报告题目：Spatial Flexibility and Labor Elasticity in Ride-Hailing with Autonomous Vehicles
报 告 人：唐润宇副教授
工作单位：西安交通大学
报告时间：2026年10月16日 15:30-17：30
报告地点：数信学院211会议室
报告摘要：Ride-hailing platforms match uncertain demand through two forms of flexibility that operate at different times. Owned autonomous vehicles (AVs) provide spatial flexibility: capacity is committed before demand is known but can be repositioned after a contextual signal. Independent contractors (ICs) provide labor elasticity: local capacity can be recruited after demand is realized, but additional participation requires increasingly costly incentives. We develop a three-stage stochastic model in which the platform first sizes and locates its AV fleet, then repositions AVs after observing contextual demand signals, and finally recruits ICs to serve residual demand. The analysis shows that spatial flexibility is not simply a substitute for fleet size. Reducing repositioning costs expands the combined AV--IC fleet when profit margins are low, because it makes additional demand worth serving, but contracts the fleet when profit margins are high, because the platform can cover imbalances with fewer committed vehicles. The effect of IC incentive costs likewise depends on profit margins: weaker IC labor elasticity may reduce service when profit margins are low, but increase combined fleet size when profit margins are high. A midpoint depot adds no value in the two-region model: there is always an optimal deployment with no AVs at the depot because regional placement preserves local readiness and the option to reposition. Under asymmetric demand, higher mean demand attracts base AV capacity, whereas higher variance attracts AVs only in high-margin markets where upside demand is valuable. An NYC-based numerical study illustrates these mechanisms under real-world spatial and temporal demand heterogeneity. The results show that AV adoption should be evaluated jointly with IC incentives, deadheading costs, and fleet controls, rather than as a stand-alone technology-investment decision.
报告人简介：
唐润宇，现任西安交通大学管理学院副教授，博士毕业于清华大学经济管理学院。目前主要研究兴趣为智慧城市运营管理，数据驱动建模与优化，动态机制设计等。相关成果发表于Operations Research, Manufacturing and Service Operations Management, Production and Operations Management, 管理世界等国内外权威期刊。主持国家自然科学基金青年基金及面上基金各一项。
