Researchers develop equitable flight scheduling model

by Charles Chen | 4/11/18 2:00am

Airline congestion is a perennial headache for airlines and millions of travelers around the world. Moreover, it cost the U.S. approximately $30 billion in 2007. However, a new study published in the journal “Transportation Science” by Thayer School of Engineering professor Vikrant Vaze and Carnegie Mellon University professor Alexandre Jacquillat sought to remedy these expensive delays by applying game theory to flight scheduling models at airports.

This study, titled “Interairline Equity in Airport Scheduling Interventions,” is the first-ever to address both efficiency and equity considerations in its proposed flight scheduling model, which the researchers named the “Integrated Capacity Utilization and Scheduling Model with Equity Considerations.” Previous models that focused solely on efficiency inadvertently prioritized the flights of certain airlines more than others, which creates interairline inequity, according to Vaze.

“We wanted to show how we could include other constraints beyond what has been done so far, in particular that of equity,” Vaze said. “[This is] to explicitly ensure every airline — and as a result every passenger — is treated fairly.”

Jacquillat and Vaze applied their model to computer-generated scenarios at one of the busiest airports in the world — John F. Kennedy International Airport in New York City. According to the study, the model was able to equally prioritize different airlines’ flights with little to no loss of efficiency.

According to Jacquillat, recent advances in technology have enabled researchers to develop optimization models that address air traffic management and congestion. However, he noted that these developments might also fuel inequity between airlines.

“There has been growing recognition of the equity concerns often created by [these] scheduling models,” Jacquillat said.

Vaze said that airlines commonly complain about inequity when scheduling models are proposed, which is a barrier to the implementation of congestion-easing schedules. Although Vaze said that he plans on further refining the scheduling model, several airlines and airports have already expressed interest in the study.

Massachusetts Institute of Technology chancellor and professor of engineering Cynthia Barnhart, who was credited in the study for providing suggestions and comments, praised the research model because of its ability to appeal to the concerns of regulators, airlines and passengers.

“The solutions that [the researchers] generate are solutions that each stakeholder would like,” Barnhart said. “With this sort of win-win-win, the chance of being able to implement these solutions is much increased.”

Jacquillat said that while many European airports already have scheduling mechanisms in place that need only slight modifications to achieve inter-airline equity, scheduling models in the U.S. might be in greater need of overhaul.

The study noted that their model could be applied more broadly to other industries that involve high-demand systems with finite capacities, such as railways and telecommunications. Shipping companies and healthcare systems that face similar constraints could also benefit from their scheduling model, according to Jacquillat.

“We believe the technology and policy designed in this paper can have broader impacts,” he said.

Vaze, who has spent a decade researching systems optimizations, added that his collaboration with Jacquillat allowed the paper to reflect each of the researchers’ own specializations. Vaze said that while he approached the research question from an airline-centered perspective, Jacquillat brought knowledge of airport operations to the study.

“This collaboration has been very fruitful because we tend to think of different sides of the same coin,” Jacquillat said. “That results in an output that is greater than the sum of its parts.”

Next, Vaze said that he plans on studying how regulators can prevent airlines from “gaming” scheduling models like theirs to obtain desirable flight schedules.