Mid-air collision in Brazil: When precision kills.

The recent mid-air collision in Brazil of a new regional airliner (fitted out for use as a business jet) and a Boeing 737 has people baffled.  How could two brand-new airplanes with advanced avionics, flown by two professional pilots in each plane, collide at 37,000′?  The precision of modern avionics may well have contributed to this collision.

Airplanes under instrument flight rules fly from one navigation beacon to another along published standard routes.  In the old days, with radio navigation receivers and pilots flying by hand, a plane wouldn’t fly its clearance exactly.  The airways include a tolerance for error of +/- 4 miles.  If you’re 4 miles to the right of course, in other words, you’re still legal and safe from hitting mountains or other obstacles.  Altitude was similarly sloppy.  If you reached for a drink of coffee or to look at a chart, you might drift up or down 200′.  Air traffic control wouldn’t get upset.

How does it work now that the computer age has finally reached aviation?  The GPS receiver computes an exact great circle route from navaid to navaid.  All GPS receivers run from the same database of latitude/longitude coordinates, so they all have the same idea of where the Manchester, New Hampshire VOR is, for example.  The autopilot in the plane will hold the airplane to within about 30′ of the centerline of the airway and to perhaps 20′ in altitude.  If two planes in opposite directions are mistakenly cleared to fly on the same airway at the same altitude, a collision now becomes inevitable.

Almost any other system would be safer.  If you sent airplanes up to fly in random point-to-point paths, e.g., from Boston to Denver, they’d be less likely to encounter one another.  If you kept the airway system, but introduced some extra logic into the avionics so that planes always flew 1 mile to the right of an airway and + or – 200′ in altitude, they’d be less likely to encounter one another.  If you replaced the precise autopilots with imprecise humans, planes would be less likely to encounter one another.  If you replaced the high-precision GPS receivers with low-precision VOR receivers, planes would be less likely to encounter one another.

34 Comments

  1. Geoff

    October 7, 2006 @ 5:28 am

    1

    Automated boat foghorns have a random factor built into them precisely for this reason. A few years ago two boats crashed because the horns were synchronized and they never heard each other.

  2. A Nony Mous

    October 7, 2006 @ 10:09 am

    2

    This has been a common problem among amateur boaters for about a decade now, ever since GPS systems were tied into autohelm system. Captains will key in the GPS coordinates of a channel marker, not maintain lookout, and then drive straight into the channel marker.

  3. Mark

    October 7, 2006 @ 10:25 am

    3

    Why couldn’t the computer just compute the exact GPS great circle, and then offset it by a random amount each time, or perhaps with an amount computed from the geographical start and time or something?

  4. Joe

    October 7, 2006 @ 11:09 am

    4

    Not really true, because these advanced airplanes have collision avoidance systems on board (called TCAS) that should have alerted them. Makes the whole event even more puzzling.

  5. Robert Link

    October 7, 2006 @ 11:18 am

    5

    I doubt that precise avionics are much of a risk for mid-air collisions. Most mid-air collisions in the US (and I suspect other countries, but I haven’t seen stats) are between VFR aircraft that are flying random routes. Or consider the bizjet-glider mid-air from last month. Introducing slop into avionics wouldn’t have prevented that.

    In fact, sloppy avionics creates a new mid-air risk; viz., that two aircraft that are in theory separated are in practice on a collision course. Your +/- 200 foot altitude standard means that, for instance, a VFR plane at 4000′ and an IFR plane at 4500′, theoretically separated by 500 feet might in practice have only 100′ of margin. With precision flying you know when two aircraft are going to come in conflict; with random errors injected, you are never entirely sure. The result is larger protected zones around IFR aircraft, reduced capacity in the national airspace system, and higher workload for air traffic controllers. The latter will lead to more operational errors at ATC and possibly even more mid-air collisions than you would get if aircraft flew their assigned clearances precisely and relied on ATC (rather than random offsets) for separation.

    On an unrelated note, does GPS really compute great circle routes? I would have expected it to use a rhumb line, although I’ve never tried to plot a course long enough for the difference to matter.

