Indeed. Search for ICAO Doc 8168 PANS-OPS.

thank you! Is that the only document which supports that claim or are there more, do you know?

Also i meant to ask, i have a document titled CAP413, do you know if this is accepted by either ICAO or EUROCONTROL as binding to all contracting states or is it a UK only material?

You may read this:

http://en.wikipedia.org/wiki/Traffic...oidance_System
http://www.caasd.org/work/project_de...fm?item_id=153
http://www.aerowinx.de/html/tcas.html

for those who want facts and please remember the conclusion:

http://www.eurocontrol.int/msa/galle..._21Apr2005.pdf

I have been installing and testing TCAS on several aircrafts and have to say that position accuracy is really not precise on the display, therefore follow the ATC instructions first and if there is a RA command then follow the TCAS vertical manoeuver!
Never use your TCAS display as a RADAR, it is not designed for that purpose!
The only way for TCAS to be fonctioning correctly is that everybody have to follow the same rules and procedures...if not there will be other cases to discuss...I am affraid

Carpediem: I always like those who have no time to read but want immediate answers,and now it includes future lawyers ! :E

The answer to your question is simple and was already given to you : ICAO PANS-OPS Doc 8168 : this is the bible that set the world standard and its current version says :
" Pilots shall follow the RA even if there is a conflict between the RA and an ATC instruction to manoeuver "
Now all the rest of the world regulations ( like the UK CAP, which , you are correct only applies in the UK ) and all airlines SOPs should reflect that standard. They might use other words , translations, etc,.. but that is the standard.
Please stay away from Wikipedia and other source of info, we are not at school making a project .

Airmen :
Caution not to mix up ACAS displays azimuth resolutions and RAs. That was not the question asked and your answer can confuse the non-initiated. Once an RA has started , the position Display is useless , Pilot concentraton should be on the VSI , not on the position display.
Before an RA , ( and that includes during a TA , pilots should indeed follow ATC instructions , but once the RA has started, pilots should follow the RA, even if it contradicts the latest ATC instruction.

That partially explains why still today only 70% of the RAs are correctly followed , 20% are ignored and worse, in 10% of the cases pilots still choose to manoeuver opposite to the RA.

The other part of the explanation as to why this is still the case were/are discussed here.

I will take that to be driven by the long-established hatred for my future proffession :p I had indeed read almost all posts yet i couldnt find what i was looking for.So i figured i'd post first before i finished my reading of all 9 pages of the thred ;)
And one more random comment: you have no idea how difficult it is to follow your posts with all these details and terminology, it sometimes feels as if i am reading chinese hehe.

Yep i did find the document and it is indeed what i was looking for. Am i correct that this document was not ammended as such before Ueberlingen? Hence if it indeed existed as such, fault could technically be attributed to the pilot who did not follow the RA, correct? Also, both the pilots of the aircraft, didn't they have to inform the ATC that there was a TCAS-descend for the one and a conflicting order for the other? I am not sure that i read something that clearly states so in the posts of the thread.

And one last question to ATC Watcher. I read in a few of your posts that when TCAS II came out, the manufacturing company for obvious liability reasons used the name advisory instead of command for the RA. Is there any document or article etc of the time that i can find this written on, or is it just a rumour/common knowledge in the world of the air?

Carpediem :
Not really, but after having been scr..wed a few times , relying on the ethics of your future profession could be assimilated to masochism...:E

Back to the subject : Correct > the old version was not clear and explicit about this situation. It was changed after Ueberlingen.
The Russian Pilots were not at fault technically as they followed ATC before the RA started. After that, according the rules published at the time , it was left to the pilot to judge to follow the RA or not .

Yes they had to : the B757 pilot did it, but very late ( after 23 seconds ) and that message was blocked out by another transmission, so the controller did not get it .The Tu154 pilot did not advise ATC, (fact) probably because they never followed the RA.( my supposition)

Not common knowledge but a fact stated at the beginning, but on whose insistence I do not know. The company that built the TCAS software is the MITRE Corporation , on contract from the FAA. whether it is the Manufactures, the FAA or the MITRE lawyers that insisted on the change I cannot say. The first manufacturers were both Allied Bendix and Sperry/Dalmo Victor. They might also have influenced the change of terminology . Check their web sites and ask questions if you want to pursue that question .
I would be personally very interested to have the answer if you find it. . ( you can PM me or post it here )

Good luck .

I just emailed them about it so they will reply soon i hope. Thank you very much for your help and i will get back at you if i have something :)

The B757 reported the RA to ACC Zurich, but the controller was working with another aircraft on another workstation, where he could not hear the B757 (or the TU154M).

You are Correct on the other workstation ( and the different frequencies ) I should have said "covered" instead of "Blocked". But the end result was the same.

For info to others , the other workstition was only 2,5m away and the controller could follow both transmissions and reply to them, but as the frequencies could not be collapsed together he had to go from one workstation to teh other to transmit. . The problem was that at the time the B757 called, the A320 also called and his call covered the 757 call.

