of operations in logistics centers, particularly in picking activities. According to the study by De Koster et al. (2007), most logistics centers remain labor-intensive industries. Picking operations are not only costly but also labor-intensive, with related personnel accounting for over 50% of the workforce in these centers. Therefore, to adapt to the era of "small quantity and diversity" demands, timely introduction of automated equipment and planning an appropriate picking strategy is crucial in significantly impacting the cost, capacity, and efficiency of logistics centers.
Based on the aforementioned reasons, this study aims to delve into the picking strategies in zone-picking warehouses similar to those using LocusBots systems, setting it apart from previous studies on LocusBots systems. We have broken the traditional constraint that pickers can only operate within their designated blocks. In our system, pickers are allowed to assist each other within blocks that belong to their own zone. This study explores the "Block Selection Problem for Pickers," proposing two types of rules. The first type, designed in this study, governs the picker′s block selection, while the second type extends the first, aiming to preemptively select busier blocks within their zone for picking. The evaluation method involves calculating the ratio of the number of robots to the number of pickers (R/P ratio) in a picking block. If the ratio exceeds 1, it indicates that the block is busy. Once busy blocks are identified, the first type of block selection rule designed in this research is applied. If there are no blocks with an R/P ratio greater than 1, pickers directly use the first type of block selection rule.
Finally, we utilize the Arena simulation software to experimentally analyze and compare the first and second types of block selection rules and evaluate the performance of the proposed rules under different performance indicators. Through this, we aim to identify the most suitable combination of picking strategies, providing valuable insights and contributions for future similar research. |
