機器人-人工協同整合之智慧物流系統

Abstract

現代電子商務供應鏈面臨不斷增加的顧客訂單挑戰。訂單量上升、勞動力限制以及快速、個性化的訂單交付需求，給倉庫帶來了巨大壓力。為應對這些挑戰，貨到人的訂單履行系統應運而生。該系統利用自主移動機器人取回物品並將其交付給人類工作者，在效率和適應性方面具有潛在的優勢。然而，要在貨到人系統中充分發揮人機協作的潛力，就需要深入了解人與機器人的協作。人為因素對貨到人系統的整體成功起著關鍵的作用。整合這些因素的周密計劃對於優化系統效率至關重要。本論文深入探討了貨到人的訂單履行系統中人為因素和訂單優先排序的具體方面，採用分析模型來指導戰略性決策。因此，本文分析在貨到人的背景下的人機協作，探討了倉庫規劃、設計和運營，以及提供相應的見解。 本文首次研究了感知工作負荷量，特別是人為因素對貨到人的訂單履行系統性能的影響。借助排隊理論，本文分析了部署機器人數量與隊列長度和系統吞吐量等指標之間的關係。此分析有助於揭示機器人的最佳部署水平，強調在設計系統時考慮人為因素的重要性。第二個目標探討了訂單優先級（例如“高級”訂單與普通訂單）在貨到人系統中的影響。利用賽局理論模型，該目標分析了接收不同訂單類型的機器人的隊列鏈接策略。研究結果可以動態管理訂單流量，最大限度地提高吞吐量，確保及時完成優先訂單，提高顧客滿意度。最後，本文研究了將壓力等額外人為因素納入貨到人的訂單揀貨系統的潛力。通過擴展以人為中心的分析，該目標旨在進一步優化在機器人輔助訂單履行環境中揀貨員的壓力和系統性能。 這項研究有助於深入地了解電子商務供應鏈背景下的人機協作。本文提供的見解可以改進訂單履行流程，提高顧客滿意度，以及創造對員工更友善的倉庫環境。

Modern e-commerce supply chains face relentless challenges of fulfilling increasing customer orders. Rising volumes, labor constraints, and the demand for rapid, personalized order delivery place immense pressure on warehouses. To address these challenges, the parts-to-picker order fulfillment system has emerged. This system leverages autonomous mobile robots that retrieve items and deliver them to human workers, offering potential gains in efficiency and adaptability. However, realizing the full potential of this human-robot collaboration in parts-to-picker system requires a deep understanding of the coordination between humans and robots. Human factors play a pivotal role in the overall success of a parts-to-picker system. Careful planning that integrates these factors is crucial to optimizing system efficiency. This thesis delves into specific aspects of human factors and order prioritization within parts-to-picker order fulfillment systems, employing analytical models to guide strategic decision-making. Accordingly, this thesis presents analyses of human-robot collaboration in the parts-to-picker context, exploring several objectives that offer insights for warehouse planning, design, and operation. This thesis first investigates the impact of perceived workload, a key human factor, on the performance of parts-to-picker order fulfillment systems. Drawing on queuing theory, this objective analyzes the relationship between the number of deployed robots and metrics such as queue lengths and system throughput. This analysis aims to reveal optimal robot deployment levels and highlight the importance of designing systems with human factors in mind. The second objective explores the implications of order prioritization (e.g., “prime” vs. regular orders) within a parts-to-picker system. Using game-theoretic models, this objective examines the queue-joining strategies of robots carrying different order types. The results can dynamically manage order flow to maximize throughput and ensure timely fulfillment of priority orders, enhancing customer satisfaction. Finally, the thesis investigates the potential to incorporate additional human factors, such as stress, into the operational parts-to-picker model. By broadening the human-centric analysis, this objective seeks to further optimize pickers’ stress and system performance within a collaborative robot-assisted order fulfillment environment. This work contributes to a deeper understanding of human-robot collaboration within the e-commerce supply chain context. It provides insights that can lead to enhanced order fulfillment processes, improved customer satisfaction, and the creation of more worker-friendly warehouse environments.