One can find this in the report of the controller and on the communication/time chart published with the BFU report.

downlinking of RA
 

ATC watcher, it was still on my to-do-list to react to your statements about TCAS downlinking to ATC.
First, there may be an average of 29 seconds before a pilot informs ATC. But that does not mean that ATC is unable to become aware much earlier, when an aircraft deviates from its assigned level. At the most 9 seconds after the pilot's following the RA instruction, the climb or descent will become visible in the data block on the controller's screen. Visible, but the controller will only observe height readout if he happens to look to that particular spot on his radar screen. And in the case of a TCAS RA, chances are high that ATC has missed to see the development of a close encounter and consequently will not look to the spot where collision risk exists (because he is too busy elsewhere or because it's an unknown military or GA), unless a simple message from the TCAS RA downlink mechanism through Mode S draws his attention to it.

Second, in case of doubt wether the pilot follows the RA, ATC might ask. Might ask, because I am not sure the question is welcome at the flightdeck during those hectic and critical 29 nine seconds before the average pilot comes to informing ATC himself. But, to all intents and purposes, a simple question is much better than a contravening instruction from ATC which cannot be other then confusing.

Third, either ATC trust that pilots act in accordance with their SOP's and adhere to the RA, or we take it for granted that pilots ignore a potentially life-saving warning. Shouldn't we just agree that it is bad airmanship when pilots play around a TCAS RA. Ok, we know it happens, 10 % according to the study you quote. That means that there is still work to do at this front. But bad controllership also exists and I maintain that it is bad controllership if ATC prefer to ignore knowledge of RA's which affect the progress of a flight. Politique de l'autruche, the ostrich putting its head in the sand.

I'd think this debate is academic and by extension false argumentation. In the case of an RA, the provision of separation by ATC has failed (in the neutral sense of the word) and is replaced by the pilot taking ultimate responsibility for the safety of flight. The only action to be taken by controllers is... to refrain from any action, to shut up except for advising other aircraft in the vicinity which might be affected by the evasive action.

Finally, I must say that I have never understood the lax attitude of the controllers' international organisation IFATCA vis-a-vis the possibility of getting technical life-saving improvements, which could to a high degree prevent a collision from happening.

In the aftermath of the Uberlingen accident technical improvements remained limited to some improvent of procedures and airborne TCAS upgrades. This was necessary and useful, but we are still waiting to see ATC improvements that preclude ambiguity with RA instructions through downlinking of the RA. ATC administrations and airlines who have to pay the bill were wary of the financial implications. Uberlingen was quickly put back in their minds in favour of the rise in traffic figures, resultant capacity problems and cost reduction.

But the controllers had to get through the assassination of one of their colleagues. His recent acquittal by a Swiss court could not bring him back. Every TCAS RA event happening in their sectors reminds them of the shortcomings of the system. Yet, their representative body IFATCA, instead of putting pressure on their administrations to provide them with the best possible technical information about TCAS RA's, kept pressing on legalistic argumentation about responsibilities and accountabilities.

Carpediem,

I work with real lawyers on real aviation cases, amongst other things trying to keep real people out of jail. But my day job is with a university.

I think I can appreciate how difficult it is for people not in the field to understand complex technology. I have been in that position. My inclination is to do what it takes to understand all I have to know in order to make correct informed judgements (I am known for this, and people pay me for it occasionally). There is another way of dealing with things, followed by most experienced lawyers, which is to employ the right experts (that is where your judgement comes in) and believe (and have them transcribe for the court) what they say.

In a real court case, you would be offering ATC Watcher lots of money to say what he has to say. In a moot court, you are unfortunately faced with the first option: no matter how hard it is, you have to get your mind round the details.

[QUOTEcarpediem86]Hence if it indeed existed as such, fault could technically be attributed to the pilot who did not follow the RA, correct[/QUOTE]

That is, if I may say so, far too naive a stance for an putative aviation lawyer.

Obvious question, which should be the first one you ask as a lawyer: who has jurisdiction?

Answer: obviously Germany, since the airspace was German.

So you had better look up applicable German law. If you don't read German, you are lucky in this case in that the applicable German law is translated in the accident report from the BFU, which I may presume that you have read, as a lawyer.

You will therefore have observed that the advice about following RAs is not strictly identical with the advice in PANS-OPS 8168.

But Germany is an ICAO signatory. So which advice applies and why?

You will also observe that there is another applicable German law (which is more or less universal), which says that the airplane on the right (i.e., the Tu-154) has right of way. According to that regulation, your suggestion that
is not correct.

You're the lawyer. You sort it out.

Whoever suggested you work on this either had it in for you or thought you were a legal genius. It is certainly worth trying to sort out, but don't expect anyone here (or anywhere else) to have packaged answers for you.

PBL

Even without TCAS RA aircraft are allowed to deviate from their assigned flightlevel. Only if the deviation exeeds 200ft ATC can become aware of the deviation.

As the radarsystems updates the displayed information in intervals, the timespan between begin of the deviation and recognition by ATC can be several seconds.

Even with all these factors for delay, technically there are two institutions trying to control during a conflict situation (ATC and TCAS). These two obviously need to be coordinated in some way.

