Scheduling Aerial Vehicles in Large Scale Urban Air Mobility Schemes with Vehicle Relocation

In recent years, a revolutionary idea has been proposed to leverage the third dimension and dramatically transform urban transportation. The concept of Urban Air Mobility (UAM) involves Aerial Vehicles (AVs), piloted by humans or flying autonomously, that can transport commuters across pairs of locations devoted to takeoff and landing. The scheduling of such vehicles, aiming for the maximization of the number of commuters serviced by AVs, is a challenging problem. This work initially formulates the AV scheduling problem in a UAM-based scheme using Integer Linear Programming (ILP) techniques, providing an offline optimal solution to this problem. As this problem is highly combinatorial, the exact solution has limited scalability and for this reason two greedy algorithms are also developed, that perform near-optimally with considerably lower execution times. The first heuristic algorithm calls and solves the ILP incrementally for each vehicle, while the second heuristic relies on a look-ahead search approach. The algorithms are then extended to support vehicle relocation in order to increase the travel requests that are serviced. All algorithms are evaluated in a number of realistic scenarios to verify their performance and demonstrate their applicability.