Slashdot Mirror
Inside FAA's GPS-Based Air Traffic Control
longacre writes "With the growing number of planes in the air setting its archaic radar-based air traffic control on a course toward 'total system collapse,' the FAA has quietly begun testing a new GPS-based system on Alaska Airlines 737s. While radar can take over half a minute to determine a plane's location, GPS technology known as ADS-B broadcasts an aircraft's position to controllers and nearby pilots essentially in real time. If all goes as planned, travelers will see fewer delays as planes will be able to fly closer together and in reduced visibility conditions, and airlines will achieve significant fuel savings by flying more direct routes. The feds plan a gradual rollout over the next two decades that may cost up to $40 billion." There's still some contention about where the funding will come from.
LIke said above, using 3 sats. Calif altitude is identical to all others, x feet above sea level. Terrain avoidance is done by on board GPS units tied to maps, as well as radar altimeters. This is not so much about avoidance of terrain but avoidance of other aircraft.
"If the King's English was good enough for Jesus, it's good enough for me!" -- "Ma" Ferguson, Governor of Texas (circa
Delta owns the following:
a tion/delta_stats_facts/aircraft_fleet/
71 Boeing 737s @ $50 million per
68 Boeing 757s @ $65 million per
75 Boeing 767s @ $140 million per
8 Boeing 777s @ $200 million per
63 MD 88s @ $40 million per
16 MD 90s @ $45 million per
68 CRJ 100/200/700s @ $24 million per
that brings this one airline's fleet cost to just about $25 billion. And I was giving the low estimate for the cost of the planes.
http://www.delta.com/about_delta/corporate_inform
"Can anyone offer some financial perspective.."
c le/2006/08/17/AR2006081701485_pf.html
I don't know about airplane costs, but here is some perspective on other government upgrade projects... each upgrade involved both hardware and software systems.
The IRS attempted to update their systems (originally designed in 1962). The project began in 1999 and was spread over several 'projects.' The 1999 plan was eventually scrapped after the main database was already around 40 million over budget and way over deadline. Further attempts to modernize the system in a more compartmentalized fashion lead to the $318 million lost due to excessive tax refunds in 2006 (for tax year 2005 returns). The system responsible was also scrapped and the old one was put back into service.
Though not mentioned in the overview that I link below, a GAO report I saw a couple years ago put the total actual losses (internal/external/disaster recovery etc...) at several times the publicly reported loss numbers.
Here is a general overview: http://www.crn.com/it-channel/192502071
The FBI attempted a complete systems overhaul (agents still can only use one search term in many of their databases, and much info is still paper file only). That was finally scrapped in 2005 after $170 million in costs, and over 170,000 lines of code... the project had been in progress for three years. The Washington Post put total upgrade costs since 9/11/2001 at around $600 million.
Here is a general overview: http://www.washingtonpost.com/wp-dyn/content/arti
I suppose the successes don't make as much news as the failures, but the real perspective we should be looking for here is who the F*** is going to plan and manage a project that will be responsible for our air safety? The upgrade attempts that I know of all ended with the old systems being put back into place.
NASA probably gets my vote. I have heard that their software design requirements and beyond insane, and that despite the catastrophic structural failures they have endured, the shuttle software systems are beyond rock-solid. They still use multiple levels of 'readers' to proofread every line of code as you would a thesis manuscript in addition to all other testers/unit-tests/sims etc..
Regards.
Mode-S a very nifty datalink system that uniquely identified aircraft and can beam all sorts of useful traffic and navigation information. It was designed *WAY BACK* in 1975, only to be ignored by the FAA (the airlines the FAA works for didn't want pay for it). So they ignored it until a mid-air collision in 1986 woke up Congress, who mandated it in 1993. ADS-B (the Popular Mechanics article seems to be describing) AFAIK uses Mode-S to broadcast your aircraft's position using Mode-S, but the FAA have started shutting down Mode S transmitters 'because the safety benefits are not worth the cost'. Nice idea, but I hope it doesn't take another costly "wake up call".
y .html http://www.aopa.org/whatsnew/air_traffic/tis.html http://www.aopa.org/whatsnew/newsitems/2005/051020 mode.html http://www.avionicswest.com/myviewpoint/modestrans ponder.htm
http://web.mit.edu/6.933/www/Fall2000/mode-s/toda
Lots of technogibberish here: Hey, Wiley! When are you writing "Air Traffic Control for Dummies"?