This need not be true in all cases. In the case of Ueberlingen, the TCAS RA was issued before the aircraft underrun the 7NM separation limit.

Thanks Peter, I am indeed too naive, Capediem was also asking me in PM a lot more questions ( which I gave him ) .
I realize now that I could perhaps make a living out of this !:O


Now answering songbird :

A lot of things to comment upon in your post.
The easy ones :
First : the 29 seconds is indeed an average. That means some are sooner , but some are also taking longer or missed altogether.
On the sooner ones, you say after 9 seconds max for a radar return, this is not universal . The Zurich radar return was 12 seconds in Ueberlingen for instance .
Also many controllers always wait the second Mode C update to verify trend , as many small deviations are just garbling or aircraft altitude hold variations.

Second : In doubt why not ask the pilot if he is following an RA. : Why not indeed . Just that in case of imminent collision, as a controller, you are more likely to want to issue avoiding instructions instead of loosing precious seconds asking the pilot if he has a functioning TCAS and if he is following an RA.

Thrid : ATC at the moment prefers to refrain from Downlinking RAs , not to ignore RA knowledge. On the contrary ATC urges pilots to report to ATC immediately if they are getting and following and RA. ( In Ueberlingen should the 757 crew had done so earlier , the outcome might ( emphasis on might ) have been different )

A little more info ( to confuse you even more ) on RA downlink to justify IFATCA ( and my own ) position on RA downlink:

A very recent Bretigny Survey of a single Mode S radar messages show the following picture :
90% of RA downlink messages are false RAs not displayed in the cockpit, but are broadcasted. . Half of those were identified as coming from a special transponder manufacturer on certain Boeing 737s ( 51 airframes identified so far , mostly on the UK register ) and a solution for those is expected.
The other half are short false RAs whose reason is not yet understood.
Of the 10% that remains, a further study indicates the following percentages : in 20% the pilots actions are inappropriate, in 10% pilots are acting in the opposite direction of the RA, and only 70% are more or less followed correctly. this confirms other larger studies both in Europe and in the USA.

Now to answer you final questions :
No, IFATCA on the contrary has the correct approach to this. They believe that hastily introducing RA downlink for the wrong reasons ( The Japanese pushed this in ICAO after their JAL/JAL encounter) while there are still so many issues unsoved will create more additional problems than it solves.

I for one believe that downlinking RAs Mode S messages at the moment will increase controller involvement and contribute to more confusion , or Ueberlingen type scenario.


This is a fundamental issue : valid for TCAS, but also for ADS-B and ASAS : Who is responsible for anti collision and when.

Nobody has resolved this issue yet and will resolve it soon I think.

For the moment I am responsible for anti collision TCAS is only a safety net that i should not take into consideration when I work. ( I am even not sure/ aware who has a functioning one ).
If the pilot is coming on the frequency and say "Moving , TCAS RA , " I Know my responsibility ends and he takes over . ..,I do not think any regulator will change this on account of RA downlink. So this is why IFATCA wants the legal bit solved before getting the Downlinks on their radar scopes. Also, based on the current surveys , Mode S RA messages are full or garbage anyway.

We are not putting our heads on the sand. We have been burned already by a hasty and immature introduction of version 6.04 ,often without training , in life traffic. We have had Ueberlingen, so time for the Regulators to take their responsibilities I would say. I support IFATCA 100% on this one .

The ball is in EASA and the FAA camp.