    -rpl

  6. Jim Kellett

    October 7, 2006 @ 11:44 am

    6

    There’s a story, perhaps apocryphal, that during the Battle of Britain, with hundreds of airplanes struggling valiantly to shoot each other down, often at night, there was never an unintentional mid air collision.

    There is a good point in philg’s post, in that by putting control into a remote, out-of-cockpit site, whether a person or a computer on the ground (ATC), there is an increase in the risk environment. And, yes, this strategy can be traced back to the very first ground-based “ATC”, developed by an American airline, and used at Midway field in Chicago. From that humble beginning, the flawed idea that safety could be improved by moving the decision-making from the cockpit became embedded in policy and perception.

    The only exception, a fortunate one, is the use of TCAS technology for airliners and bizjets – at least this system exploits technology for what it can do better than a human (see better, react faster) but put the decision making back where it belongs – in the pilot’s hands.

    Now . . since both of the Brazilian aircraft were TCAS equipped, what went wrong?

    Jim Kellett, Resident Curmudgeon
    “Men would rather believe than know.” – Edward O. Wilson

  7. Jim Howard

    October 7, 2006 @ 2:09 pm

    7

    This is one of the strangest accidents I’ve ever heard of.

    Why did not at least one of the TCAS systems issue a resolution advisory? Or could it it be than an RA was issued, but ignored. That happened in Switzerland a few years ago.

    I know never to trust anything you read in the newspaper, but I’m sure you’ve seen the article from the reporter on the Legacy. He seemed to think the Legacy wasn’t talking to anyone, and just found this military runway they used on their own.

    Surely FL370 is in controlled airspace? The Legacy and/or the airliner couldn’t be up there talking to no one, could they? What was ATC doing?

    Very strange.

  8. Roberto

    October 7, 2006 @ 4:39 pm

    8

    In the route that the two planes crashed even altitudes (36000, 38000 feet) are reserved for planes flying in one direction, and odd altitudes (37000) for the other. The Legacy was flying in the wrong altitude when the crash happened — it should have lowered to 36000, but kept to 37000 since it was unable to confirm its flight plan with a ground station.

  9. ColoZ

    October 7, 2006 @ 8:41 pm

    9

    but kept to 37000 since it was unable to confirm its flight plan with a ground station.

    This is the first thing I thought of when I heard the report of the collision. Several months ago, I was a passenger on a United flight to Sao Paulo, and crossed this same region of airspace. United lets you listen to radio traffic, and I was amazed at how bad the communication was over Venezuela and the Amazon. The ground radio spacing must be huge, since much of the time we were effectively out of contact. A good fraction of the radio calls I heard were along the lines of “Amazonica Center, United 123, relaying. Varig 456 is checking in at FL350. He can’t hear you.” Garbled instructions and heavy accents don’t exactly help either.

    On the main point of the post, I do agree. Here’s another way to look at it. While few like to admit it, the primary collision avoidance method for VFR flights is “big sky, little airplane,” followed distantly in effectiveness by “see-and-avoid.” If the skies were congested enough that traffic conflicts were really common, there would be midairs galore. In IMC (or at high enough altitude, where true airspeeds are very high), see-and-avoid is completely useless so ATC takes over. But ATC (and pilots’ responses to commands) is not perfect, so BS-LA makes a good backup.

    However, the sky is only big because it has three dimensions! Putting lots of aircraft on the same line reduces the 3-D sky to 1-D and therefore makes its effective volume approximately zero. There’s no such thing as a “big” line, so the combination of ATC, pilots, and TCAS must be much closer to perfect. Is it? (As an aside, I never really did understand the reason for airways in Class A. Do they just make it easier for humans at ATC to keep track of flights?)

  10. philg

    October 8, 2006 @ 12:25 am

    10

    Folks: It is true that there are conventional altitudes for eastbound/westbound traffic under instrument flight rules. However, air traffic controllers are free to deviate from convention and I have often, especially here in the crowded Northeast, been instructed to fly at, for example, 4000′ while eastbound. As a pilot, you know that 4000′ is generally supposed to be for westbound flights, but you still fly your clearance and it is not considered polite to take up radio time in crowded airspace with “are you sure you didn’t clear someone in the opposite direction at this altitude?”