Technically, you'll need four or more satellites in view to calculate altitude. In general, you'll need (n+1) points of reference (satellites) to triangulate a point in an n-dimension space. (Assuming you want the position calculated in all n dimensions).
g ulation
In practice, three satellites are adequate for ground and altitude calculation (since most other spatial possibilities can be ruled out as being 'ridiculous').
For more info: http://www.beaglesoft.com/gpstechnology.htm#Trian
No you're not missing anything. The altimeter provides the altitude readout and it's just sent as a "sentence" along with the other information. Altitude data is already encoded from the altimeter and sent to the ground based radar as part of a coded signal from a transponder in the aircraft. This has been the way for decades. There is no need to get altitude from GPS. Position data from GPS is another thing. It is theoretically more accurate than a radar position, but there a bunch of innaccuracies that have to be built into a 'tolerance" that has to be applied to the aircraft position as reported.
TCAS is a traffic collision avoidance device also in use today that transmits altitude data between aircraft. Again the data comes from the altimeter.
Automatic Dependant Surveillance (ADS) data provides position (from INS or GPS) and altitude from the altimeter. The data can be sent via radio link or satellite. The amount of times per minute (or hour) that this data is updated to the ground station provides the basis for seperation of aircraft. If you update quite often you can run planes just a few miles apart. If you update every thirty minutes or so by expensive satellite links (trans-ocean) you might have to run the aircraft 100 miles apart. Some of the cost is in the aircraft but much of the cost is in ground station receivers, computers to interpret the data, displays to show the aircraft positions and then training for everybody along the line to use it.
The benefit is in better routing and less time in the "stack" when you arrive. Less fuel burnt is a cost saving but also think in total cost per minute of crewing and running a 747. It costs a bundle to switch to this but the longer term savings are far greater.
That's what they own. They lease a bunch more, for a total of 600 planes. In terms of fleet size, they're behind American, Northwest and Southwest (and FedEx - those packages don't fly themselves, you know...). They're just ahead of United. If you regularly fly with one of the majors, it's not unlikely you'll meet the same planes repeatedly. Counting just major domestic carriers, you come out at ~3500 planes. But this excludes the foreign carriers you'll see at US airports (most of whom will be tending toward the 747/767/777 end of things and away from the CRJs), excludes freight and perhaps most importantly, excludes the code-share partners you see flying regional jets from small airports to major hubs. These latter are usually flying with the livery of the associated major, but don't have much more to do with them. They're external and just flying those routes under contract.
A plane's registration number is unique. You can generally see it somewhere around the rearmost door of the plane. You might find airliners.net interesting.
PenguiNet: the (shareware) Windows SSH client
Read http://en.wikipedia.org/wiki/AeroPeru_Flight_603 to see how a blocked pitot-static port caused the crash of Airoperu 603 and how the air traffic controller had no idea of the true altitude of the plane because the altitude that was being shown on air traffic control radar screen was being broadcast from the aircrafts transponder which was incorrect due to the blocked pitot-static port.
Despite the somewhat high cost of the avionics, the real expense is the ground stations and the infrastructure to process all the data.
Let me fix that for you:
The altitude calculated by the GPS is way more imprecise than the value measured by the altimeter.
Which means that if there is a solar flare or something of the sort, the potential for disaster is enormous.
The chances of a solar flare killing a significant proportion of the GPS satellites seems very remote.
Loads of planes flying around close together using a system that depends on vulnerable satellite links.
The GPS ranging sats are in reasonably low orbits so not especially vulnerable to solar activity. Of course, they may be requiring SBAS signals too, which rely on a small number of satellites in GEO - much more vulnerable. But even if you lose the SBAS signal you've got a reasonably long time before the ionospheric corrections it provides are out of date (on the order of 15 minutes), which would probably be plenty of time to space the planes out a bit. In reality, I imagine that SBAS would only be required for GPS augmented landings.
http://blog.nexusuk.org
Yay for uninformed scare-mongering posts...
And satellites are far less reliable than radars.
Care to qualify that statement? Satellites are pretty reliable (I'd be inclined to say a single satellite is probably more reliable than radar, although I have no figures to back this up). Afterall, satellites are designed to run without maintenance whereas radars are not, so it makes sense that radars would be less reliable.
In addition, you would need to lose several satellites at the same time to render the GPS inoperable.