Joernstu wrote :
True for the 200 feet deviation, it's actually 299 feet. But a TCAS originated FL deviation will very soon get beyond 300 feet. I agree this has some impact on the time lapse before a controller could see the FL deviation without a TCAS RA downlink. But remember the context of my point. Because there is no attention getter related to the FL deviation, it is not very likely that the controller will see it because his attention is drawn somewhere else. What the controller needs is the attention getter in order that he can take the right measures, which normally will be to refrain from action vis-a-vis the RA affected aircraft.
Yes, several seconds, but always better than the average 29. In busy areas, with multiple ground radar antennes integrating the signals for display on the controller's radar screen, the delay is only two or three seconds.
Very well put, this is the heart of the matter. And it requires a good solution, see my further points to conclude that a good solution is achievable and is better than a do-nothing attitude, waiting for the best solution.
-
ATC Watcher wrote:
You put the 29 seconds and the 9 seconds in one basket but they apply to different magnitudes. The 29 seconds gives the average time lapse before pilots inform ATC on the radio, whereas the 9 seconds is, at least according to the Qinetiq study, the average radar return. My point was that on average there is 20 seconds to gain for getting the information where it is needed. Now on the detail, in Zurich 2002 the radar return was 12 seconds. However, the Swiss violated here the Eurocontrol standard of 8 seconds. I think this is mentioned in the investigation report but I haven't checked up yet.
As you will have read, I have my doubts myself, but it's better, or shall we say less harmful, to ask a question than a contravening and confusing instruction against TCAS.
Exactly, the instinct of a controller is to do as you describe. But the difficulty is that in the case of a TCAS RA, he should not follow his instincts any more. TCAS has taken over and the controller should act accordingly which in this case is to refrain from action, other than removing third party traffic. You will have seen my own hesitation to put up this possibility of asking the pilot. The main point is that the controller should, at the earliest possible time, be stopped from doing the wrong thing, that is to issue a contravening instruction which is so much confusing in the hectic situation of an RA.
So, connect the downlink to those RA's which are displayed in the cockpit. Problem solved.
A remaining problem indeed. As I said before, a lot of work remains to be done. On the detail again, it is not clear to me wether the '70 % more or less correctly' is of the total population, or only of the 10% that remains. If the latter is the case, then, logically, the real problem is only for 3 % of the TCAS total?
I don't think your statement is supported by the downlink simulations which were organised in Bretigny. Furthermore, how can knowledge, acquired by a downlink, be confusing. It is lack of knowledge and bad decisions taken based on a lack of knowledge, which is confusing. Not only in ATC but anywhere else. In this context but nothing to do with TCAS the Amsterdam crash of the El Al a/c flying into an appartment building springs to mind. The crew couldn't see what had happened with the engines which had fallen off, by this lack of knowledge they decided to turn in the wrong direction which caused the dramatic going down of the aircraft.
STCA (short term conflict alert, the ATC safety net) is only a safety net. I'm sure all area controllers take it into consideration when they work (admittedly, provided the number of false alerts is reasonable, but if this is the case management and technical staff should urgently repair the tool). Why should TCAS, as you say equally 'only a safety net', be treated differently when this is made possible? Why wait the average 29 seconds before it is taken into consideration anyway? I am pleading for improving the controller's situational awareness, rather than impeding. Improved situational awareness will improve his decision making, and not confuse. In my previous message I called this good controllership.
True. The first TCAS versions were introduced, pushed through even, without safety cases as to the impact on ATC. However, that's spilt milk, ATC has to live up to the situation and should embrace the improvement at hand, which brings the individual controller in a better position to apply his judgement when he gets involved in a horrifying RA situation.
I have noticed with satisfaction that you have put in bold 'for the moment'. That would seem to imply that you, and dare I say through you IFATCA, provide an opening to accept the downlinking of TCAS RA. You are right of course that the legal stuff should be solved. But the statement that the ball is elsewhere (what has the FAA to do with Europe, but that's another matter, I would think worldwide ICAO is the Ageny to address), alludes to a passive attitude from the side of IFATCA. Laxism may have been a too strong expression, but I would expect IFATCA to be active on this front. If it's only a few sentences in some documents to be changed : don't wait but go for it.

One scenario I can think of and where this would be problematic is a situation between one aircraft with transponder and TCAS and another without with ATC providing separation.

If aircraft 1 (the one with TCAS) for some reason deviates from its flightpath towards the flightlevel of aircraft 2 and almost simultaneously its TCAS transmits a bogous RA to ATC, the controller will not advice aircraft 1 to return to its flightpath as heshe would see the TCAS RA indication on the screen and could miss to recognise that aircraft 2 cannot be the cause for the RA.

So perhaps for downlinking TCAS RAs in addition to transmitting only real RAs (those issued to the crews), ATC would also need a database on all aircraft listing their transponder state. This database itself could introduce other failure scenarios.

The indication of TCAS RAs to the controller has to be correct 100% when it is made - or the indication should not be made at all.
Reason for this is, that no indication made at all will not change the current situation where ATC and TCAS simultaneous try to control the aircraft - not changing the current level of safety. TCAS RA indications to ATC which aren't 100% correct could lead to controllers acting as though they were 100% correct but in those cases where they aren't lead to dangerous or fatal situations - in extreme lowering the level of safety instead of improving it.

Songbird , here we go again :
A bit more in depth this time.

On the delay : .
Problem is that ATC using radar always look at the past. whether it is, 4, 9 or 12 seconds. TCAS is a very dynamic system, which can issue corrective RAs and whose version 7 can even issue corrective reversal RAs. Those, in order to be effective have to be followed by pilots within 2.5 seconds according the manufacturer manuals. We agree so far right ?
My point :
This will only happen in some particular situations, I agree, but in a mode S download situation what will be displayed to the controller on his screen will be totally different of what is happening in real time, hence increasing the possibility /likelihood of controller intervention.

On the technical downlink medium :This will mean re-encoding the displayed RAs and re-broadcasting them .I am not an engineer, but this will be complex and possibly expensive , and adding an avionic box is not the idea.
Also in Eurocontrol they have already decided that should we go for RA downlink , it ill be using the mode S broadcast. The cost of using the other (better) method of using the 1090 Extended squitter was judged prohibitive for that function only.

Replying to your question :
The 70% is within the initial 10% which are valid correct RAs. The rest is raw data , you cannot start to make percentages values on unvalidated data !

Increasing Controller involvement 2 :
Be careful in analyzing the outcome RADE simulations. the rather large RADE 1 was only a simulation on the HMI ( i.e the display) of RAs, and indeed most controllers there found the HMI adequate.
The 3 subsequent low-scale RADE simulations start to show another picture, and the 4th one ( RADE2T, only with 4 controllers from 2 APPs ) show the real limitations. The number of RAs shown to controller during those simulation was so low that I would refrain from drawing firm conclusions one way or another.
What I myself found interesting (and worrying),in those simulations is that despite this simulated environment, and the low numbers, on 2 occasions controllers intervened after an RA ! Training is still a major issue, also for controllers..

