A Stochastic Model of Airline Operations

Decision makers in air transportation face many uncertainties. In spite of this, airline planning models do not explicitly consider uncertainty in operations. As a result, there is often a notable discrepancy between an airline’s planned and actual performance. We present a stochastic model of the daily operations of a domestic airline. Its primary purpose is to evaluate plans, such as crew schedules, as well as recovery policies in a random environment. We describe the structure of the stochastic model and give computational results. Finally, we give future directions for the study of airline recovery policies and planning under uncertainty. In airline operations, there are many uncertainties. A disruption is an event that prohibits an airline from operating as scheduled. Anecdotal evidence suggests that major domestic airline carriers almost never experience a day without disruptions. Furthermore, on-time performance is deteriorating. Average daily flight delays increased 20% from 1998 to 1999 [9]. Consequently, customer complaints also increased by 130% [23]. In June 1999, the airlines began a “Customer First” campaign to avoid additional government regulation [23]. Unfortunately, airline performance has not improved. In May 2000, there were four days with over 2000 air traffic delays each [1]. In June 2000, flight delays increased 16.5% from June 1999 [22]. Several factors in airline systems account for the current frequency of disruptions. There are more passengers, and planes are more crowded than at anytime in history [22]. Air traffic controllers are struggling to accommodate the number of flights the airlines want to offer [22]. Mechanical failures can disrupt the planned schedule. On June 10, 2000, 71 flights were cancelled due to mechanical problems by one major