Bruce Benson 28/10/97
Abstract
Recently there has been much discussion in the aviation world about how to ease the increasing congestion in the skies and reduce the attendant danger of collision involved with air travel. As the demand for air travel continues to rise, the regulatory authorities are looking at ways to attempt to ensure that the incidence of mid air collisions does not rise in proportion with the congestion. The latest suggestion being considered involves abolishing the existing system of defined air corridors down which all the traffic flows. The benefit being that by allowing the airliners to "go their own way" they will naturally spread out, increasing their separation and reducing the collision risk. The problem with the "free flight" scheme is that it allows aircraft which would normally have been constrained to fly on the same heading as they tracked down the airway, to choose individual headings which may be convergent thus increasing the risk of collision. The mitigating factor that the authorities are playing on to get around this problem is the mandatory implementation of TCAS ( Traffic and Collision Avoidance System). TCAS watches the airspace surrounding the aircraft and assesses collision risks with other aircraft and alerts the pilot to any danger. In theory this is fine but it begs the question "where does the buck stop"? Is the controller still responsible for separation or is the aircraft commander or even the TCAS?
This report will explore:
Introduction
Free flight is not a new concept. Orville and Wilbur Wright were very familiar with it since there was no Air Traffic Control (ATC). They could take off whenever they wished and fly to wherever they wanted using any route and altitude that their creation would allow. Since there were no other aircraft on the same continent they could be sure of their separation. As the number of aircraft increased there was still very little trouble since they were well spread out and would not often arrive at the same airfield at the same time. Even if they did, the chances of them having independently arrived from the same direction was so small that control was not even thought of. Problems started when the aircraft became a form of public transport. With this came the advent of fixed routes between airports chosen because they were the most direct. As they were the best routes all aircraft used them, very often at similar altitudes, but still the idea of control enroute was unheard of. One incident changed that in Britain. One of the first mid-air collisions occurred between two British aircraft operating a cross channel route to Paris from an airfield near London. The route was very successful and in what must have been an inspirational moment they decided to run the London-Paris service at the same time as the Paris-London and the inevitable happened. On a misty day both aircraft were flying just under the low clouds on reciprocal headings and they met mid channel. This was the beginning of the need for modern ATC systems and the reason for the development of airways. Airways provide controlled corridors in which aircraft can be assured of their separation thus improving the safety of air travel.
Now that air travel has become a huge industry this "hands on" individual control is coming under attack particularly in the U.S., for causing delays due to being unable to cope with the traffic density and being uneconomical. Controllers are more likely to make mistakes and even if they don't, delays are inevitable at busy periods because of the mandatory separations. Controllers have to keep aircraft in holding patterns to stop dangerously high traffic densities at choke points and terminal areas above airports. In addition, the lack of direct routing costs the airlines vast sums of money in fuel and time.
The Federal Aviation Authority (FAA) in the states has accepted Free Flight as it's solution to these problems and have defined it as:
"A safe and efficient flight operating capability under instrument flight rules (IFR) in which the operators have the freedom to select their path and speed in real time. Air traffic restrictions are only imposed to ensure separation, to preclude exceeding airport capacity, to prevent unauthorized flight through special use airspace (SUA), and to ensure safety of flight. Restrictions are limited in extent and duration to correct the identified problem. Any activity which removes restrictions represents a move toward free flight."
In order to do make the new system work there is a need for much greater aircraft autonomy as regards collision avoidance. One of the technical requirements implied by this is the Traffic and Collision Avoidance System (TCAS). TCAS builds up a picture of the aircraft around it's host and uses sophisticated collision prediction logic to detect separation violations and generate a warning and an appropriate course of avoiding action which is verbally issued to the flight crew. Under the free flight system it is proposed that the pilots will be responsible for the separation of their aircraft in certain types of airspace which they can do with some degree of safety because TCAS will pick up any mistakes and issue instructions to rectify the situation. Or so the theory goes.
This report will discuss some of the arguments in the "current ATC system" versus "free flight" debate and look at the technology of collision avoidance to see what can go wrong.
Free Flight
A little history.