    Robert: Regarding your question, “Does GPS really use great circle navigation?” … I don’t have too much experience with GPS, but the Garmin 530/430 use great circle routes (your desired track change constantly on a long route; see http://www.garmin.com/manuals/GNS530_PilotsGuide.pdf) and so does the old standard King KLN 90B (see http://www.ocflightcenter.com/uploads/documents/ACF6103.pdf (more complete explanation)).

  11. Anthony

    October 8, 2006 @ 1:49 am

    11

    This article gave me a bit of a chill. Just this past week I was on an evening flight from Toronto to Denver. As we were cruising along at high altitude, out of the corner of my eye I saw something out the window coming at us. I only saw about 1 second of it, but there was another commercial airplane wizz by from under us. I swear the plane was no less than 200-300 ft from under us when it zoomed by. But it could have been farther than that – it looked scaringly close!

  12. JRT

    October 8, 2006 @ 10:34 pm

    12

    Most AA pilots I have talked to believe the pilots in the smaller jet didn’t file a flight plan and turned off the anti-collision instrument as they were trying to avoid the typical BS, Brazil, redtape protocol by sneeking across the border .

  13. rayfraser

    October 9, 2006 @ 5:02 am

  14. Dick Madding

    October 9, 2006 @ 12:46 pm

    14

    This accident was the result of the usual accumulation of events, the absence of any one of which would result in no accident occurring. Reasons for the following will be interesting.

    The Legacy transponder was apparently inop at the time of the accident. I’ve seen no mention of ATC remarking earlier to the Legacy that the transponder was not working. [Possibly some technical problem on a brand new aircraft]

    The Legacy was cruising westerly at 37,000 ft., an altitude normally inappropriate for that direction of flight. However, ATC always has the option of ignoring that convention in the interest of maintaining smooth traffic flow.

    The Legacy pilots reportedly erred in taking the new radio frequency during a handoff and were not in contact with ATC at the time of the accident. [Here’s a point at which a sharp controller would have taken action.]

    I agree with Philip that the degree of accuracy introduced by the use of GPS significantly increases the chances of a midair. This occurs not only on standard routes, but between any two points pilots elect to fly while VFR. While direction of flight altitude conventions are helpful, the climbing/descending are more risky.

    Among other techniques for avoiding midairs, I have long used the slight deviation right of course centerline and slight off-altitude for collision avoidance. The altitude deviation has saved me several times [even IFR].

    New technology always brings with it the liklihood of unintended consequences. The Comair crash was another instance, where there is no requirement to set a mechanical directional gyro just before takeoff.

    Here is the best review of the Brazil accident I’ve seen so far, but you need NYT free registration to view it. Only one technical error involving how TCAS works.
    http://www.google.com/url?q=http://www.nytimes.com/2006/10/06/world/americas/06crash.html&e=15314&sa=X&oi=news&ct=result&cd=2&sig=__7zhkEOyosk5E8-hqZHlkhRTc2VY=

  15. Solon

    October 10, 2006 @ 3:58 am

    15

    JRT: the Legacy did file a flight plan, which stated it should go down from FL370 to FL360 after changing airways. it also was en route to Manaus, where these aircrafts usually make one last stop for refueling before heading to the US.

    so far, according to the Air Force, the only facts are that the Legacy’s transponder was inop at the time of the accident (but working when it landed), and that Brasilia center never contacted Manaus center after apparently losing contact with the aircraft. the rest, at this point, is just rumor.

  16. hack

    October 11, 2006 @ 3:08 pm

    16

    There is a decentralized GPS-based anticollision technology for light aircraft being experimented with in Europe called FLARM. It’s a pretty obvious idea to try. One amusing thing is that the manufacturer not only explicitly forbids the device being used in the U.S, it also forbids the device being used in any aircraft carrying an American citizen or resident, such is the fear of the American tort system.

  17. Sweeney

    October 19, 2006 @ 5:48 pm

    17

    “Brazil: U.S. pilots followed flight plan
    The Associated Press – Updated 10/18/06 – 5:57 AM

    SAO PAULO, Brazil (AP) – Two American pilots followed their flight plan for at least part of the way before their plane collided with a passenger jet, the apparent cause of a deadly crash, the defense minister said Tuesday.”

