Computational Intelligence Aided Aircraft Maintenance Planning

Abstract

The aviation industry is suffering under heavy price erosion, and airlines do their best to remain competitive and profitable. Airlines can only earn money if their aircraft are flying. Therefore an aircraft undergoing maintenance is not profitable and only causes costs. In order to reduce maintenance time, a good maintenance plan is mandatory. To find the best maintenance schedule and improve efficiency digital tools become more and more popular. This thesis introduced a new integrated maintenance approach by combining the aircraft routing or tail assignment problem with the maintenance task scheduling. Flying hours and performed flight legs heavily influence the latest possible maintenance date of a specific task. The more an aircraft flies, the earlier the task has to be performed. This thesis introduces a solution to maximise the quality of the maintenance plan while ensuring, that all scheduled flights of an airline can be served by an aircraft. The concept has been implemented based on a hybrid-metaheuristic approach combiningsimulated annealing (aircraft routing) and genetic algorithm (maintenance task scheduling)using multiple criteria to assess the fitness of the different calculated solutions. For evaluation reasons, the solution directly integrated into the maintenance software AMOS. The proposed approach is proven to work with positive feedback from experts.