The Development of a Simulation Model for Airline Cockpit Crew Rostering

Abstract

This paper presents the development of a simulation model for airline cockpit crew rostering. The developed simulation model supports airlines in performing comparative analyses of different crew rostering scenarios. The model was developed in collaboration with a major European airline, leveraging actual historical pairing, crew and roster data to accurately model the airline crew rostering process. The simulation model is based on a novel solution method that is based on the Hungarian assignment algorithm and that exploits the resemblance of the weekly crew rostering problem to a minimum-cost bipartite matching problem. An assessment of the performance of the simulation model has shown that the simulation model is able to construct rosters of near-optimal solution quality in limited computational time. To test the applicability of the developed model as a scenario analysis tool, two experiments have been performed. The insights the performed experiments have provided confirm the practical value of the developed simulation model.