    Now I’ve been following this story in both the Brazilian and the American press. The American update as of yesterday was the above, misleading headline. (You either follow a flight plan or you don’t.) It is inaccurate to say the flight plan was followed when in fact it was only partially followed.

    Now I’m still hoping that this collision is proved to be a horrible, honest-to-God accident and not negligence, but I wonder why the American press is doing such a poor job reporting this story. Imagine if the roles were reversed. Imagine the scene if a foreigner was flying in American airspace and had clipped a commerical jet liner and downed it in the US; and consequently had landed their own plane with no serious injury to themselves.

    The question that I have is how do two planes collide and one has the engine shorn off and the other has a ding on the wing? At those speeds how is it that the bigger plane is completely debilitated and the other plane not only lands safely, but the passengers are not entirely sure what they hit? Any theories.

  18. Peter T.

    October 22, 2006 @ 12:52 pm

    18

    I also wonder how the smaller jet survived the collision. Sometimes I think there might have been a third unknown aircraft involved — possibly an unidentified drug running plane that collided with the 737 and went down somewhere unknown in the Amazon rainforest. Maybe the debris from the 737 and the unknown aircraft hit the Legacy jet flying beneath them causing the slight damage to the smaller jet.

    I find it very puzzling that the smaller jet didn’t have significantly more damage after supposedly colliding with the larger 737.

  19. Jim C

    October 25, 2006 @ 9:59 am

    19

    I wonder if it was the fact that the 737 tried to avoid the collision and in so doing got into an unusual attitude that was difficult to recover from given that at that altitude there is not much speed above stall. I would imagine that in an ensuing dive it would be difficult to recover without overstressing the airframe. I’m certainly no expert in flying a big jet, so it might be interesting to hear from someone with more experience.

  20. Dick Madding

    October 26, 2006 @ 10:11 pm

    20

    At 1000 mph impact speed, it doesn’t take much metal to cause tremendous damage, especially to relatively thin aircraft surfaces. There was no third aircraft, just the vagaries of human decisions that brought the two aircraft together oriented such that David fatally smote Goliath.

    While there were several contributing factors, the most significant among them being the inop transponder, in my opinion a competent ATC would have prevented this accident, since it appears that both aircraft were assigned the same altitude and, because of the Legacy error in switching radio frequencies on a sector handoff, the normal ATC instruction to the Legacy to descend to 36,000 ft. could not occur. A flight plan includes only the route and the initial cruising altitude requested. Altitude is always per ATC instruction, and in the case of radio failure, the current altitude is to be maintained [until necessary to go higher for terrain or lower to land]. Again, my opinion is that ATC erred in not instructing the 737, with whom they WERE in contact, to alter course. I suspect overreliance on their new technology had something to do with it.

    Given the same circumstances, this accident would likely not have occurred in the US.

  21. Miles Rochford

    October 30, 2006 @ 7:51 pm

    21

    I have read elsewhere that there is a difference between United States and international (possibly Brazilian) airspace rules in situations where a pilot is unable to communicate with air traffic control.

    There are two possible options in this situation:
    1. Continue at the current altitude, and assume that ATC will clear a path (and if not, TCAS will take care of things).
    2. Continue following the altitudes set out in your flight plan, and assume that ATC will clear a path (and if not, TCAS will take care of things).

    My understanding is that ATC assumed one of these options, and the pilots assumed the other (which I believe is the standard mode of operation in the United States).

    In addition, the transponder issue is also not as simple as being switched on or off. There are typically five modes on a transponder: “TST” (test), “OFF”, “SBY” (standby), “ON” and “ALT”. There have been a large number of situations where pilots in Australian airspace have set their transponder to “ON”, believing that this will provide all necessary information to TCAS and air traffic control, and that “ALT” means “alternative” or “alternate” or something similar. In fact, “ALT” actually means “altitude”, and unless the transponder is set to “ALT”, this information will not be available to TCAS or ATC (who would normally request “recycle transponder” – if they are able to communicate with the aircraft). So it is possible to have your transponder “ON”, but not providing all the necessary information (this mode is designed for use in situations where the altitude reading is inaccurate by 200ft or more). Read more about transponders at http://www.public-action.com/911/transpon/ (note the comment that even if you don’t have altitude support, you should use ALT because if you use ON you may be invisible to certain types of requests from ATC).

