The value of flexible flight-to-route assignments in pre-tactical air traffic management

Abstract

In European air traffic management, there are various discussions regarding the future role of the network manager (NM): in particular, should their role be strengthened to be able to assign flights to specific trajectories, or should airspace users be allowed to freely choose their preferred trajectory, or something in between? If the latter, how would this work? In this paper, we develop a modelling framework that can be adapted to these different settings so as to quantify their effect on key performance indicators.

We focus on the pre-tactical stage of planning air traffic for a future departure day, meaning that airspace capacity budgets are given and incoming flight intentions need to be offered one or several `trajectories products' for a (possibly dynamically determined) charge. These trajectory products differ in the amount of flexibility that they provide the NM to route the flight shortly before the time of departure. The idea is to reward greater flexibility of the airspace users with lower charges. The airspace user considers the options of trajectory products offered, and chooses one according to a choice model reflecting their preferences in light of various factors including the dynamic trajectory product charges. Shortly before the departure day, the NM decides simultaneously on the routing (within the limits defined by the purchased trajectory products) and on each airspace's sector opening scheme (within the limits of the fixed capacity budgets) so as to minimize the total displacement costs. Charges are set so as to just recover the exogenous cost of capacity budgets and to influence airspace users in their trajectory product choice.

Methodologically, the problem deviates from typical dynamic pricing problems in various major ways (such as featuring a hard boundary condition as well as fairness and revenue neutrality constraints). The problem is cast in the form of a dynamic program with a boundary condition that we show to be NP-hard. We exploit a certain structure in this boundary problem so as to formulate an efficient heuristic. Based on a numerical case study, we find that the use of these trajectory products along with dynamic pricing can be highly beneficial to the extent of achieving a cost performance close to the one obtained if the NM has a mandate to simply assign flights to trajectories shortly before departure. Therefore, this seems an attractive design for the role of the NM, giving airspace users some choice whilst achieving low overall costs.

Keywords: dynamic pricing, trajectory optimization, demand-capacity balancing, route choice