A three-stage mixed integer programming approach for optimizing the skill mix and training schedules for aircraft maintenance

This paper presents a three-stage mixed integer programming approach for optimizing the skill mix and training schedule for aircraft maintenance workers. When all workers are trained for all skills, cheaper workforce schedules are possible. However, the training that is required to acquire all those skills can become very expensive. In the first and second stage, we therefore make a trade-off between the training costs and the resulting cheaper workforce schedule. As we assume that workers are unavailable to work during their training, the resulting schedules are only applicable in practice if the required training can be performed without endangering the current maintenance operations. In the third stage, we therefore want to find an optimal and feasible training schedule in order to obtain the desired skill mix with minimal costs. A computational experiment based on real-life data of an aircraft maintenance company not only demonstrates that our models succeed in finding good solutions within reasonable computation times, but also illustrates how the explicit incorporation of skills training in the scheduling process can lead to significant cost savings.