    A tragic accident, but also another example of where human factors come into play.

    – M.

  22. Dante

    November 4, 2006 @ 5:51 pm

    22

    A question for experienced pilots: It is possible that in the chain of events, a communication error have occurred?. I have read that American pilots flying to the far east have a hard time communicating with TCAs. Is it possible that the controller in Brasilia did not understand the english spoken by the Legacy´s pilot?. The TCA of São José dos Campos, where the flight initiated authorized the Legacy to proceed to the Eduardo Gomes airport in Manaus at the same altitude of 37,00ft.

  23. Robert Patlovany

    November 7, 2006 @ 11:36 am

    23

    The latest Brazil newspaper speculation says ATC ordered the smaller jet onto the same altitude reserved for the opposite direction 737. This is in reference to the hemispherical cruising altitude rules, which generally require westerly traffic at even thousand feet (… , 32,000′; 34,000′; ….) and easterly traffic at odd thousand feet (… ; 33,000′; 35,000′; …). A pure physics rule (mean free path formula) guarantees that such a rule is more dangerous than random-altitude non-compliance. Test this out for yourself by counting bloodless virtual airplane collisions in your computer at http://web.wt.net/~b1rd/MidairApplet/JAppletMidairCollisionTestingDriver.html.
    The danger results because aircraft are artificially congested and concentrated in narrow slabs of legal airspace, while 90% of real airspace is illegal to cruise in. Additionally, near polar headings result in virtually head-on collisions actually encouraged by law. Summarizing my paper and NASA’s Russ Paielli’s paper, it is physically impossible to use the hemispherical cruising altitude rules to improve midair collision safety. Since my paper was published in 1997, 188 people have died unnecessarily in 31 easily avoidable collisions from the repeated uncorrected root causes of previous accidents. A Namibian collision killed 33 when one aircrew accidentally flew very accurately at the altitude for opposite direction traffic. A German collision killed 71 when Swiss ATC controllers failed to separate two aircraft accurately following the rules. About 100 were injured near Japan due to violent last-second maneuvering when ATC failed to separate two aircraft with over 500 potential victims. The TCAS collision warning systems were unable to compensate for the expected, repeated human error factors in the German and Japanese accidents. One of the above failures most certainly happened over Brazil. See http://www.geocities.com/rpatlovany/PreventableMidairs.html for the list of avoidable accidents and references to papers by Paielli and me. My solution is the Altimeter-Compass Cruising Altitude Rule (ACCAR) to replace the hemispherical cruising altitude rule. The ACCAR ideas is over 40 years old, cost nothing, and can be learned by any pilot with less than five minutes of practice. Every aircraft already has an altimeter and a magnetic compass, so no new equipment is required. The above link to the Java applet includes ACCAR testing compared to random altitude rules and the current fully discredited rules.

  24. George Rock

    November 8, 2006 @ 2:27 pm

    24

    Events sequence for Legacy X Boeing GOL 1907 Crash in Brazil

    Flight Plan presented to ACC – BR ( Brasilia Area Control Center ) via SBSJ – Sao Jose dos Campos Tower.( 23 14.78S, 45 51.25W )

    Legacy’s crew received Flight Plan CLEARANCE from ACC-BR via Sao Jose dos Campos Tower Controller.

    That’s the Legacy FPL route detailed:
    N0452 F370 DCT PCL UW2 BRS/N0456 F360 UZ6 TERES/N0449 F380 UZ6 MAN DCT

    Legacy’s pilots received CLEARANCE to CLIMB up and Maintain FL 370 to Manaus as CLEARED by ACC-BR.

    Departure time: 17:50 Zulu Time
    Compulsories Fixes along Airway UW2 : PCL ( 21 50.15S, 46 33.94W ), KERBO 21 06.2S, 46 44.2W), ARX (19 41.3S, 47 03.6W ), VALDI ( 18 03.8S, 47 28.5 ) and BRS ( 15 52.4S, 48 01.3 ). Heading 006 degrees from PCL NDB to BRS VOR.

