Another class of problems in fleet management aside from those discussed, is the one concerning air transport. This is characterized by network design and schedule construction, fleet assignment, aircraft routing, crew scheduling, revenue management, irregular operations, air traffic control and ground delay programs, gate assignment, fuel management, short term fleet assignment swapping. They were mostly solved by operation research techniques and the majority of applications utilized network-based models. 

The airline scheduling process is carried out sequentially so that flight, aircraft and schedules are created one after another over several months prior to the day of the operations. A detailed flight schedule might be based on marketing decisions. The first step in operational scheduling is the assignment of an aircraft fleet type to each flight and is based on the demand forecasts, the capacity and the availability of the aircrafts. After fleet assignment, an aircraft is assigned to each flight with respect to maintenance constraints such as aircraft routing. Crew scheduling can be broken down into two steps. The first phase is called crew pairing and it involves anonymous crew itineraries subject to constraints such as maximum allowed working time or flying time per duty. The second phase is crew rostering, and it involves assigning individual crew members to the itineraries. The goal of this scheduling process is to reduce costs. 

Fleet routing and fleet scheduling also affect costs but it determines the airline’s level of service and its competitive capability in the market. Network flow techniques are adopted for modeling and solving such complex mathematical problems. The full optimization problem can be hard so they are solved in parts sequentially. The output of one is input to the next. 

The limitations of the sequential approach were subsequently solved with an integrated approach that reduces costs even more. 

The fleet assignment problem deals with assigning aircraft types, each having a different capacity to the scheduled flights, based on equipment capabilities and availability, operational costs and potential revenues. When there are many flights each day, this problem becomes difficult. Some remediations include: 1) integrating the FAP with other decision processes such as schedule design, aircraft maintenance routing, and crew scheduling, 2) proposing solution techniques that introduces additional parameters and constraints into the traditional fleeting models, such as itinerary based demand forecasts and the recapture effect and 3) studying dynamic fleeting mechanisms that update the initial fleeting solution as departures approach and more information is gathered on demand patterns. In a few models, a non-linear integer multi-commodity network flow is formulated, and new branch-and-bound strategies are developed. 

Traffic disruptions are one characteristic of this problem space. This might lead to an infeasible aircraft and crew schedules on the day of the operations and the recovery to reasonable schedule must be attempted. The short-term recovery actions might increase operational costs, sometimes even higher than the planned costs. Recovery options could be factored into the scheduling at the design time and this approach is generally called robust scheduling. Sometimes this is articulated as a measure. For example, a non-robustness measure is used to penalize restricted aircraft changes according to the slack time during an aircraft change. 

Global stochastic models have been attempted to be solved with an iterative approach. The iterative approach yields a set of different solutions regarding the trade-offs between the costs and robustness whereas an integrated approach returns mostly one near-optimal solution for a given robustness penalty. Iterative approach is more favorable to a decision maker. When multiple airlines must coordinate, the models are formulated as multiple commodity network flow problems which can be solved by programs based on mathematical formulations.