Revenue management (RM) attempts to understand and shape consumer behavior to maximize revenue from a perishable resource. Various algorithms can be used to control bid-prices, and subsequently, perform differently with respect to the total network revenue that they generate. There is currently a need for some method to compare RM engines; a simulation can fulfill this need.

The first module of this thesis will create a statistically accurate representation of customers arriving at ticket purchasing channels. Each customer's attributes are: arrival time, origin and destination, number of destined tickets, and willingness to pay. Each attribute can be generated using a specific distribution.

The created customers will then be used to simulate the purchase of tickets and overall revenue for a flight network. With a valid simulation, airlines will be able to compare the performance of different RM engines under various circumstances.

Most staff planning for airline industries are done using point estimates; these do not account for the probabilistic nature of employees not showing up to work, and the airline company risks being under or overstaffed at different times, which increases costs and deteriorates customer service. This model proposes utilizing a stochastic method for American Airlines to schedule their ground crew staff. We developed a stochastic model for scheduling that incorporates the risks of absent employees and as well as reliability so that stakeholders can determine the level of reliability they want to maintain in their system based on the costs. We also incorporated a preferences component to the model in order to increase staff satisfaction in the schedules they get assigned based on their predetermined preferences. Since this is a general staffing model, this can be utilized for an airline crew or virtually any other workforce so long as certain parameters about the population can be determined.

The crew planning problem in the airline industry presents a very computationally complex problem of high importance to the business. Airlines must schedule crew members to ensure that all flights are staffed while remaining in compliance with the business needs and regulatory requirements set by entities such as unions and FAA. With the magnitude of operation of the prominent players in the airline industry today, the crew staffing problem proves very large and has become heavily reliant on operations research solution methodologies. An area of opportunity that has not yet been extensively researched lies in the planning of crew vacation. This paper develops a model driven by the idea of system risk that constructs an optimal vacation grid for the time period of one year. The model generates a daily allocation that maximizes vacation offering while ensuring a given level of system reliability. The model is then implemented using data from US Airways and model improvements are provided for practical application in the airline industry based on the output.