Volgenau School of Engineering
George Mason University
George Mason University Mason
George Mason University

Big Data Provides Solutions for Airline Savings and Passenger Satisfaction

May 20, 2015

What if you didn't have to go to the airport to rebook a cancelled flight? What if you could use your smartphone to rebook before ever leaving your home or hotel? Sound impossible¬–maybe not.
 Big Data Provides Solutions for Airline Savings and Passenger Satisfaction
Sanja Avramovic, a PhD candidate from the Volgenau School of Engineering and Associate Professor Lance Sherry may have a workable solution for passengers and airlines. 
 
Avramovic, a Serbian-born mother of a two-year-old daughter earned her master's degree in Belgrade before coming to the U.S. in 2009.  She chose Mason for its proximity to Washington and the quality of programs. Her research aimed to develop a methodology for pre-emptive rebooking of cancelled airline flights.
 
"I had no previous experience in aviation," said Avramovic, "but as soon as I started working on the flight simulator with Dr. Sherry, I was fascinated." 
 
The research studied the feasibility and benefits of preemptive rebooking of passengers on cancelled airline flights by addressing some key questions. For example, what would happen if passengers were given the opportunity to choose a rebooking time prior to their ticketed departure when they know their flight is cancelled? Who would benefit? What would the savings look like? How would this improve the passenger experience?
 
By using massive public data sets from the Bureau of Transportation Safety Avramovic discovered that pre-emptive re-booking is not only feasible, it is desirable. "Changes in technology and procedures have improved in airline passenger mobility," said Avramovic. "So many people are connected to the airline via email and smartphone apps, in a way that they weren't years ago. This communication between airlines and passengers increases opportunities for coordination."
 
Her research analyzed schedules, on time performance, and load factors. Avramovic said, "Airlines know in advance if there is a large scale event such as a snow storm, I wondered what would happen if passengers could decide what is the best for them: rebooking before or after the ticketed departure, instead of going to the airport only to find out that their flight is cancelled and that they have to wait for a later one?"
 
She is using a methodology developed at Mason that will estimate the delays and costs on a given day.  Her Monte Carlo testing engine is written in Structured Query Language (SQL) the special-purpose programming language. It allows a certain percentage of passengers to be rebooked on early flights when their flight is scheduled.  She applied the delay and cost methodology to investigate the effects of early rebooking. 
 
Avramovic has approached the Veteran's Administration about using this methodology in other contexts. "It is very exciting to think about the massive data sets that the VA has. While the air transportation has a great deal of data, the VA has even more."