As early as 1979 free flight was proposed as a solution but it took many years for it to become technically feasible so that it could be seriously considered by the FAA. In 1994, the FAA asked Radio Technical Commission for Aeronautics, Inc. (RTCA), an independent organization that serves in an advisory capacity to the FAA, to form a select committee to study Free Flight. The committee that convened in October 1994 was composed of representatives from general aviation, pilot and controller unions, the airline industry, civilian contractors, and government policy makers. In January 1995, the committee presented a report to the FAA Administrator defining the Free Flight concept and the first steps for its implementation. Late in 1995 the FAA accepted it and were quoted as saying that it was to be the FAA's "next major goal".
How does the current system work?
Currently, a pilot establishes a flight plan with air traffic control. This plan requires the aircraft to fly along a specific route. Any deviation from the designated route must be pre-approved by an air traffic controller. For example, if a thunderstorm is encountered along the flight path, the pilot must notify the air traffic controller of the need to change course, and the controller would designate a course for the plane to avoid the storm. Under the Free Flight concept, the pilot will be able to choose the route, speed and altitude to achieve the desired results, notifying the air traffic controller of the new route.
How does it work?
Air traffic control is primarily about separation. The Free Flight concept approaches this at the aircraft level by defining two airspace zones, protected and alert, the sizes of which are based on the aircraft's speed, performance characteristics, and communications, navigation, and surveillance equipment. The protected zone, the one closest to the aircraft, can never meet the protected zone of another aircraft. The alert zone extends well beyond the protected zone and, upon contact with another aircraft's alert zone, a pilot or air traffic controller will determine if a course correction is required. In principle, until the alert zones touch, aircraft can manoeuvre freely.
Implicit in all this is the requirement for some technology to carry out the airspace surveillance and conflict resolution. This can be approached from two directions, either from the ground or from within the aeroplane. Ground based separation will rely on computerised collision probes which take all available information about aircraft positions, history vectors and declared intentions to produce a prediction about aircraft near future positions and from that to check for possible collisions. Additionally there will still be a human "separation manager" watching the computers and intervening if deemed necessary. Airborne surveillance is at present carried out by TCAS II which will become TCAS IV within the next few years.
TCAS
History
Beginning in 1975, The MITRE Corporation proposed a collision avoidance system concept to the FAA that would interrogate and listen to replies from existing aircraft transponders. This concept was adopted for development, and was named the Beacon Collision Avoidance System. Development work on this and a ground based system relying on digital data links with aircraft continued for the next 6 years. Then in 1981, the FAA Administrator terminated the development of ground-based collision avoidance, and announced that the data link technology was to be incorporated in the airborne system, henceforth named TCAS. At this point the data link became known as "mode S" transponder operation. After much testing in aircraft and some logic improvements to reduce the number of false reports, in 1987, the U.S. Congress passed an act requiring the FAA to equip all civil aviation air carrier aircraft with TCAS equipment by the end of 1991. At an October 1991 U.S. Congressional Investigations and Oversights Subcommittee Hearing "To Review The Status of the Airborne Traffic Alert and Collision Avoidance System " mixed reviews were received from pilots and from air traffic controllers. Not only had TCAS costs become prohibitively high to general aviation, but a number of unresolved technological issues persisted. Among the list of TCAS deficiencies were: sudden appearances of "ghost" or phantom images and high false alarm reports. New versions of the logic and software purport to reduce these.
How it works.
There are in existence two types of TCAS (I & II), a third was being developed but was overtaken by GPS technology and so was scrapped in favour of TCAS IV which will be able to exploit the enhanced accuracy of GPS. Leaving TCAS I and IV aside for a moment, TCAS II obtains position information from nearby traffic by actively interrogating their transponders in either mode C (altitude information is encoded with the reply) or mode S ( the data link capability) formats and deriving their relative position from the replies. This position consists of:
This position information is analogous to the position determined by ground based ATC, which makes it's own measurement of range and bearing using the same transponder reply formats. The fact that ATC is interrogating the transponders as well as all the other aircraft within range has been known to lead to overload and failure to supply the required information. TCAS tracks the position reports to determine potential collision threats, and issues it's pilot a Traffic advisory and a subsequent Resolution Advisory (RA) containing manoeuvring instructions in the vertical dimension, when a collision appears imminent. The bearing information is not sufficiently accurate to be used in threat determination or in choosing an RA. If the threat is also TCAS II equipped, the RA will be coordinated between the two aircrafts' TCAS units using mode S messages.
