Skyport location problem for urban air mobility system

With rapid advancement in aerial vehicle technologies, air taxi is drawing keen interest as a new type of transportation service. One of the necessary elements for air taxi-based transportation service system is a skyport, which is a designated area for take-off and landing of air taxis. In planning a skyport network for air taxi transportation system, it is important to consider midair congestion between air taxis. Severe congestion increases the risk of collisions, which in turn increases the complexity in scheduling and routing in the subsequent operations phase. In this study, we propose a novel hub location problem (HLP) that incorporates traffic congestion caused by intersecting vehicles along hub-to-hub arcs. Modeling congestion in an HLP formulation involves nonconvex bilinear terms, which makes it difficult to solve. We tackle this challenge by exploiting a property of an optimal solution and reformulating the original model to obtain a linear formulation. We also develop a heuristic algorithm based on a genetic algorithm to handle large-size problems for practical applications. Using the proposed model and solution methods, extensive experiments have been conducted to study optimal design of air taxi network under various environments. Our analyses demonstrate that viability of air taxi service is a complicated function of many factors including collision risk, system cost, airspace design, to name a few. Our model can offer quantitative understanding of such complex relationships to assess and develop strategies to improve the viability of an air taxi transportation system.