喻园管理论坛2024年第58期(总第990期)

演讲主题: Dynamic Switching Models for Truck-only Delivery and Drone-assisted Truck Delivery under Demand Uncertainty

主 讲 人: 郭倩雯，佛罗里达州立大学助理教授

主 持 人: 盛    典，管理科学系副研究员

活动时间: 2024年5月30日（周四）10:00-12:00

活动地点: 管院大楼125室

主讲人简介：

郭倩雯博士是佛罗里达州立大学助理教授。郭倩雯博士对交通优化问题有着广泛的研究兴趣，尤其是在公共交通、智能共享出行和基础设施投资决策方面。她在交通领域顶尖期刊如Transportation Research Part A，B，C上发表了多篇文章。目前她正以项目负责人身份主持美国自然科学基金委NSF及美国交通部USDOT多项项目，包括开发用于工程教育的虚拟现实实验室、桥梁和道路基础设施网络容灾性、以及公共交通规划和运营的策略。

活动简介：

Integrating drones into truck delivery systems holds the potential for transformative improvements in customer accessibility, operational cost reduction, and delivery efficiency. However, this integration is not without its associated costs, including drone procurement, maintenance, and energy consumption. The decision on whether and when to incorporate drones into truck delivery systems is heavily contingent on the level of demand for these services. In areas where demand is low and dispersed, deploying drone-assisted trucks may lead to the underutilization of resources and financial challenges, primarily due to the substantial upfront costs of drone deployment. Accurately predicting future demand density is a complex task, compounded by uncertainties stemming from unforeseen events or infrastructure disruptions. To tackle this challenge, a market entry and exit real option approach has been used to determine the switching timing between delivery methods while considering the stochastic nature of demand density. The results of this study highlight that drone assisted truck delivery, particularly when multiple drones are deployed per truck, can offer significant economic advantages in regions with high demand density for delivery services. Utilizing the proposed dynamic switching model, both deterministic and stochastic approaches result in a 23.7% and 43.0% reduction in costs compared to a static model, respectively. Furthermore, the stochastic parameters within the real option framework asymmetrically influence the entry and exit timings, as revealed through sensitivity analysis. The proposed dynamic stochastic models are applied in Miami-Dade County area to evaluate the cost of dynamic switching services for three major logistics companies in a real-world scenario. This research illuminates the potential benefits of dynamic switching between different delivery modes and provides decision-makers in the logistics industry with valuable insights into optimizing their delivery systems.