A Passenger-Centric Approach to Air Traffic Flow Management

Abstract

Traffic flow management initiatives aim to balance the flows of vehicles in capacity-constrained transportation networks to minimize congestion costs. While existing research has focused on vehicle-centric delay minimization objectives, this may not result in most efficient outcomes from the end users' perspective when travel itineraries involve connections between multiple vehicles. This paper develops an original user-centric approach to traffic flow management. First, an analytical Markov Decision Process is presented to derive structural insights on the drivers of user-centric operations. Second, this paper augments existing air traffic flow management (ATFM) models by explicitly balancing flight delay costs and passenger delays. A large-scale integer programming model is formulated to track the impact of flow management operations on passenger accommodations. An original rolling procedure decomposes the problem over time while ensuring global feasibility, and shown to enable the model's implementation in short computational times. Computational results in the US National Aviation System suggest that this modeling and computational framework can achieve large reductions in passenger delays at comparatively small increases in flight delay costs. Analytical and computational results highlight two major levers of user-centric operations: (i) delay allocation, which determines which flights to delay or prioritize to minimize delay costs, and (ii) delay introduction, which deliberate adds departure holds to avoid passenger misconnections.