Domain-driven multiple-criteria decision-making for flight crew decision support tool

During the flight, the crew might consider modifying their planned trajectory, taking into account currently available information, such as an updated weather forecast report or the already accrued amount of delay. This modified planned trajectory translates into changes on expected fuel and flying time, which will impact the airline’s relevant performance indicators leading to a complex multiple-criteria decision-making problem. Pilot3, a project from the Clean Sky Joint Undertaking 2 under European Union’s Horizon 2020 research and innovation programme, aims to develop an objective optimisation engine to assist the crew on this process. This article presents a domain-driven approach for the selection of the most suitable multiple-criteria decision-making methods to be used for this optimisation framework. The most relevant performance indicators, based on airline’s objectives and policies, are identified as: meeting on-time performance, leading to a binary value in a deterministic scenario; and total cost, which can be disaggregated into sub-cost components. The optimisation process consists of two phases: first, Pareto optimal solutions are generated with a multi-objective optimisation method (lexicographic ordering); second, alternative trajectories are filtered and ranked using a combination of multi-criteria decision analysis methods (analytic hierarchy process and VIKOR). A realistic example of use shows the applicability of the process and studies the sensibility of the optimisation framework.