Proceedings of the 1st International Conference on Aircraft Operations Management (AOM) vol:1 pages:60-63
International Conference on Aircraft Operations Management (AOM) edition:1 location:Munchen (Germany) date:17-19 September 2012
Scheduling aircraft maintenance personnel at an aircraft maintenance company entails some special problems. First, the workforce scheduling problem is heavily constrained by labour union agreements. Second, the maintenance jobs that have to be carried out are malleable. This means that the company should decide when exactly the maintenance will take place between the arrival and departure of the aircraft. Hence, the timing of the workload is an extra decision in the scheduling problem. Third, aircraft tend to not always fly on schedule and arrive now and then with a delay. When the workforce scheduling does not anticipate delays in arrival time, the scheduled capacity may be insufficient to maintain all aircraft in time. This paper focuses on this latter problem and presents a technique to obtain robust aircraft maintenance personnel rosters that minimize the total labour costs. Robust personnel rosters are rosters that can handle delays in aircraft arrival times. We define the stochastic robustness of these rosters as their ability to ensure a certain service level; i.e., to ensure that on average at least a certain percentage of the flights can be maintained before their Scheduled Time of Departure (STD). To obtain this stochastic robustness, we propose a model enhancement heuristic that iteratively enhances a deterministic Mixed Integer Linear Programming (MILP) model by adding constraints based on information that is obtained by simulation experiments. This research is inspired by the real life problems at Sabena Technics, a large aircraft maintenance company located at Brussels Airport in Belgium.