FAA Rolls Out Revamped Air Traffic Forecasts Based on MITRE-Developed Model

July 2014
Topics: Aviation Industry, Air Traffic Management, Aviation Administration, Airports, Modeling and Simulation
Factors affecting aviation traffic levels and patterns change frequently. That means the Federal Aviation Administration must adjust its forecasts accordingly. To enhance its predictions, the FAA now uses a model with roots in MITRE's research program.

Do you ever wonder how the Federal Aviation Administration estimates just how much air traffic to expect—and where—in the coming years? Put in simple terms, the FAA combines large amounts of data, sophisticated modeling techniques, and in-depth analysis.

That's one reason why the FAA looked to MITRE for help in improving its approach to forecast future passenger demand, air traffic flow volume, and patterns across the National Airspace System.

The work began as the Future Air Traffic Estimator (FATE) capability, a research project MITRE conducted on behalf of the FAA more than a decade ago. We transferred the technology to the FAA in 2008, and it's now part of what the agency calls the Terminal Area Forecast Modernization, or TAF-M.

MITRE staff played a key role in refining the technology and readying it for FAA implementation. "Most recently, we assessed how the current TAF-M forecasts compare to those from an earlier version of the methodology," says George Solomos, a department head in MITRE's Center for Advanced Aviation System Development, the MITRE-operated FFRDC sponsored by the FAA. "We also performed analyses to determine how the FAA could make the forecast even more accurate. Passenger route choice, airport choice, and the impact of certain input assumptions—such as economic growth—are some of the factors we helped the FAA study so it could improve upon the forecasts."

The FAA used the latest version of TAF-M to produce an enhanced forecast, which it officially rolled out in March at the Federal Aerospace Forecast Conference. Sponsored by the American Association of Airport Executives, the conference brought together aviation industry stakeholders, analysts, economists, and government representatives to discuss the FAA's forecast for 2014–2034 and the analyses behind the predictions. (See "How the Tool Works," below.)

TAF-M's Beginnings

MITRE analysts planted the seeds for this new forecasting methodology 10 years ago. They saw the potential for expanding both the data sources the FAA used to produce the forecast, as well as the scope of the forecast itself.

"The FAA used to look at arrivals and departures from airports and make forecasts about air traffic and passenger demand based on that information," says Dr. Dipasis Bhadra, the former MITRE principal economic and business analyst who created the FATE tool in 2004. Bhadra later joined the FAA and has headed the technology's implementation effort since 2008.

"Because MITRE's modeling, simulation, and analysis work addresses airspace management, we wanted to process those airport-specific forecasts to understand how complex the airspace would be," Bhadra recalls from his early days on the project. "So we sought ways of determining what routes passengers would take, whether or not there would be stops along those routes—and if so, where—and the aircrafts' specific trajectories."

As a beginning, Bhadra considered information the Department of Transportation published quarterly indicating how passengers traveled between any two airports. "That enabled us to look at the flow of the system."

After developing flow forecasts based on that new data, CAASD analysts took their research a few steps further. "We demonstrated that, with this information, we could identify the probable type of aircraft that would be used for any given flight, so we were able to map passengers to aircraft type," he says. The MITRE team also tied flights to timetable information. "We identified where the peaks and valleys were in each airport's arrivals and departures and integrated that into FATE as well."

Later work on FATE added even more data to enhance the integrity of the forecasts. In addition to historical traffic flows and flight schedules, researchers incorporated data on local demographics and economies into their calculations. This helped the FAA better understand how traffic forecasts would differ with population changes or economic growth or decline in a particular city or region.

More recently, CAASD developed models for forecasting international and general aviation traffic as part of the TAF-M calculations.

A Forecast that Meets the Needs of Many

At this year's Federal Aerospace Forecast Conference, the FAA rolled out its forecasts for 141 of the nation's 515 commercial airports. "We have forecasts for all of these airports, and we are hoping to roll out the remainder later this year," Bhadra says. "We staggered the forecasts because they affect so many industry stakeholders and decision-makers. We want to make sure all the numbers have gone through a thorough check before we release them."

Many organizations within and outside of the FAA use the forecast for a variety of purposes. For example:

  • The FAA's Air Traffic Organization uses the forecast for workforce staff planning, such as calculating how many controllers it will take to meet demand.
  • The FAA's Office of NextGen uses the forecast to help understand the current and potential impact of NextGen improvements at different airports.
  • The FAA's Office of Airports incorporates the forecast into its calculations as to whether certain airports will require more capacity in the next 5 to 15 years.
  • The FAA uses the forecast as an input to make decisions on approving federal funding for airport infrastructure improvements.

"At MITRE, we use information from the forecast as one of the inputs for generating the schedules we use in our simulation models of the NAS," says Felipe Moreno-Hines, a group leader in the Performance and Economic Modeling and Analysis department.

The model is proving to have benefits outside of aviation as well. Solomos explains that MITRE is now working with the Department of Transportation to apply the TAF-M methodologies to determine what influences freight-rail route choices.

"It's very gratifying when something we've developed in our internal research program has been successfully transitioned and becomes a vital part of a sponsor's capabilities," says CAASD chief engineer Glenn Roberts. "In this case, the technology we developed is helping us to support not only the original sponsor but is informing our work for others as well."

—by Marlis McCollum


Publication Search