On the safety nets :
Ah, the old discussion between a safety tool and a safety net !
We could debate this for hours,. Short version : NO controllers should not use STCA and TCAS as safety tools. ( and they are trained and told not to )

Finally :
The FAA owns the TCAS system . Changes to its software and procedures have to be discussed with them trough RTCA. It is not a European system. ICAO do only issue SARPs for ACAS . They have no influence on TCAS.( and frankly no expertise either )

As an aside, I do not represent or talk for IFATCA here, but it happens that I agree with what they say on this particular point , which also happens to be very close to what the IFALPA ATS committee says by the way.

In a nutshell : No against Downlinking of RAs per se , but not now. Solve the problems first to make sure we will not add more problems than we want to solve, and increase, not decrease the likelihood of controller intervention .

Absolutely against introducing it now here and there to " see what happens" and " gain experience " etc...

Do all operators use TA/RA all the time?
Do any operators or countries allow for selecting TA only on , say, parallel runway approaches.

As far as I understand it is an FAA recommendation to consider TA only. Original Airbus manuals state TA or TA/RA AS RQRD with a note on the FAA stance.

My employer does not go any further and the accepted practice is always TA+RA. Never had any problems with parallel runway operations (not always approaches) in FRA, MAD, BCN, BRU, AMS, LHR, FCO, LED, SVO, OTP, HEL, MAN, OSL, CDG ... However, the runway spacing is not very tight and there is little or none GA/VFR traffic, which I suppose may be the main reason to select TA only above 800 ft on approach.

FD (the un-real)

I have had 2 RAs at DEN. Out of 5/6 visits this year.
On 1 occasion we were 4 across - to the 4 parallel rwys.
We had a/c on both sides, very close, gear down, and if we all reacted, or did full go-arounds, we would certainly lose sight of each other. And could I trust the software to cope with 3 or 4 simultaneous reactions?

As a followup......This quote was made by Joernstu earlier in this thread.
http://www.eurocontrol.int/msa/galle...n_9_Jul-07.pdf

I direct you to the Eurocontrol ACAS Bulletin #9 and quote...

"Five. The factor of collision risk reduction thanks to TCAS II in the operational world, taking into account some non TCAS II equipped aircraft, inaccurate pilot responses and lack of responses, altitude reporting inaccuracies, etc."

There have been some irresponsible statements made much earlier in this thread. It is rational to Follow The RA even if it contradicts an ATC instruction.

Reading your quote I thought this was point "5" from an enumeration of various points. The meaning, however, is that TCAS has reduced the risk of a mid-air collision by a factor of FIVE.

And I'm sure they arrived at this figure by finding out that there were 80% fewer mid-air collisions since the mandatory introduction of ACAS-II per billion flight hours, compared to before. Or did they?

Otherwise, this figure is completely bogus. (I'm certain it is bogus anyway, since luckily the number of mid-air collisions is too low to derive any statistical significance from it.)

Trying not to re-iterate the entire argument: we did not doubt that following the RA was a rational choice. The point was that it was neither the only rational choice, nor necessarily the best (as the DHL reaction clearly shows: they followed the RA, and had a mid-air collision. QED).


Bernd

You would be sure of that if you believed risk reduction is the same as fewer actual midair collisions.

I don't have the details of how they came to their conclusion or Joernstu came to his conclusion, but I am left with a choice of believing an organization called Eurocontrol and their statements and excellent newsletter or the supposed risk evaluation of some guy on a website called Joernstu.
Choice made. I believe the collision risk is less with TCAS installed.

Interesting that they underlined "operational world" in their statement. Maybe they are reading this thread. Perhaps this "risk evaluation of would be episodes" should be sent to them and they can reply with their "bogus" data as you call it. After all, safety is our common goal.

Of course it is rational to follow the RA and blend everything else out. This will well be the action that comes nearest to the mathematical "rationale".
Problem is, as soon as you introduce humans into your system, the question of rationality becomes a lot wider. A course taken by a human will be rational, if he acts acording to the limited information he has. If he chooses, the best way for him to solve his problem will be course A, course A will be a rational choice. You cannot decide on the rationality of a humans actions by looking at the outcome alone - the pilot's cristal ball isn't near as good enough for this ;-)

Well thank you, I'm glad you agree with me. After all it was a statement(way back in the thread) that contradicting the RA was a rational choice that got me involved in this thread.

Well, it can be, as I described before.

Yes, but you should be aware that until you can adequately solve the decision problem posed to you then, few TCAS experts will take your interventions seriously.

PBL

I'd say Eurocontrol and I are on the same wavelength about how to proceed in such a situation. The BFU as well. I take them seriously and don't refer to statements from them or one of them as bogus. It is not for the TCAS experts to take me seriously. It is me taking them seriously which I do.

