Airline Disruptions: Aircraft Recovery with Maintenance Constraints

Airline schedules are rarely performed as planned because of irregularities such as delays, unpredicted maintenances or bad weather conditions. Recover from a disrupted schedule as quickly as possible is a hard and high priority problem for practitioners: given a disrupted situation they have to take decisions in order to minimize the number of canceled flights and the average delay respecting all technical requirements. We present a column generation based algorithm to solve the airplane recovery problem including maintenance constraints and show, through computational results, the added value of including the maintenance scheduling in the recovery problem.