Robust crew pairing : delays analysis and implementation of optimization approaches

Abstract

With increasing delays and airport congestion that disturb airline operations, the development of robust schedules is becoming crucial. Increased traffic and poor weather are a few of the causes of airport congestion, rising delays and lengthening passenger trips. In this thesis, we identify the latest trend in the flight arrival and departure delays, differentiating major U.S. airports from other smaller airports. We also quantify the types of delays airlines should work to mitigate. We then analyze the effects of schedules changes that were implemented by a major U.S. airline at their largest hub. We measure the effects of these schedule changes on on-time performance, taxi time, plane utilization, and passenger connection and total travel time. We also analyze how extensive is the practice of adding buffer time to flight times to improve schedule reliability. Finally, we propose and implement a new model to achieve robust crew schedules, that is, crew schedules that are less likely to be inoperable due to disruptions during operations. We show that with an increase in crew costs of 0.2%, we can decrease the number of times crews must connect between different aircraft by 32%.