    At least for twice after leveled on 37.000 feet Legacy’s pilots had confirmation about Flight Level 370 to Manaus.

    Overflying VALDI Compulsory Fix ( 95 NM to BRS VOR = 12 minutes ) Legacy’s pilots called Brasilia Center and received these instructions from Flight Controller:

    Legacy – Brasília, N600 transfering. ( before that Brasilia Center Flight Controller instructed Legacy’s pilots to make VHF Radio frequency change to next Sector frequency to be flown )

    Controlador – N600 squalk identification, maintaining flight level 370, under radar surveillance.

    Legacy – Roger.

    The plane overflown BRS VOR at 19:00 ( GMT – 3 ).

    Next Compulsory Fix to be flown it would be TERES ( 12 28.54S, 51 28.09W ), ( 282 NM from BRS VOR on UZ 6 airway ). Where Legacy’s pilots should CLIMB UP to Flight Level 380 as APPROVED in original Flight Plan, but overruled by LAST Brasilia Flight Controller AUTHORIZATION “Maintaining Flight Level 370”.

    Legacy aircraft had lost communications with Brasilia Area Controle Center.
    Captain Joseph Lepore and co-pilot Jan Paladino affirm they tried for many times to make VHF radio contact with Brasilia Area Control Center without success. They have claimed never turned down VHF radio volume and, they did not desengage the Autopilot for any reason after take off and they felt insulted by some newspaper accusing them by turned off the TRANSPONDER .

    The distance, over UZ 6 airway, from Compulsory Fix TERES to Brasilia Area Control Center boundary is 156 NM, heading 334 degrees. The next boundary is Manaus Area Control Center.

    The distance from Manaus to Brasilia Area Control Center boundary is 609 NM.
    Airway UZ 6 from Manaus VOR to Brasilia VOR has 1049 NM.

    The collision took place at 19:58 ( GMT – 3 ) around 171 NM from Compulsory Fix TERES near Air Force Base RESTRICTED area. This area is RESTRICTED to overflight from 18.000 feet until UNLIMETED airspace.

    Airspace from BRS VOR to MAN VOR is RVSM area. ( Reduced Vertical Separation Minimum ).

    The question is:

    Why ever didn’t Manaus ACC Controller put Boeing 737-800 GOL1907 under RADAR VECTORING to keep the plane from COLLISION?

    Boeing GOL1907 Departure was at 18:35 ( GMT – 3 ).

    After Legacy plane has overflown BRS VOR at 19:00 ( GMT – 3), ACC-Brasilia Controller had 58 minutes to coordenate with ACC-Manaus Controller to ALERT and DIVERT Boeing Flight GOL1907 to a CROSS-TRACK route parallel 10 or 20 miles to UZ 6 airway centerline.

  25. Randy

    November 8, 2006 @ 3:50 pm

    25

    This is just but another example of fuel for further technological advances. Like the mid-air over Europe, the question here is why TCAS didn’t do its job. The European accident was human intervention in what otherwise would have been a fairly straightforward decision-making process. What will be looked at eventually will be connecting the autopilot to the TCAS system to avoid mid-airs such as this. There will be pros and cons for each argument, for and against, but I can see the writing on the wall. But the question here that isn’t answered yet, is what went wrong?

  26. Robert Patlovany

    November 9, 2006 @ 6:54 pm

    26

    What went wrong is the by-design reliance on both humans and technology that fail repeatedly in a system that cannot tolerate faults. Compare this to fault tolerant safety systems in nuclear power plants and elevators. Fail electrical power, and the control rods fall into the reactor killing reactivity. Fail an elevator cable, and lack of brake releasing tension causes brakes to automatically deploy holding the elevator in position.

    Since the 1970s, Robert Machol and others have documented the “navigation paradox” that maximizes midair collision risk if there is any human or hardware failure. In the Namibian collision (1997) the German flight crew accurately flew the altitude reserved for opposite heading traffic–a human pilot failure. In the German collision (2002) and Japanese near-hit accident (2001), humans on the ground failed and conflicted with automated TCAS hardware. In the Denver collision (2003), a bad solder joint prevented the ATC display of altitude on the controller station (disabling automated altitude conflict alarms), so the ATC warning wasn’t transmitted until the moment of the collision.