TCAS I is also in frequent use since it is substantially cheaper than version II and is now mandated for use in aircraft with between 10 and 30 seats in the US. It derives the same information in the same way but it cannot use mode S and gives only Traffic Advisories.
TCAS IV will use a GPS derived, position report broadcast to keep other aircraft informed of it's whereabouts. Addtionally it will transmit dynamic information such as velocity and heading , relieving TCAS of deriving that from a series of position reports. Finally, the aircraft can provide future intentions, such as heading or altitude changes. This would improve the TCAS' capability for collision avoidance, which hitherto could do no better than project the recent path of a threat assuming there would be no change. Additionally version IV is expected to give horizontal as well as vertical RAs.
What's wrong with it?
Like everything, TCAS is not 100% reliable, it suffers from all the usual radar/radio problems of interference, noise, reflections and in addition it uses some very complicated logic to decide whether an aircraft is a threat or not and if so what to do about it. These problems manifest themselves as sudden appearances of "ghost" or phantom images, false alarms, saturation of air traffic control radar frequencies resulting in limited densities of TCAS II units that can operate safely in this interrogation dependent fashion. Disturbing is the fact that some of these deficiencies appear during taxi, departure, approach, and landing phases where the system's integrity is vital to safety. This has created a serious problem, not only for pilots, but for air traffic controllers and with free flight relying on TCAS as the last line of defence before the pandemonium sets in it is not difficult to see why TCAS IV is in development. It should solve some of these problems because the number of interrogations required will fall due to more intelligent selection of targets, position information will be very accurate and far more reliable than the derived information currently available and new logic is being developed to reduce the number of false alarms.
TCAS is widely liked by pilots, especially with the latest version 6 logic, and is certain to play a large role in the future of collision avoidance with or without free flight. With free flight implemented it will form the final layer in the separation maintenance arsenal. The first and second being the separation manager and the ground based collision probes which will also benefit from the planned, more informative mode S broadcasts.
When will free flight be implemented?
So far the talk has been of things to come but there have been some small steps taken towards the implementation of free flight. Elements of the Free Flight concept are part of the current expanded National Route Program (NRP), the Central Pacific Oceanic program, and future communication, navigation, surveillance, and air traffic management technologies. The NRP expansion is designed ultimately to permit aircraft flying above 29,000 feet to select their own routes as alternatives to published, preferred IFR routes, thereby removing the restrictions and constraints currently imposed on these users. The NRP expansion is successfully using a phased approach. As of June 1, 1995, the expanded NRP includes all flights above 35,000 feet. In the western U.S., the level was lowered to 33,000 feet. The FAA estimates that the NRP saved the aviation industry $40 million in 1994 by allowing pilots to fly more optimal routes.
In the airspace over the Central Pacific, advanced satellite voice and data communication is being used to provide faster, more reliable transmission to enable reductions in vertical, lateral and longitudinal separation, more direct flights and tracks, and faster altitude clearances.
Will it work?
On first acquaintance the free flight scheme may seem like sanctioned mayhem or as one respected pilot put it:
"Picture a large metropolitan area with thousands of automobiles clogging the highways. Although there are occasional accidents and the red lights cause delays, the flow of traffic is fairly orderly. Now, remove all the guard rails and barriers, pave the entire city, and allow anybody to go in any direction they please. Pandemonium would result."
Joe Benkert online
On closer inspection this analogy does not hold up very well. The "city" referred to represents the Terminal Manoeuvering Areas (TMA) above major airports and these will not be within the free flight system since to allow that would result in pandemonium. They will remain as controlled airspace and will extend out up to 50nm as required by peak loads, to allow time and space to manoeuvre all the joining traffic ready to be fed into the runway. Perhaps a more accurate way of using the car analogy would be to describe free flight as paving over the land between cities so that it would be possible to drive in a straight line between two locations once outside the city. Then consider the vertical dimension, there is a lot more room in three dimensions than there is in two. All of a sudden it doesn't seem so crazy.
The problems start to come back when one considers that there will be several aircraft all wanting to use the optimal route and height between two airports at very similar times. How are they regulated? Who gets to fly in the airspace that they wanted and who is forced to take another route? The solution to this may lie in the fact that the aircraft cannot all take off at the same time and so will have their initial separation enforced by their differing take off slots.
