This article is reproduced with the permission of New Scientist for exclusive use by Nova users.

Is it time to give airliners the freedom of the skies?
13 July 2002
From New Scientist Print Edition.
Gerry Byrne, Dublin

Enlarge Free flight for all

Last week's mid-air collision over Germany might have been prevented if the pilots had been allowed to choose their own routes. That's the view of aviation experts, who for years have been campaigning for a radically different air traffic control system called Free Flight, in which aircraft would be able to use the vast tracts of airspace outside today's fixed air corridors to get from A to B.

Free Flight cocoons each aircraft in a protective virtual "bubble". This is designed to give aircraft advance warning of any possible collision. As a result, planes would be able to fly more direct and fuel-efficient routes without continual interruption by air traffic controllers on the ground. The US Congress says it makes economic sense to free up unused airspace in this way and the Federal Aviation Administration is committed to implementing it. But its adoption is piecemeal and, for Free Flight's backers, agonisingly slow.

Today's air traffic control (ATC) procedures funnel aircraft into narrow air corridors. Planes often have to wait for a slot to get into an air corridor, and for safety's sake they can only do that at intervals of 5 nautical miles (9.25km). Planes kept strictly separated from each other in this way can't possibly collide, runs the reasoning. But ATC doesn't always get it right, as shown by last week's collision and the steady succession of near misses around the world. And even when the system is operating safely, its many critics say that it leads to unnecessary congestion and delays.

Critics also complain that air traffic controllers have too much power in the cockpit. In effect, the controller is the plane's navigator, deciding an aircraft's route, height and speed. Pilots have to follow navigation commands blindly except in the case of an imminent collision when pilots are supposed to follows instructions from the aircraft's collision avoidance system (TCAS). But in the crash over Germany, voice recorders show that a controller countermanded instructions from one plane's TCAS.

This could never have happened with Free Flight, says Michael Baiada, the United Airlines pilot who championed the idea. "The current system creates a massive conflict at the merge point of two jetways," he says. "With Free Flight, the conflict between the two aircraft would have been detected 20 minutes earlier and their courses changed."

It's now eight years since Baiada first persuaded Congress of Free Flight's benefits. In Free Flight, much of the navigational control is restored to the cockpit, leaving pilots free to abandon traditional air lanes and choose the most favourable route according to the weather, distance and altitude by using the huge amount of airspace outside the air lanes. It should reduce take-off delays and cut fuel consumption, saving airlines billions of dollars. And tests show that the computerised route planning that is a cornerstone of Free Flight is likely to make it safer than today's system, which relies on overworked ATC staff.

And the burden on the current system is only getting worse. In the past 20 years the number of airline flights in the US, Europe and Asia has grown by 72 per cent- and it's not slowing. As a result, the International Air Transport Association, the umbrella group representing airlines, sees a crunch coming. "In 2012 we will hit a capacity wall and current methods of air traffic control will not take us beyond it," says Steve Zerkowitz, IATA's air traffic control analyst.

It's a scenario that means controllers become increasingly stressed. In Britain earlier this year reports of "overloads" in air traffic control centres, with too many planes coming into a controller's airspace in too short a time, have more than doubled over the last year.

In the present ATC system, planes fly along air corridors marked by ground-based radio beacons, and are kept at a safe distance from each other - five miles apart horizontally and 1000 feet vertically - by air traffic controllers who can see at a glance not only which aircraft is which, but where it is and where it is heading. They can do this thanks to a device called a transponder that is fitted to all modern aircraft. When triggered by a signal from the ground or another aircraft, the transponder broadcasts or "squawks" its identity, height, course and speed to the local ATC centre and cockpit radar screens on nearby planes.

In a Free Flight scenario, this would all be handled by a network of air traffic computers. Before a flight, the pilot would enter the chosen flight plan and estimated journey times into the network. Using a program called a conflict probe, ground-based software sets aside a 100- to 200-mile long bubble of airspace in front of the plane and calculates if it will meet another aircraft's bubble during the flight. If so, the computer alerts whichever aircraft is due to take off next and asks the pilot to adjust its flight path. Controllers watching on radar only need to intervene if the bubbles of two aircraft threaten to meet.

But this is where the problems start. Running Free Flight is a huge task for any computer system. Although Baiada says the software is relatively straightforward, the hardware that would be needed for the central route planning and conflict computer is not. Reliability would have to be total. Back-ups, and back-ups for back-ups, would be vital.

But Baiada is encouraged by a Free Flight test in the airspace around Malmo, Sweden, in 1999. The test, carried out by Eurocontrol, the pan-European air traffic control organisation, proved the basic concept of the conflict probe idea and showed that controllers and pilots can be comfortable with the system. Eurocontrol is planning similar tests in Italian and Dutch airspace next year. But it's still early days for Free Flight. The FAA, which controls US airspace, says the system has no chance of widespread implementation until key supporting technologies are in place, such as a text-based cockpit-to-ATC communications system and more accurate navigational aids.