    Flight operation regulations and aviation technology fails to implement fault tolerant collision avoidance systems that maximize the mean free path to collision. In other words, a fault tolerant system must use all available airspace volume in order to end the “naviagation paradox” that rewards inaccurate pilots with more safety. No airspace safety system can afford the extravagant waste of airspace volume required by the hemispherical cruising altitude rules, especially as the very light jet economic revolution takes hold. For a comparison of fault tolerant highway design to airspace design see http://www.geocities.com/rpatlovany/AviationSafetyMarch2003AsSubmittedUneditedWithReferenceLinks.html.

    If nuclear science failed (as aviation safety science has) to recognize the applicable physics of the mean free path formula, it would be impossible to build a nuclear reactor. As long as aviation safety science refuses to apply the mean free path formula in midair collision safety systems, the “navigation paradox” will be an accident-encouraging fact of life. The Altimeter-Compass Cruising Altitude Rule (ACCAR) promoted in 1968 by the last of four articles (http://web.wt.net/~b1rd/Historical.pdf) on the subject in “Air Facts” by Leighton Collins (father of “Flying” editor-at-large Richard Collins) would have prevented the Brazil collision and would have saved 188 lives in over 30 easily avoidable collision since 1997. Currently, the most skillful accurate pilots are the most dangerous pilots after the inevitable human or hardware error has occurred, because of the “navigation paradox” designed into the current system. The ACCAR idea credited by Collins to an Australian aviator 70 years ago, actually rewards skillful accurate flying with improved safety, while allowing the totally inept ACCAR pilot at least the benefit of random altitude safety over the current system. Purely random cruising altitudes were proven by me and by Paielli to be five times safer than the hemispherical cruising altitude regulations (Figure 1 in http://web.wt.net/~b1rd/JournalManuscript.htm and Table 4 in http://web.wt.net/~b1rd/RussPaielliFinalPaperFall2000AirTrafficControlQuarterly.pdf).

  27. Sweeney

    November 10, 2006 @ 8:02 am

    27

    Given the conclusions drawn by various organizations thus far as to the cause of this accident, is there any reason for the continued house arrest of American pilots Lepore and Paladino? I have no reason to suspect they are being abused, but it has been two months.

  28. Rob Mark

    November 19, 2006 @ 1:16 pm

    28

    I think Sweeney’s point is where we should be focused right now. It’s certainly something I’ve been tracking as a business aviation pilot. That could be me under house arrest in Brazil. And the question would be why?

    No one has completed any sort of investigation.

    No one seems to know why these two men have been unable to leave, except for the practical one – they have no passports.

    No one seems to be covering the situation at all at this point either. As I write this today, it is Day 52 for these guys. Most of the discussion are able the technical aspects of the collision.

    None of the Washington alphabet groups seem to find this the least bit concerning although the Air Line Pilots Association issued a statement last week calling for the men’s release.

    What are we missing? Or have these two guys simply fallen through some immense bureaucratic crack?

    And what does the Brazilian government’s ability to grab two U.S. pilots – or two pilots from any other country actually – say about international aviation law?

    Would the results be the same at this point if these were two British Airways’ pilots, or two business aviation pilots from Italy?

  29. Alex Fields

    November 20, 2006 @ 7:22 pm

    29

    About Rob Mark’s comment: we have to keep in mind that any person is bounded by the laws of the country where he/she is, no matter what is this person’s citizenship.

    The discussion about the two american pilots’ passports custody, decided by Brazilian Superior Court, can not be made under passion. We have to know exactly what local laws and regulations say about that situation, and what powers have brazilian authorities under that laws and supported by juridical decisions.

  30. André RD Baptista

    December 4, 2006 @ 7:52 am

    30

    “Next Compulsory Fix to be flown it would be TERES ( 12 28.54S, 51 28.09W ), ( 282 NM from BRS VOR on UZ 6 airway ). Where Legacy’s pilots should CLIMB UP to Flight Level 380 as APPROVED in original Flight Plan, but overruled by LAST Brasilia Flight Controller AUTHORIZATION “Maintaining Flight Level 370″.”

    This is a point that still not clear to me. I´ve read in brazilian newspapers that under international law the Flight Plan can not be overruled, and that in the loss of comunication the pilot should stick to the original plan.

    Still, I´m no specialist, so I would like to read a fundamented opinion about that.

    I also regret the way the american´s pilots are being treated, wich seems to me a bit of anti-americanism. Still, when I discuss this in Brazil people always point out “how do you think a brazilian pilot would be treated in US, had him hit an american plane killing 154 american citzens?”. That I can not answer, but reinforces my guess about the anti-american bias.

    André, Brazil

  31. Charles Gage

    December 15, 2006 @ 9:14 am

    31

    After hearing this story, the first question that came into my mind: The TCAS units on TWO aircraft failed? How is that possible? Bizarre.

    The premise Philip has about being overly precise has some merit. To me, the problem is GPS technology and that degree of accuracy is being applied to an old airway system. You recall the airways were designed to provide navigable routes, not efficient routes. The VOR is straight-line and made sense at the time. This was a huge improvement over NDBs, of course. But, following any segmented, straight-line route, as are the current airways, does force everyone down a single pipe, thus increasing the risk of collision.

    We saw some improvement with DVOR (mind cramp – is that right?) where we could plug in the offset in miles and degrees off a DME-equipped VOR to make a waypoint, thus straightening out the course.

    The use of GPS to overlay the airway system is a great improvement over NDB, VOR, and DVOR, but it misses the point — you do not really need to follow airways to get to where you’re going. Of course, you do need to follow them for legal reasons and “that’s the way the system works” but everyone I know requests “Direct” at every opportunity. The FAA is getting there with the “Open Skys” program where everyone flys “Direct”. Certainly, some studies need to be completed to weigh the pros and cons, but getting away from the fixed airways is a step in the right direction.

  32. Bernardo

    December 17, 2006 @ 5:44 pm

    32

    Flight plan is a suggestion for the flight. First, that can be changed, and normally it is changed, on the initial contact with the ground clearance control or anytime during the flight. The Ground clearance could gave him the same route as planed by saying – fly to Manaus as filled – but they didn’t.

    The only time you fly in accordance with your flight plan is on communication failure. The rule say’s after 20 minutes of the communication failure proceed to your flight as filled. The crash took place about 15 minutes after the failure. That means was not pilot’s error.

    They can blame as much as they want to on the pilots, but the pilots didn’t do anything wrong and everything leads to controllers error. Sure everything that happened to the pilots was an anti-americanism act. They wouldn’t treat two Brazilians pilots like this.

    Shows how bad the controllers in Brazil are. They complain about salary and infrastructure, but wouldn’t make any difference if they got all. They are really bad trained and they make mistakes all the time. Still an old military mind on the system and would take about 5 to 7 years to build a safe flying environment.

    I study the case really hard and there is no way these pilots are going to be convicted.

  33. Sweeney

    January 10, 2007 @ 9:48 am

    33

    “December 10, 2006 — The two Long Island pilots charged in Brazil’s deadliest air disaster finally had something to smile about yesterday.

    After 10 weeks of detention mostly spent in a Rio de Janeiro hotel, Joseph Lepore and Jan Paladino returned home to cheers and tears from a crowd of friends, relatives and supporters. …..
    Rep. Peter King (R-L.I.), who pressured the State Department to aid in their release, said the charges were just a ‘face-saving gesture to explain why they held these men for 70 days.’ King said he doesn’t expect them to set foot in a Brazilian courtroom again. “

  34. Carlos

    August 7, 2007 @ 1:59 pm

    34

    “Next Compulsory Fix to be flown it would be TERES” is not true. Next Compulsory Fix to be would be BRASILIA. In the last contact with ATC they were 40 miles to BRASILIA. The should descent to FL360 on BSB VOR and then climb to FL380 on TERES. Flights that arrive BRASILIA have levels 310,330,350,370…. Flights that leave BRASILIA have levels 320,340,360,380… Did Joe Lepore know that? Should Joe Lepore know it? If their flight plan indicated it, shouldn’t Joe Lepore guess it? Why this accident happened with a pilot that never make this route, and not with a dozens of commercial airlines pilots that use to make this route?

Log in