I don't want to trust my life on the optimistic hope that solar flars won't be at peak when I'm traveling
As mentioned above, you would need to lose several satellites at the same time to cause a problem.
Additionally, which is harder: disrupting radar systems, or shooting down a few satellites ?
Disrupting radar is probably a lot easier to do than shooting down several satellites.
which is harder: sending a few people to fix a broken radar in a few hours, or sending people up to fix a satellite in six months ?
NAVSTAR has been running for a long time without much trouble. There are more than enough satellites to cope with a few breaking at any one time and the satellites are fairly routinely replaced and deorbitted with no disruption to the service.
how many satellites would we need to cover reliably the whole planet before they can switch totally to GPS
24 satellites are required to cover the whole planet - there are currently 30 in operation.
http://blog.nexusuk.org
GPS Aided Geo Augmented Navigation aka GAGAN is what India calls it. This system is supposed to be operational in 2008.
wiki page
Details in google's cached copy of the announcement.
Google search would also get more details on this.
The FAA has been trying to upgrade the ATC for nearly two decades and is roughly seven years behind schedule from the original plan's timetable not the one they just changed to make themselves not look like total asses. The FAA has FAILED miserably and it is all of us who suffering. From longer ground delays at our nation's largest airports to few flights in smaller communities due to unnecessarily constricted airspace - the FAA's making it more difficult for all of us to fly.
I would suggest everyone read Michael Boyd of Boyd Aviation, an aviation consulting firm, that has been highly critical of the FAA and over a decade ago brought the idea of "Free Flight" to Congress but since that time has been ignored. Boyd has his pulse on the aviation world better then anyone I know and writes a column each Monday.
http://www.aviationplanning.com/asrc1.htm
Most planes have a dedicated instrument for just that purpose: the altimeter. IMHO GPS altitude is at best a backup.
For higher altitudes, the altimeter usually measures the air pressure. This isn't a problem-free method. You have to set the altimeter before each flight (to compensate for the height above sea level of the airport you're at). It's also not very accurate, as the indicated height varies with the barometric pressure. Incorrectly-set altimeters have been known to cause crashes.
A GPS altimeter would solve all this. Connect the GPS unit with a terrain map, and you're even better off: you'll know both your altitude above sea level, and above the local terrain.
Yes, you can get altitude from GPS. You need a minimum of 4 satellite signals, btw, for 3-axis calculations. However, until WAAS (Wide Area Augmentation System) was added, the z-axis was not accurate enough for air navigation.
Aviation altimeters only correct for air pressure variance (i.e. the barometer setting in the Kollsman window) they do not correct for temperature variance. Therefore, GPS with WAAS would be more accurate than a standard pressure altimeter. However, knowing what your absolute altitude is at 31,000 feet isn't important as long as every other aircraft in your area is using the same method for calculating altitude.
There currently is zero contention on where the funds will come from. Under current fee schedules, the FAA with have some 120 million extra in their coffers AFTER they complete their ADS-B deployment.
How do I know this? Because it's in the public record. The airlines and their lobbyist have been spreading misinformation and FUD on a make believe funding crisis. They have been doing this to take control of the FAA. What? Ya, sounds odd, but here are the details.
Right now, ever ticket sold has a tax which pays for infrastructure costs. Plus, every gallon of fuel sold (per gallon tax) pays for infrastructure costs. The airlines, by far, are the largest users of FAA services. What they want to do is to have the per ticket tax waived, pocket it, reduce their tax on Jet fuel and increase the taxes on the planes that hardly or rarely use FAA services. On top of that, they then want to create a "user fee" system where the FAA is free to set their own rates. The want to charge for items such as weather briefing, landing fees, IFR (instrument flying) service fees, in route update fee, etc. This means two things. One, and most importantly, the FAA would no longer have to own up to Congress on how and where they spend their money. Which is sad because right now they can not even explain where some 20 million went. And two, the small guy would be expected to pay the airline's share in taxes. Worse yet, even by the FAA and airline's own admission, they would suddenly create a significant funding short fall.
In a nut shell we have:
o Airlines want per ticket tax waived so they can pocket it (ticket prices would not be reduced)
o Airlines want a tax reduction forcing small guys to carry the airline's tax burden
o The airlines/FAA and crying the current infrastructure will not pay for new tech deployment
o Both the FAA and airlines have finally admitted their scheme will fall short of the existing taxes by hundreds of millions. AOPA has been saying this for a long time using the FAA's and the airline's own numbers with VERY conservative accounting.
