CORVALLIS, Ore. - Engineers at Oregon State University and NASA have created a new system for air traffic control that could significantly improve congestion in the airways - a problem recently determined to be costing the United States economy up to $41 billion a year.

The system has been developed over five years of research, was recently adapted to make it even more flexible for voluntary use by air traffic controllers, and should be able to improve system wide performance by as much as 20 percent - a potential savings of billions of dollars and enormous amounts of frustration by travelers.

The concept is designed to leave control of aircraft in the hand of experienced controllers, but give them additional advice they could use at their discretion to improve the flow of aircraft on a regional and national basis. With some additional work the approach could be ready for its first test, researchers say.

"It takes a decade and billions of dollars to build a major airport, and with the growth of air traffic in the U.S. it's pretty clear we're never going to be able to build our way out of this problem," said Kagan Tumer, an associate professor in the School of Mechanical, Industrial and Manufacturing Engineering at OSU.

"What we can do is improve the efficiency of air traffic control, by giving controllers better information to make complicated decisions that benefit not just their airport but the region or nation as a whole," he said. "And our approach will do nothing to interfere with the safety of the existing system, which is extremely high."

Existing approaches, the scientists said, are absolutely safe - they're just inefficient, largely because it's not practical for a controller to be landing planes in Chicago while worrying at the same time about a weather delay in Kansas City, some mechanical problems in Miami and a growing bottleneck in Los Angeles. But sophisticated computer systems and monitoring devices using advanced algorithms developed in the new research can do exactly that.

"The technology may sound complex, but it's actually nothing more than sometimes telling aircraft to speed up or slow down to maintain certain spacing, or sometimes delaying a takeoff a few minutes, things like that," Tumer said. "This is already being done to some extent, but only on individual and local levels, not with an approach that rapidly considers changing conditions and new developments over entire regions or the whole nation at the same time. That's where advanced computer systems can help."

Right now, Tumer said, thousands of air traffic controllers at more than 5,000 public airports are making their best judgments on more than 40,000 flights a day with limited data and little in the way of a support system. The new approach would have aircraft constantly monitored by new technology, and make optimized suggestions that can instantly reflect multiple issues - congestion at one airport, bad weather at another, planes that have to be routed around thunderstorms.

Early studies were theoretically able to reduce airport congestion by 50 percent or more, the researchers said, but that presumed that air traffic controllers would take and implement every computerized suggestion. The new system is more realistic, allowing experienced controllers to have much more influence on the process and hopefully be receptive to what it can offer.

"In effect, we're trying to balance what computers will offer with the need for air traffic controllers to remain in full control of the situation," Tumer said. "The computer has to make suggestions that humans can live with and are not too radical a departure from existing approaches. But even having made those concessions, we think this new system could cut congestion by up to 20 percent."

In this system, a computer, in effect, will suggest "here's what I would do" as aircraft become bunched up on a certain line of flight. The air traffic controller can consider the recommendation and then do whatever they wish. The approach balances the interplay of humans and computers to improve overall performance, in an effort to transcend the problem that technology may have advanced more quickly than the trust that it will work.

"New approaches are difficult when something has always been done a certain way, and there's a suspicion the new technology may not work," Tumer said. "And it must be acknowledged that air traffic controllers have a remarkable safety record, which is not something anyone wants to jeopardize. Also worth noting is that nothing in this system would change the most critical aspects of air traffic control, at takeoff and landing.

"We have to be respectful of this existing expertise and work toward improved systems as a partnership."

Early work on this was supported by the Next Generation Air Transportation Systems Program at NASA, and new studies are funded by the Cyber Physical Systems Program of the National Science Foundation. The latest findings on this research project were reported last year in Advances in Complex Systems, a professional journal.

The stakes on these problems are high, experts say. Air traffic in the U.S. is expected to triple in the next 20 years. A 2008 analysis by the Joint Economic Committee of the U.S. Congress indicated that domestic air traffic delays in 2007 cost the economy $19 billion in increased operational costs for airlines, and $12 billion in lost time for passengers.

Air traffic congestion, experts say, now costs more than the average damage done by all the hurricanes hitting the U.S. each year.

Click photos to see a full-size version. Right click and save image to download.


Kagan Tumer, 541-737-9899