Why don't you send this problem or situation to the appropriate Eurocontrol department(we already know the BFU's opinion) and we can see what their reply is. I will start holding my breath now.:ooh:

Well, my friend, this is where your youth and lack of experience are showing through again, as also my inclinations as a professional educator.

As far as I know, the colleagues at Eurocontrol were reading this thread when there was technical discussion (I doubt they are bothering any more, and I can't be bothered to ask).

The BFU has nobody expert in safety-critical digital systems,
and why should they. They don't design them, build them or analyse them.

PBL

Hmm, so now I am youthful and inexperienced by reading Eurocontrol newsletters that say to FOLLOW THE RA, ageeing with them and feeling we are on the same wavelength. As well, the BFU in their final report recommends "never manouver in the opposite sense to an RA(directly from the report that deals with your scenario). I am youthfully on the same wavelength with them despite my inexperience.

I thought about your scenario the other day as I was flying along at a high altitude. There was 180° opposite traffic almost on our track that we knew was several thousand feet below us. We saw their contrail well before we crossed paths. For much of the time their contrail appeared above us. It was definitely above the horizon until it got much closer and became obvious that it was much below us. Sort of like the cloud tops that often appear to be at our level initially the pass by down below. Your scenario has the option to visually manouver to avoid the visually aquired traffic which is opposite to the RA that has alerted you to a strong collision potential. As I have previously said, visual illusions, especially at night can make it extremely difficult to accurately assess the trajectory of another aircraft in this situation.

In your scenario, which is the accident scenario, it is safer to climb. Why, because we know how extremely dangerous it is to descend as proven by the European collision and other near collisions and it is unlikely that the TCAS is wrong. The other guy will likely be descending as well because he is well trained. Manouvering visually may make things worse in many cases (especially at high speeds, high altitude and night). How easy is it to manouver visually around a rotating beacon and a set of nav lights closing in at 500 miles per hour, possibly against a backgound of many ground lights). Is this your actual experience and non-youthfulness speaking on how easy it would have been for the 154 to descend and visually manouver around the 757).

How dangerous is it to climb or maintain level flight(assuming that there really is an aircraft out there that you can't see)? I don't know, but I say less dangerous. Why? The big sky theory. The chances of actually hitting that aircraft are still extremely small. ATC is going to calling for an immediate climb or descent if your aircaft will be within 5 miles of each other(or whatever the minimum separation is). TCAS is calling for a manouver when the collision potential is much higher. It is nice to have hours upon end in an office to analyze various scenarios. When you have few seconds to make your decision, certain procedures should already be clear in your mind. Sure, there is always the obscure scenario where acting contrary to the SOP saves the day. But that is a rare day. and it wasn't the Day(or night) over Germany.


That the BFU has no experts on TCAS is not surprising. Safety boards typically bring in outside experts during an investigation.


It is unfortunate to hear that you are close enough to Eurocontrol to be
aware that they were reading this thread yet you "can't be bothered" to attempt to discuss with them your scenario and their recommendations and reasons for their recommendations for your scenario.

In case you happen to find the time, their address, fax and phone number are on their newsletter.

punkalouver, in all your argumentation, you continously overlook, that the topic of this threat is the discussion on TCAS philosophies. This cannot be limited to the teaching material published by any organisation.

I wonder - if you really have read the Eurocontrol bulletins as you said you have, why hasn't it occured to you, that all the cases published there have one causal factor in common: TCAS. Some of the problems described are related to training, but some of them seem to be very hard to mitigate, e.g. the interception of a A340 by a military jet transmitting altitude information described in Bulletin 9. Apart from changing the practice by the airforce there seems to be little, that can prevent this kind of incident from happening. What strikes me in this incident is, that only the disregard of the A340 crew to follow the TCAS RA when reaching critical altitude led to a reversal RA. TCAS could have noticed that the intruder (airforce) was not acting according to the TCAS logic and could have issued a reversal earlier - but this will be solved in TCAS 7.1 as Eurocontrol promised.

Interestingly enough, you quote the BFU report but seem not to recognise, that the first point stated by the BFU report under "systematic causes" (page 119 in the german version) is the insufficient integration of TCAS into the aviation system.

I haven't overlooked that at all. I have just responded and perhaps added to 2 very, in my opinion, irresponsible statements.
1) that the actions of the TU-154 crew were rational
2)The statement "I cannot decide, what kind of action would induce the lowest risk - following an RA, not following an RA or flying without TCAS at all as the basis for my risk evaluation would be episodes. I think the same goes for you as your information basis is still only episodes."

There are occasional pilots out there who will read or hear about about some obscure or one-off scenario(not necessarily TCAS related) and as a result decide that they have a better plan of action than the SOP. Perhaps it is how to deal with a certain type of fire or ditching procedure. I have flown with these types. Statements like your coming from a supposed accident analysis panel or whatever it is called I believe increases the likelyhood of some people disregarding or going against an RA.
I think, perhaps it is you who should read the bulletins. The are all kinds of examples where TCAS worked as programmed and saved the day including two in Bulletin 9(one an altitude bust,the other is an ATC error). The A-340 incident was caused by proper procedures not being followed which can cause an incident in many ways in aviation.

I read this statement which is located a second time earlier in the report. I will print here the next sentence in the same paragraph. "The regulations concerning ACAS/TCAS published by ICAO and as a result the regulationsof national aviation authorities, operational and procedural instructions of the TCAS manufacturer and the operators were not standardized, incomplete and partially contradictory."
So the training and regulations were not proper or were contradictory. This could lead to problems in many areas. I believe the BFU is correct. This is a regulatory issue which I believe has been corrected.

Young man, this is getting really tedious. Let me try to put another perspective on things.

You introduce the notion of responsibility. This is not a foreign notion to engineers, but part of most of the codes of conduct of most professional engineering societies.

The engineering facts which you choose not to like, inter alia that, given the decision presented to the Tu-154 crew, it would have been rational for them to choose to go against the TCAS RA, were discovered, as far as I can tell, by me, and have been published in an engineering organ after peer review by two published TCAS experts, one a Fellow of the Royal Aeronautical Society who liked the paper, and one by an engineer for a major ANSP, who did not like what he read but was unable to construct any valid counterarguments.

It is in the code of conduct of many professional engineering societies that if one becomes aware of significant safety issues with engineered systems, one is obliged to make these issues known at least to the profession and to users of these systems. This obligation falls under the concept of "professional responsibility".
I did that, and continue to do so.

Instead, you suggest this very action is "irresponsible". That is a perjorative term. Were you to be a professional engineer and a member of a professional engineering organisation, you would be censured by that organisation for saying such things publically. For most societies censure the public defaming of work of fellow engineering professionals.

Defaming is or course different from technical criticism, which is generally encouraged. But you have been repeatedly asked over almost six months for arguments contravening the engineering fact that you do not like, and you have repeatedly evaded the question. While repeating your defamation.

If you pretend to be a professional, the only way you can get away with that behavior without censure is that this is an anonymous forum and nobody knows who you are. I think that it in general a good thing, because there are all sorts of facts that people don't like and which they attempt to suppress, and it is good to have a place in which they can be brought up and discussed without political consequences. But by the same token I don't think the forum should be used to propagate abuse.

So I'll just ask you to behave like a professional. If you are not sure what that involves, let me suggest you check out a few codes of conduct of engineering societies which are available on the WWW.

PBL

All I can say in response to the last post is to suggest that anyone(preferably people not associated with myself or PBL) who is interested in this back and forth posting, to read what I have printed and post what they think I have said here is unprofessional.

I don't disagree or necessarily know what what engineering facts you have discovered. There are no facts on this thread or any of your studies that I have ever said I don't like. I assume they are correct.

I strongly disagree with the two posts that I previously mentioned and stand by them. I don't have technical studies to throw into the forum to back up my arguements. All I have is real world results to look at. Publications giving example after example of saves by TCAS, near collisions caused by improper procedures regarding TCAS and near and actual collisions caused by ignoring TCAS. Then of course there is the occasional TCAS related incident.
As for the not responding for six months, I have no technical engineering or studies to give mathematical probabilities of how much safer it is to follow the RA in the midair collision incident. Just real world actions and opinions.

Perhaps you could respond on how to visually avoid that 757(probably seen only as dim lights) closing in at 1000 km/h or more, probably descending as well, and quite possibly against a background of lights. There are no mathematical formulas for this and theory and statistics don't work in a split decision moment like this. Perhaps blind luck does.

Now new terms have been brought in.

Defamation. Which rightly or wrongly makes me think lawsuit in my part of the world and makes me glad for anonymity.
abuse. Please, contact the moderator and ask him to review this thread and give an opinion on what abuse I have given out on this thread and please publish the response. There has been no abuse here. I know there will be no follow up on this one.

Maybe its time to lighten up with this enlightening link.:D

http://www.successunlimited.co.uk/humour/engineer.htm

Punkalouver:For the record : I am not associated with PBL (nor yourself.)
I am also not an engineer, just a simple Controller with some extra knowledge.

I however think that you should open your mind beyond what you read in the Eurocontrol ACAS bulletins. They are meant for the education of the public users , (i.e pilots and controllers) , and provide standard basic solutions to selected illustrative incidents . They are not the TCAS Bible .They do not cover the philosopy of the system , which is what we try to discuss here.
You said :
We only have one in Europe :so let's take it as an example ( because we all know this one quite well by now) I asume you wrote this because you mean the TU154 pilot ignored TCAS ?
Now this is how I see it :
The collision was not caused ONLY because the 154 ignored TCAS.
To occur this collison needed more :
1) To have both aircraft fitted with TCAS ( because with no TCAS there would not have been a collision ) and
2) for the other aircrfat to follow his RA .

The 757 followed his own RAs more or less as the book said ,and still collided.
(so TCAS did not help/protect him at all )
Finally on this point : based on the regulations in force in July 2002 , it was not irrational for the Tu154 crew to follow the ATC instruction.

Next :
Nobody has. Therefore when I see statements like it is 4 times safer to follow RAs , I smile a bit behind my PC. In any case the numbers are far too small to deduct any useful statistics.

Remaining in Europe ,and only looking at facts : since 1970, there were only 2 high altitude collisions before Ueberlingen.( Nantes in 1973 and Zagred in 1976 ) Those 2 collisions would most probably have been prevented if TCAS would have been available then. However in the period 1976-1994 . No collision and No TCAS either. For the period 1994 -2000 a large portion of aircraft at High altitude in Europe were fitted with TCAS , but not all, especially not those of the former Soviet Union States .Again no collision during that period.
But since mandatory carriage of TCAS in Europe ( 2000) 2 years later one collision partlially caused by TCAS.
So, allow me to question the philosophical statement that " TCAS is good for you "
So when you say that ( quoting you ) :
I also smile a bit. We did not have a collision in 26 years despite many , many millions of flights in that period, and suddendly, we now have a system that would have saved dozens of collisons every year ?
TCAS was introduced in the USA without a proper safety case , for political reasons. It was later introduced in Europe for legal reasons (it would not have been sustainable legally for a State to have a collision in its airspace that could have been prevented by a system that was technically available then ).

Now it is its technical ( in abroad sense ) shortcomings that are being discussed.

The moderator thinks that folks should have the occasional ten deep breaths prior to getting too excited and agitated over a discussion ...

Thanks for your reply. I fully agree that there are probably technical shortcomings and PBL is quite possibly aware of them. Have never said different.

It has been over 21 years now since the midair collision over L.A. of a DC-9 and a mode-C equipped Piper. For most of the past 21 years in the U.S. all airliners have had TCAS and there have been no airliner midairs. In the 21 years before the L.A. collision there was:

1) 25 September 1978; Pacific Southwest 727-200 vs Cessna172; San Diego, CA-143 dead.
2) 6 June 1971; Hughes Airwest DC9-31 vs. F-4; Near Duarte, CA-50 dead.
3) 9 September 1969; Allegheny DC9-31 vs. Piper Cherokee; Near Fairland, IN-84 dead
4) 19 July 1967; Piedmont 727 vs.Cessna 310; Hendersonville, NC-82 dead.
5) 9 March 1967; TWA DC9-14 vs. Beech Baron; Near Urbana, OH-26 dead.

Keep in mind that there have been way more airline flights in the 21 since the L.A. collision than the 21 previous years. This reality is what makes me smile behind my PC.

From the final report in part:
The TU-154 AOM contains general recommendations regarding the use of TCAS. The AOM reference draws attention to the fact that manouvering opposite to TCAS can lead to a collision.
The TU-154 AOM stated that manouvering opposite to the RA is PROHIBITED.

One final thought about the rationality of manouvering visually to avoid that 757(against the RA) based on "engineering facts". Exactly what facts were used to determine that one can successfully visually manouver around this fast moving jet. Is it just assumed that it can be successfuly accomplished or is there actual data on how easily it is done(in daytime, at night, over brightly lit areas, in marginal visibility or with different cloud formations in the background and the illusions they can create with false horizons which I have also seen with northern lights, etc.). If so please publish this engineering data.

Like I said I have no engineering data. But I do have real world knowledge of a midair between two airliners caused by improper evasive action by one of them when none was required due to illusion.

http://www.prop-liners.com/midair.htm

And I know that according to a Flight Safety Foundation(perhaps the best known safety study group) publication, a JAL 747 captain figured the could go against the RA and visually avoid a DC-10. He was right. He missed by 135 meters and only injured 99 people on his plane and lodged a galley cart in the ceiling(following an RA is a one-quarter G manouver by the way).

http://www.flightsafety.org/fsd/fsd_mar04.pdf

And finally our much talked about TU-154 captain who attempted to visually manouver by a 757. Results catastophic.
I eagerly await these "engineering facts".

That statement is hard to reconcile with your statement that it is "irresponsible" for me to publish them on this forum.

You list five midair collisions in the US in 11 years to 1978, and then there was one in the next 13 years until TCAS was mandated. It is hard to tell what you want to make of these figures, such as they are. But you do not attempt to account for confounding factors, and that omission is going to invalidate any conclusions you may want to draw.

The most obvious confounding factors are that there were two major revisions of airspace in the U.S. in those thirteen years from 1978 to 1991. The introduction of TCAs and then "Class" airspace had far more effect of separating commercial fast jets from GA traffic than any technical collision avoidance system could have had. If you think the midair collision figures say anything at all, then they say at least that.

PBL

The Piper involved in the Aeromexico 498 collision (31.08.1986) was not equipped with a mode-c capable transponder.

Please provide me with the exact location of your quote. I searched the report and the only occurence of "AOM" does not nearly state anything you quoted.

The important section (which is diametrically to your "quote") can be found on page 53 in the english report, where the TU154M FOM is quoted (although in translation from russian):

To my experience, almost no accident can be traced to one single failure. There are always a large number of occurences, that interactively lead to (near-) disaster. Perhaps you should have a look at Reasons "Swiss Cheese Model". In the JAL-JALincident for example, an important factor was the mix-up of callsigns by ATC.