Contrary to the assertions made in the article, there are fewer planes flying now than there has been since the 1970s; which is the US's peak in aviation. Even the current infrastructure can handle the load. The FAA's concern is a new category of jet has been created; the Very Light Jet (VLJ). The problem is projections indicate the FAA's current tax schedule will be able to handle the growth until at least 2030.
Long story short we have the airlines and the FAA working to break free of Congress' funding oversight. Currently, the US's FAA model is considered the best model in the world for both funding and safety, bar none. In all other places in the world where user fees have been implemented, GA has been destroyed, costing thousands and thousands of jobs. Worse, most analysts exist aviation safety will begin to decline almost immediately as pilots will now be reluctant to use federal services because it costs a per use fee. This means more pilots in higher densities without being in contact with each other. Worse, this means more pilots flying into unknown weather conditions.
Long story short, the funding for this system is already well established. Any short falls will be addressed by congress. Their current effort is to break free of congress and create a windfall for the FAA and the airlines; as they would be free to charge anything they want for their services. If they get their way, US skies will very likely become a dangerous place to be, even in commercial planes.
If this concerns you, I highly recommend you contact your representatives and congressman to let them know you expect the airlines to pay their own way and you demand the skies remain the world's example of safety. Tell them absolutely no user fees.
If you want more information, please go to http://aopa.org./
One last note, there is a FAA crisis looming. Right now, there is a mandatory re
If you trust the planes to tell you where they are, there is a potential that the planes could lie to you. I really hope they take that into account when designing the system.
I think that they already rely on the planes to transmit a lot of data correctly.
To the best of my understanding, civilian flight-control RADAR isn't an "active" system. It doesn't put out a whole lot of power and look for reflections, like a military system does. It's just a receive-only system, which listens to the signals being transmitted by the planes' transponders. If a plane changes its transponder code, it effectively "becomes" a different flight (with everything that entails: the ATC would think that it's a different type of plane, etc.). Short of going and looking up in the air, there's really nothing to prevent that, aside from whatever anti-tampering provisions the transponders themselves have.
But more to the point, if you don't trust the pilots in the planes, you have a much greater problem, since they are effectively big flying bombs. If a pilot wants to create havoc, they're more than able to, and they probably don't need to mess around with the GPS signal or their ATC transponder in order to do it. So a certain amount of trust is implicit in the design of the system. (In contrast, military systems or systems protecting critical parts of national infrastructure should NOT make the same assumptions, and shouldn't rely on any signals being transmitted from the aircraft; they should be active systems and assume that every possible attacker is going to be flying a stolen B-2 with its transponder and IFF turned off.)
"Ladies and gentlemen, my killbot features Lotus Notes and a machine gun. It is the finest available."
Just make sure it's not any young people you actually like. My wife and I have three children (and about 45 years combined in ATC), and we've made sure they understand all the good reasons NOT to follow in their father and mother's footsteps. 466 days to retirement....
No, but they can do 936km/h. 737-200 economical cruise is listed at 796km/h. 221 meters per second. GPS almanac transmits every 12.5 minutes and GPS Ephemeris every 30 seconds. So yes a few seconds matter when it means being off course by 2 football fields in crowded airspace. Obviously GPS data should be simply be "one more input" in the whole navigation system.
Civilian radar can do "skin paints" but you're right that in general it depends on aircraft to self-report.
As an example, in 1996 an Aeroperu 757 took off with tape over the static ports. This resulted in the crew being confused as to their
airspeed and altitude. When they asked ATC for information on their airspeed and altitude, ATC echoed the data from the transponder (i.e. what
the crew was seeing on their displays). The crew assumed that data came from radar returns and reduced their altitude and well, jet engines don't work too
well when they ingest sea water instead of air...
GPS has always had lousy altitude resolution compared to position over the ground. I've always thought it was both satellite geometry and earth model problems, but I can't really say for sure.
In any case, grandparent is currently correct: baro altitude is more precise than GPS.
Comparing Maastricht's area of operations with that of the FAA seems unfair.
a ge/FactsFigures.html
d -factbook/geos/us.html
Maastricht: 260,000 square km
http://www.eurocontrol.int/muac/public/standard_p
U.S. 9,160,000 square km (land area)
https://www.cia.gov/library/publications/the-worl