Another fundamental tool for Free Flight is the pre-filled flight plan. The requirement to file a flight plan will not be removed, infact it is even more important to free flight than it is to the current system since it is the main way that the people on the ground can tell what the pilot is going to do next. It also allows the predictable airspace conflicts to be resolved before take off as they are today but using integrated Traffic Flow Management (TFM) systems. Having had the flight plan approved the pilot will be expected to stick to it or request changes as necessary, free flight is not about random actions just random paths.
The arguments for free flight under normal, fully functional circumstances are powerful and hard to beat. The heated discussions are now arising over what happens when something goes wrong. A classic and highly plausible scenario is a full communications failure in an aircraft that is in free flight airspace, has just been authorised to change altitude and is heading towards a cross stream of peak time traffic about to enter a TMA. This scenario was put to a panel of four air traffic controllers and a pilot to see whether they could find a solution as part of a study into the human performance aspects of free flight ATC. The researchers wanted to explore whether, when and how much a human should intervene to prevent a collision. It brought up some important questions to do with responsibility and information dissemination. The panel found that the biggest problems they faced were:
Without being able to talk to the pilots or use the aircraft data link there was no way of assessing the damage or to ask what the pilot wanted to do. The only other possible method of obtaining aircraft information was through the airline's dispatch communications which may have been able to tell if the aircraft was maneouvering, but this would have required a lengthy telephone call which they were unhappy to spend time on.
Under the current ATC rules the pilot would be obliged by law to follow a certain course of action which would let the controllers deal with the situation since they would be able to predict where he would take the aircraft. Under free flight they were unsure as to whether he would follow his flight plan and continue to his destination ( which was the airline's home base) so that he could get the aeroplane fixed or whether he would follow the prescribed course of action and land at the nearest suitable airfield. Hence they could not decide whether there would even be a problem.
Had they been able to ascertain that the TCAS was working properly they would know whether the pilot was well enough informed to be responsible for his aircraft's separation, whether other aircraft would be able to see it and whether they should intervene to enforce separation. If they did intervene what should they do? A question they found very hard to answer.
Intervention itself is a tricky problem and at different times throughout the discussion various proposals were offered and rejected because they all had serious complications. They ranged from taking control of all of the incoming stream and routing them around the problem aircraft to merely issuing a traffic advisory to the stream of the last known position, altitude and heading of the stricken aircraft and hoping that the TCAS would be able to advise each crew in the stream as to what to do to avoid it. The former option was regarded as requiring far too much of the sector controller and the latter solution as possibly too little too late.
None of the possible solutions were satisfactory because the information required to solve the problem was distributed between the sector controller, the aircraft and the aircraft's dispatch. The issue of who is responsible could not be ascertained because responsibility was found to go hand in hand with the information distribution. As one of the controllers put it:
"If I'm responsible there's the caveat that I need to know what's going on. "
From this example it can be seen that there are many questions to be answered on the road to free flight and that for the system to handle exceptions such as equipment failures there need to be multiple independent communications channels and revised emergency procedures to put some predictability back into the situation. Separation is the aim and communication is the means to achieve it safely.
Conclusion
The benefits of free flight are many, economic, greater capacity, better safety, lower controller workload, fewer delays and less bureaucracy. Most of these are beyond argument however people still disagree on whether the concept can actually be turned into a working reality. The FAA readily accepts that the road to full implementation is long and tortuous and they don't have all the answers yet but they have reviewed the proposal at great length and believe it to be the best solution. The further expansion of the current system is not a viable option since it is already reaching the natural limit imposed by the size of the airways and the minimum separation permissible between aircraft. Under the present rules the next generation collision avoidance systems, envisioned as arriving within three years, will not be able to significantly increase the number of aircraft that can be fitted into the current system. What is needed is a completely new approach which turns out to be a return to the ways of old, albeit with an absolute requirement for a vastly superior arsenal of collision avoidance technology to make sure that everything stays apart. To enable this technology to work to it's best advantage, new higher capacity data links need to be implemented so that aircraft can tell each other and the ATC computers exactly where they are (based on GPS information), where they are going in the next two minutes and ask each other how they want to resolve the impending collision. What to do if it all goes wrong is something the powers that be will be arguing about for some time to come, but free flight is slowly on it's way.
Total hitz: