Mid-air collision over Brasil

[Brian Abraham]

From Aviation International News today

Report Blames U.S. Pilots, Controllers for Brazil Midair
According to reports in several Brazilian newspapers, U.S. Legacy 600 pilots Joseph Lepore and Jan Paladino and Brazilian controllers will be blamed for the September 2006 midair in which an ExcelAire-owned Embraer Legacy 600 collided with a Gol Airlines Boeing 737-800 over Brazil’s Amazon jungle, killing all 154 aboard the airliner. A final accident report, leaked to the Brazilian press over the weekend and expected to be officially issued tomorrow at noon by both the NTSB and Brazilian aviation safety agency Cenipa, claims the Legacy’s transponder was “inadvertently turned off by the hand of one of the pilots,” which was the “central point in a chain of errors” leading to the collision between the Legacy and Boeing 737 at FL370. A transponder turned off or set to “standby” mode also places its TCAS into “standby” mode. The controllers will be taken to task for failing to note the drop in transponder returns from the Legacy, miscommunication about the Legacy’s altitude and failure of communication between the crew and ATC. ExcelAire executive vice president David Rimmer said, “The transponder is a distraction from the true cause of the accident: ATC put two airplanes on a collision course for about an hour. It was compounded by multiple catastrophic errors and weaknesses within the ATC system.”

[GroundProxGuy]

Nation & World | Pilots outraged by release of cockpit recordings in crashes | Seattle Times Newspaper

IFALAP's statement expressing outrage at Vanity Fair for publically releasing full CVR audio of a fatal accident.

I have downloaded both files and listened to the 2+ hours of ExcelAire CVR in its entirety. As someone who works in aviation safety I find it most interesting to hear with my own ears the ATC comms and try for myself to make sense of what they heard.

That said, I agree with IFALPA's position.

[Roadtrip]

Even if the transponder was inadvertently turned off and the aircraft was under radar control, the ground controllers must inform the flight crew who would have recycled the transponder.

Either you are under radar control or you are not. Brazil can't have it both ways.

[ATC Watcher]

Even if the transponder was inadvertently turned off and the aircraft was under radar control, the ground controllers must inform the flight crew who would have recycled the transponder.

Agree, but for this you need all of the following :

a) a controller watching his display all the time
b) an ATC system that indicates clearly that the transponder failed
c) 2 way (VHF) communications

let's wait if the report will answer those basic 3 questions.

[broadreach]

One issue the Gol accident highlighted was the difference between a job and a profession.

With ATC being an essentially military function in Brazil, it’s difficult to know what improvements were put in place. Difficult here, let alone outside Brazil. One thing I’m pretty sure of – and BR aircrew on this forum might confirm or deny – is that substantial changes have occurred and that the airforce didn’t just hunker down and wait for the PR storm to pass but that they quietly lobbied for and got whatever they felt they needed to improve. And I doubt very much that whatever they got was limited to a handful of English lessons.

The overwhelming question, and one I suspect ATC Watcher may already have a notion as to the answer, is whether the corporate ethos has changed in any way.

[Richard_Brazil]

Cenipa final 1907 report available online

The final report is available online at: http://media.folha.uol.com.br/mundo/...o_voo_1907.pdf

I hear rumours that the NTSB will make an English version available. This is, of course, in Portuguese.

Local news reports show heavy ";spinning&amp" with vocabulary used only by lawyers, and frequent comments from "the families", which are always the same two or three individuals. Always intent on attacking the pilots.

There is some interesting new information. Up to now, only CENIPA knew which RMU turned off the transponder (the left one). Missteps by controllers are more clearly detailed; for example the Legacy was handed off to Brasilia Sector 7 very early, when still well inside Sector 5. The Jeppesen chart had an invalid frequency, 134.70 MHz, not foreseen on Brazilian charts.

The file, a total of 277 pages, includes from page 262 onward, a response by the NTSB, disagreeing with certain conclusions. Some of those are related to the automatic altitude chance function of Brazilian ATC software. An audit of the system by the independent Federal Court of Audit (TCU), in August, agrees with the NTSB. Interestingly, one of the objections made by the Air Force in the TCU report was that the controllers could use the criticism of the function as a legal defense in the Gol 1907 case.

As to the optimism on changes in ATC management expressed by Broadreach, I have seen no reason to share it, but agree with him that the forum would benefit from first-hand reports, instead of what I can offer second-hand.

- Richard

[GroundProxGuy]

fixed link:

http://media.folha.uol.com.br/mundo/...o_voo_1907.pdf

[broadreach]

Richard,

You have been the prime source of news on this thread, for which my thanks and, I'm sure, those of many others.

If I might quibble over your last post, it’s not just optimism although I suppose that does enter. More, perhaps, an awareness that, despite outward appearances to the contrary, things do change here, and I am constantly reminded of that in my own industry, shipping; frequently two steps forward, one back, but the overall direction is still in the desired one.

What I was getting at earlier was, I suppose, whether the airforce has been able to instill a more professional attitude amongst their ATC people. By professional, I mean “this is my vocation” as opposed to “this is my job for the next two years”. Not the easiest of challenges for a hierarchical organisation. If I’m at all optimistic it’s that brains within the airforce and without - politics aside - will be discussing the issue and find a way forward.

[kappa]

The latest comments by Joe Sharkey, a pax on the Legacy aircraft

Joe Sharkey, who writes for a number of U.S. newspapers, has posted three more blogs commenting on the latest matters.

Edit to say I'm trying to fix the link

Here are three links:

http://joesharkeyat.*************/2008/12/brazil-crash-key-charge-dropped-against.html

http://joesharkeyat.*************/2008/12/more-on-brazil-crash-charges.html

http://joesharkeyat.*************/2008/12/brazil-crash-was-primarily-caused-by.html

If they don't work, try joesharkeyat.*************

[Machaca]

Reports in English...

... available from NTSB:

CENIPA Draft Final Report

NTSB Detailed Comments

[ATC Watcher]

Finally got time to go through (most) of the 300 pages .
Nothing really new , basically confirming what was reported before by IFATCA /The Controller magazine in their special issue of the collision in March 2007.

But what I find interesting are some of the safety recommendations (pages 15-19),
which infirm some "official statements" .

Rec 266A06 for instance orders the Brazilian authorities ("shall immediately") to propose off-set procedures in regions which present communications/radar coverage deficiencies .
These deficiencies have been reported by IFATCA but until now fiercely denied by the Brazilian air force.

Similarly :
Rec 267 orders the ATS software to be changed to alert controllers of Mode C loss.
Rec 122 : about lack of 121,5 coverage.
Red 81 about (lack of) proficiency evaluation/training of controllers
Rec 77 about bad staff selection and staff shortages.

All these recommendations ( and some more) point out the real deficiencies of the Brazilian ATC system. Hopefully those recommendations will be followed.
Betters selection and training of controllers in sufficient numbers to operate a better designed ATS system , and correct the VHF coverage “deficiencies” will do a lot to rectify the situation. But , for this the BAF needs to recognize that ATC is a profession and not a military “ specialty” that need proper training and recognition. Recruiting off the street, giving them 3 months training , the grade of corporal or sergeant and pay them 250 USD a month will not contribute to that. If this is nor addressed , they the Brazilian controllers will continue to have another job ( even 2 ) on the side to make ends meet, and leave the ATC job at the first opportunity they get.

On the airborne side, Rec 204 about Pilot training requirements for international operations ,: I find personally this recommendation very pertinent in the general sense, especially in conjunction with the imminent arrival of the VLJs worldwide .

The rec 206 about certification of RMUs and the position of instruments is very relevant to this case. The report main text spends some time on the position of feet rests and crew training on how to use those feet rest whereas the real problem is the design of an RMU that allows to stand by the transponder (and disable TCAS) with a simple touch that can remain unnoticed by the crew.
Honeywell is apparently of the hook legally , but the problem remains in all EB135/145 family and might kill again one day.

Will something change in Brazil after this report ? Maybe our Brazilian friends here can answer that one.

[RatherBeFlying]

Design Errors

Left to themselves engineers can come up with brittle human interfaces.

ATC software that quietly changes cleared flight level is one obvious invitation to disaster that has been well discussed.

Device control is changing wholesale from single function knobs to multi function buttons. We don't know exactly how the xpdr got flipped to standby, but I would not be surprised if a corner of the laptop contacted the button. That a single inadvertent button push can silently put the xpdr into standby is not a design philosophy that allows for human error. Perhaps an audible click would have alerted the crew that a button had been pushed. The other question is whether xpdr mode selection should have remained controlled by a knob.

Displaying TCAS Off in white is reasonable on the ground, but needs to be more attention getting in the air. The message likely would have been caught if either crew had an eye on the PFD when it first displayed, but the report shows that they were busy working the numbers for the landing and subsequent takeoff from the reduced length runway at Manaus.

Same deal for the display of standby in the RMU -- in the air a case could be made for it to be flashing.

Losing TCAS in the air would seem to merit a chime and a line in the EICAS message list. It's worth as much attention as losing a hydraulic system.

[A-3TWENTY]

This brazilian report is a joke. They try in all manners two things:Accuse pilots and controllers

1They say that Brazilian airspace didn`t contribute to the accident, because the lack of communucation in that area is just "sporadib" .Lie,lie,lie.I`ve flown there for many years (I`m brazilian).

2. They spent almost the whole report trying to accuse the pilots and their flight preparation and the fact they turned the xponder off.
1. Pilots could be last one week preparing the flight, but at the moment they are cleared to Fl370 throught out the route..
2.They are in controled airspace so they can`t change a flight level by themselves.

3.TCAS is just a secondary way to avoid midair colisions.The very first one is ATC .

4. If pilots did the flight with just 5 simulation hours,is because law permits that. They are under the law. So guilty here is the guy which did the law.

To me, the first one cause of this accident was brazilian airspace infra structure and controllers( who are poor guys who have to work in dailybasis with poor equipment and very poor working conditions).
Pilots had just a small contribution to the accident when hey turn the xponder off.If they effectively did that.

A-3TWENTY

[Heilhaavir]

How about implementing SLOP in that region too, just like over the North Atlantic and the Chinese RVSM ...

[marciovp]

The military

In Brazil ATC and Cenipa are runned by the military. From the beginning they made sure that the ATC was out of the loop. Finally they are blamming the controllers (not the equipment, the software, etc) and the US pilots. Yes, the controllers did not see what was clear in the screen, the Transponder was off close to one hour! And yes, the US pilots did not notice either that the Transponder went off. Nobody know why. But the software gave wrong information (370) and the transponder off signs are now going to change to produce an alarm when a Transponder goes off. But this will take months. I agree, the ATC abandonned the pilots of the Legacy despite all the signs that they were without a Transponder and heading for a disaster. And, again the ATC is runned by the military,

[broadreach]

A few observations on the report, with a proviso: my knowledge is more of the maritime variety, so I look at the report as if it were a maritime collision case. Sealanes are not centrally controlled but the politics and other aspects are very much alike.

First, the report is not the whitewash many expected. I was happily surprised to note how far Cenipa went in its criticism of the existing ATC system and not at all dismayed at its criticism of the Legacy crew. ATC Watcher and RatherBeFlying go the nub of each aspect.

Re ATC, “being run by the military” is almost immaterial. The issue is whether it’s run professionally for its clients, or as a side issue which ensures revenue or some other advantage for the corporation, e.g. prestige or a bargaining chip in influencing political decisions related to budget allocations. Up to this accident it’s been the latter. The accommodations necessary for this situation to change are, I believe, already well underway, but you don’t change a corporate culture overnight. The process will be slow and the public won’t hear about the final result as it affects ATC until well after it’s happened. An indication of change though, is the recent discussion of airport privatizations and of new defense kit for the airforce.

Regarding the Legacy side, I wonder if one of the biggest holes in the cheese might not have been the very fact that ATC appeared to be working reasonably well in Brazil, its deficiencies being of the type we all work around in our day-to-day business, aviation, maritime or in any other industry. In hindsight, a "Tips, Tricks and Traps" for the Legacy men might have helped but who would have written it, a dozen AA, UA and TAM/Gol crew?

[bubbers44]

Had a situation over Swan Island over the Carribean one day in the clouds at FL350. We reported Swan Island on the hour at 350. One minute later World reported Swan Island at 01 at 350. I asked them to confirm their position report and they confirmed it. I then asked them what direction they were going and they were on a 130 heading. We were on a 180 heading so knew we were clear. A few minutes later the controller cleared them to FL370. No reports were filed but things like that happen a lot down south. TCAS is a wonderful thing down there.

[100BMEP]

Bubbers:
I'm impressed! You too remember it as Swan Island instead of Isle de Cisne!
It is a different world down south. It was not too long ago no one had enroute radar..so everyone would 'report traffic in sight' to get a climb or descent. Time have changed for the better...sort of!

[Blip]

If everyone simply made it a habit of offsetting a fraction of a mile to the right, ESPECIALLY those guided by GPS, no opposite direction aircraft would come close enough to hit each other.

Why does it have to be such a big deal to off-set say 0.2 or 0.3 of a mile (or even 1.2 nm to protect against those aircraft that are not GPS equipped)??. At least then if there is a collision, it is by chance rather than by design!

[ATC Watcher]

lateral off set

AH Blip ! just search this forum with the words " lateral offset" and you will have enough reading material for the whole of the Christmas period !

It is one of these old idea where everyone is for ( IFALPA ATS committee previous 2 chairman had seriously lobbied and tried to have it ) but the idea does not pass the mathematical model .
Without going into too much details, embedded ( i.e. inside the FMS ) offset is the solution, 0.1 NM right of track will be enough , but 1 NM was retained for the mathematical model. Result: offset increases the risk of collision at intersections. Then there was also the notion that voluntary decreasing precision of navigation goes against some other principles, therefore ICAO will not pursue the matter. There are other problems as well.

I see that Brazil says in the report that they will look into the matter, but little chance of getting anywhere I fear, unless they do it alone and file an ICAO difference.

The solution is not offset it is dual one way tracks/airways . This is what was finally done in Europe.( very easy to do with RNP5 and without loss of airspace ).It also increases capacity as all levels can be used if necessary on both sides.