Baiada, however, insists that current technologies are adequate for Free Flight and parts of the Free Flight idea are already becoming reality. From this summer, for example, pilots can request direct routes when flying over Europe at altitudes above 29,000 feet. And in the US, the FAA has installed conflict probes in several ATC centres to manage ad hoc direct routeing requests.

But this piecemeal style of introduction angers Baiada. "The FAA has spent billions on this but if you ask me am I flying any differently now than 10 years ago, the answer is no," he says. "I want to be able to pick my routeing, navigate accurately according to my requirements and I want the controller to stop me hitting anybody. I don't need them to navigate for me, I can do that myself."

Not everyone agrees, however, that giving pilots this autonomy would bring any major benefits. For example, Eurocontrol simulations suggest that Free Flight would reduce flying distances by only about 5 per cent. "We expected the gain would be greater," says Alex Wendels, a programme manager for air-to-ground communications at Eurocontrol. He accepts that the ATC community is traditionally hesitant when it comes to taking up new technologies, but he insists Eurocontrol is not being difficult or looking after its own interests, Free Flight simply doesn't deliver as much as initially suggested, he says.

Opponents of Free Flight also raise the issue of deciding what to do with all the aircraft as they descend and enter the increasingly congested airspace around an airport. "Nobody has yet come up with an answer for terminal congestion and I suspect the reason is because there isn't one," says Rod Fewing, a researcher at the Air Transport Group at Cranfield University in Bedfordshire.

Baiada doesn't agree. One reason for the amount of congestion around busy airports is that there is no control over the numbers allowed to arrive in the terminal airspace at any one time, he says. "Nothing happens until everybody shows up all at once, yet they should have known this was going to happen two to four hours in advance. A pre-sequenced arrivals flow would solve that problem."

With this in place, planes could then tailor their departure and journey times to arrive on a "just-in-time" basis, precisely when they are expected. But it will also mean airlines will have to be sensible about take-off times, and not insist on only the "marketable" slots on the hour or half hour.

Nonetheless, Free Flight is unlikely to be adopted any time soon, especially in Europe. On top of the delays in proving the complementary technologies, British airlines are showing less interest than their American counterparts. Seven of them, including Virgin and British Airways, hold a 46 per cent stake between them in Britain's part-privatised air-traffic system, NATS. British Airways says it will leave the debate over such matters to the experts. Free Flight doesn't yet form any part of future plans at NATS. "Strategic trajectory planning is likely to remain a ground-based function controlled by air traffic controllers," says NATS's spokesman Richard Wright.

Despite last week's crash, ATC-controlled flight has a good safety record. But its critics say that its philosophy of waiting until conflicts arise before solving them doesn't make sense. Free Flight's conflict probes are designed to forecast an aircraft's position and decide what action to take up to 20 minutes, or at least 100 miles, ahead so that aircraft are separated in plenty of time. That said, traffic conflict alerting systems are already installed in European air traffic centres (see "How planes avoid each other").

On the other hand, ATC-controlled flight could be accused of having a lower margin of error by allowing aircraft to come within 5 miles of each other in the normal course of events. But its supporters say it can assure safety even when some of its essential systems, such as radar and computers, fail.

This was demonstrated in the skies above England in May when every main computer system failed in the en-route ATC centre at Swanwick, near Southampton. No mid-air emergencies ensued. American controllers in older centres routinely endure radar failure without mishap. And in Ireland, controllers in Dublin once famously kept busy air traffic moving with a walkie-talkie set when their VHF radio caught fire, forcing the evacuation of their control centre. In such eventualities, controllers fall back on long-established flow control methods to space traffic even further apart than normal.

But such emergency techniques work only with traffic which is already separated and confined to narrow, predictable airways, not spread out across a broad airspace. Baiada says Free Flight would cope in these circumstances thanks to the distributed nature of its computers, sharing functions around the network, and that the 20-minute separation between aircraft gives a far wider safety margin.

In the end, most observers predict a gradual improvement and upgrading of air traffic control equipment, rather than a quick and radical switch to Free Flight. Baiada, meanwhile, is quietly despairing. He feels that his vision of a quick and easy change in the way air traffic is managed is being obscured by bureaucracy, inertia and what he sees as unwarranted concerns that it isn't safe.

"Traffic increases over the next 10 years will eat away at any efficiencies we gain from new technologies," he told New Scientist. "What we are doing now is a waste of time."

From issue 2351 of New Scientist magazine, 13 July 2002, page 12

For the latest from New Scientiist visit www.newscientist.com