 |
 |
 |
 |
| |
|||||
|
|
|
Home > News & Events > MITRE Publications > The MITRE Digest > | |||||||||||||||||||
| Travel Forecasting Takes Off March 2008
Planning to fly from New York to Los Angeles seems like a simple enough proposition. However, there's an array of factors for the traveler to consider. For starters, in the New York metro area, it's possible to depart from one of at least three airports, while in L.A., there are no less than five airports in the region, all of which support varying degrees of air traffic and passenger travel. As for the flight itself, is there a non-stop route? If not, which hub airport should be chosen to make the connection: Chicago, Dallas-Fort Worth, or Phoenix, for example? Some travelers take the type of aircraft being flown into account. And perhaps the most important issue of all is timing: the departure and arrival schedule can make or break a plan. "It's these kinds of variables that make air traffic forecasting for the federal government a bit complicated," explains Jacqueline Kee, principal investigator at MITRE on an innovative project that seeks to foretell air traffic by digging into data on local demographics and economies, historical traffic flows, and flight schedules. In the U.S., the National Airspace System (NAS) is constantly evolving. While low-cost carriers challenge the business models of the major airlines, the future of current hubs, schedules, and airport regulations are also in flux. To address this uncertainty surrounding takeoffs and landings, MITRE engineers have devised a means of forecasting to help the Federal Aviation Administration (FAA) better understand how air traffic projections will differ with varying assumptions about local economic growth in emerging cities, population changes, and airframe technology. Known as the Future Air Traffic Estimator (FATE), the tool is the centerpiece of a project that examines underlying passenger demand for air transportation. "We used data from the Department of Transportation (DOT) to start building a model to better forecast passenger flow through the NAS," says Dipasis Bhadra, a principal economic and business analyst at MITRE who created the original FATE tool four years ago. "While the airlines do this for their own routes, there is no aggregation at the NAS level. Demands change over time, and the FAA wants to be ready to meet those demands, both regionally and nationally." Since Bhadra created FATE, a team from MITRE's Center for Advanced Aviation System Development (CAASD)—the federally funded research and development center (FFRDC) that MITRE manages on behalf of the FAA—has been further modifying it. The other FATE team members who lend their complementary expertise to the project include Jennifer Gentry, a lead economics and business analyst; Gene Lin, a senior simulation and modeling engineer; and Lorrie Smith, a senior software application development engineer. They have made FATE more modular for wider use, which has now been incorporated into other MITRE projects. FATE originated within MITRE's research and development program. The program seeks both to create new technologies and to apply existing tools and technologies in innovative ways to deliver value for customers. Who's Flying Now (and Later) Bustling airports are a way of life in America's largest cities, fueling the nation's economy. FATE's analytical framework incorporates city of origin data in order to calculate system-wide use. "FATE examines the characteristics driving domestic air travel demand from one metropolitan area to another," Kee explains. "It considers local economics and income, and statistically determines the sensitivity of the demand to the underlying drivers in order to forecast demand in the years ahead." The three main elements comprising FATE are passenger forecasting, flight forecasting, and time schedules, which are implemented using Visual Basic software and Microsoft Access programming. But that doesn't reveal the whole package. Peering under FATE's hood reveals an innovative mix of software design, economics forecasting, and statistical models. The tool merges passenger ticket information by origin/destination pair, which is data the MITRE team obtained from government websites, along with local economic and demographic forecasts purchased from Global Insight, a forecasting company that provides comprehensive data on countries, regions, and industries. For example, to analyze journeys between Chicago and Washington, D.C., the tool delves into how many passengers are most likely to travel, and allocates those passengers to the origination airports such as Chicago's O'Hare International or Midway International, and to the various destination airports, like Washington's Dulles International, Ronald Reagan National, or Baltimore/Washington International. "FATE also analyzes exactly how they will get from Chicago to D.C.—such as via nonstop flights or by connections in other cities like Cincinnati, Detroit, or Philadelphia," Kee says. The tool then assigns passengers to aircraft—whether it's a narrow-body jet seating 130 people, or a wide-body accommodating 200—and factors that into the projection of the number of flights. "We also developed different algorithms for forecasting international and general aviation traffic," Kee adds. Each forecasted flight is also assigned a departure and arrival time. "FATE is designed to produce detailed flight schedules," Bhadra points out. "So, if you want to know how the National Airspace System will look in 2010 during a particular quarter, or even weekdays versus weekends, it can generate those schedule projections. Those schedules are then used for NAS simulations for specific times in the future." A Twist of FATE FATE doesn't just produce numbers, however. It also contains a graphical interface, so users can compare the forecasts with historical data, and explore the effect of model assumptions. "In 2006, we made the tool more flexible so users can change assumptions and generate different scenarios, which is valuable to analysts," Kee says. "We also ported it to a new, more user-friendly interface." In 2007, the MITRE team continued improving the quality of the economic models and evaluated the impact of differing assumptions on the forecasts. "FATE has allowed us to gain insights into the main drivers of air traffic—both passenger and aircraft traffic—by market," Bhadra says. "This approach ultimately lets us estimate passenger demand, and even respond to changes in NAS policy." FATE's modeling scenarios can be used by government agencies to see how passengers typically move through the air system. "This data helps with decision-making, whether it's installing new runways or taking other measures to alleviate congestion," Bhadra says. "FATE even gives our sponsors an idea of how things might evolve in a certain situation, such as the spread of disease amongst the traveling population." The tool has already delivered value to customers. FATE forecasts have been referenced in the FAA's capacity needs section of the NAS analysis, which takes a frank look at where capacity "hot spots" will emerge. The FAA's Office of Aviation Policy and Plans has also expressed interest in FATE's origin-destination passenger demand forecasting methodology. Sharing Knowledge Inside and Out Meanwhile, the MITRE team has published its ongoing research results in peer-reviewed publications such as the Journal of Air Transportation, Journal of Aircraft, Journal of Air Transport Management, and Journal of the Transportation Research Forum. Similarly, they have presented their work at industry conferences. Information about FATE was included in a document by the DOT and FAA about assessing the country's future needs for airport capacity. That document, entitled "Capacity Needs in the National Airspace System," was circulated to Congress and can be found on the FAA's website. Ensuring that FATE meets users' needs has required a great deal of collaboration between MITRE and its sponsors, and that same spirit of collaboration is also evident within the corporation. "Anyone at MITRE may access the FATE forecasts, because the results are available via CAASD's Data Repository," Kee declares. "Someone, for example, in MITRE's Center for Enterprise Modernization FFRDC can look at FATE's forecasts and perhaps adapt it to an aviation-related project." —by Cheryl B. Scaparrotta Related Information Articles and News
Technical Papers and Presentations
Websites |
||||||||||||||||||||
| Page last updated: March 25, 2008 | Top of page |
Solutions That Make a Difference.® |
|
|
