You mean, apart from the Russian pilots at Überlingen ?

Who sadly didn't survive to tell the tale.

IIRC 6 months earlier A JAL 747 and DC10 came awfully close due to conflicting instructions from ATC and TCAS.

TCAS RAs must take precedence over ATC instructions. Controllers understand that and so should pilots.

The problem occurs when the controller becomes aware of a conflict late and will try to resolve it, but is unaware of the RA. The controller may have a completely valid conflict resolution which may be contradictory to the equally valid RA. The ability of the TCAS to co-ordinate both aircraft simultaneously (or instruct one aircraft with knowledge that the other is not RA capable) with none of the latencies of voice comms means it should always take precedence.

Modern surveillance systems are now able to downlink the RA information during the event. However providing this to the controller is a long way from universal. The purpose is not for the controller to intervene in the RA, but simply to be aware of its occurrence, not risk applying potentially contradictory instructions, and be ready to deal with any consequences involving, for example, reclearance.

@angryrat and Don'thangup: Absolutely.

1) TCAS RA's are a coordinated response between two conflicting aircraft. The TCAS units actually talk to each other and agree a course of action. One says, 'OK I will climb', the other says 'fair enough, I will descend' (or whatever). But the point is there is agreement and a coordinated response. This is far more exact, quicker and safer than a vocal exchange between aircraft and ATC, neither of whom - by definition - have spotted the conflict approaching until it is too late.

2) TCAS targets displayed on the ND or equivalent can be up to 30 degrees in error in azimuth. So are you really looking at the correct aircraft??

3) We can only see forwards from our cockpits and about 135 degrees to the right and left. We have very limited views below, and directly above. TCAS targets can be behind, above, below you or otherwise invisible to you. Many pilots seem only to think that the only conflict will be between two aircraft approaching each other more or less head-on.

4) Unless the sky is gin clear with no cloud, an intruder could very well be hidden from view.

5) As has been mentioned, very few of us have been trained in combat or avoidance manoeuvres (other than TCAS RA's). Even those that have are not in current combat practice, and our aircraft are not nimble fast jets either.

6) (And this is a note to self): There have been many cases of crews mishandling TCAS RA's. Time spent trying to spot an intruder aircraft would be much better spent looking inside at the PFD/ND and mentally rehearsing the RA manoeuvre.

There seem to be a few people on here who are determined to justify not following an RA.
Let's try and simplicate it for you.
Following an RA may result in some minor inconvenience.
Not following an RA may result in your death, and the death of your crew and passengers, and the other aircraft's crew and passengers and maybe even some people on the ground.
Do you want to argue with that?

I do not often read R & N, and I post on it even more rarely. I can now see why some refer to it as Ravers & Nutters.

We were doing TCAS RA's in the sim this week. A dual TCAS was given as a selection for the program, and the student elected to look out the window, and decided not to react correctly. They missed the fact the threat was from below, unseen and not the one they had in sight!
End of story!

Well said.

Why not comply with RA demands? If they're followed in a reasonable timeframe it’s generally a gentle manoeuvre. Do you have 360deg vision above and below your aircraft? No. So just do it.

It’s not just your life you’re playing with, it’s all the others mentioned above. The only time I would even consider not following an RA would be if it was going to result in an immediate impact with terrain and how often does that happen?

If ATC said climb or descend you'd comply, yes? So what’s the problem complying with TCAS RAs, considering they have a higher priority than ATC clearances? There should be no confusion here.

Next, we'll be arguing about ignoring EGPWS warnings.

Yep, and the earlier version (not "E") saved my butt as well.

AirBlue A321 at Islamabad, Sukhoi demonstration flight at Mt. Salak...

Sure. There are rare, quasi-legitimate reasons for judging the veracity of these warnings. But the percentages are simply not there to do so.

I don't recall any reports in the literature that state that a mid-air collision was prevented by ignoring TCAS and following ATC instructions. We have the one case that demonstrates the opposite.

This is a by-the-numbers industry. The data just isn't there to support ignoring the warnings even when one thinks one's own judgement in-the-moment, is superior. One is just throwing dice by ignoring the engineering and the SOPs behind these systems.

Just to nitpick but RAs are not always coordinated. Coordinated RAs may happen if both aircraft are TCAS II equipped. But there are plenty of aircraft without TCAS II (and perhaps only Mode C/S).

If the other aircraft doesn't have TCAS II, an uncoordinated RA may be generated. That means there's a higher chance that the other pilot might make a conflicting evasive maneuver. Bottom line is to always be prepared for additional RAs (reversals, etc.)

I still can't believe anyone who flies professionally is still advocating not following an RA.

This argument about ex mil pilots seeing and understanding better is a total red herring and shows a total lack of understanding of accident avoidance.

TCAS IS THE FINAL DEFENCE AGAINST A MID AIR COLLISION AND AN RA MUST ALWAYS BE FOLLOWED.

If you have a TCAS RA, 1) ATC have failed to separate you. 2) Your SA ref listening to other ac if on freq and knowing where they are has failed to separate you. 3) You have failed to acquire and visually separate yourself from other ac. 4) Every other method of separating you from other aircraft has failed.

Yes there is a possibility the RA has been caused by a high ROD/climb and if both aircraft adhere to their clearance separation will be maintained but you cannot assume that. Follow the RA.

TCAS is inhibited below 1000ft, and there is an SOP to not follow a TCAS descend RA below this height. A TCAS RA is not a violent manoeuvre.

Less well known is the reason behind the SOP radio call of "xxxxxx TCAS RA" which should be made in a timely, clear and unambiguous way when an RA is initiated. It is an instruction to ATC (and other AC) to cease all transmissions until the clear of conflict call is made to allow the manoeuvring crew(s) to concentrate on complying with the RA and prevent conflicting ATC avoiding instructions.

Again, in my best wide body jet shouty instructor / examiner voice

ALWAYS FOLLOW A TCAS RA.

You are in the last chance saloon and failing to follow an RA has killed far more people than it has saved. TCAS 2 is very reliable and rarely malfunctions.

Captain refused to follow RA. Said he saw the traffic and it wasn't a threat. FO's screaming that there are two targets. Captain was looking at the wrong one. Near miss!

In a RA there are always two or more players and RA is issued to all to take them through. Any body who wants to disregard RA without reason must also be clairvoyant to know the other aircraft's response. There may be RA reversals for them or he may also disregard his RA leading to chaos and dangerous situation. Only time you can consider is with a EGPWS alert.

I have never seen any GA aircraft without transponders at FL360...

...but then again, it could surely happen.

Airmanship and SA goes hand in hand. No SA, outside controlled airspace and with more than one target on TCAS, sure I will follow it, but if only one target at FL360 in controlled airspace and visual (since 80 miles in the contrails) plus maybe SA from ATC communication, I will assume the responsibility and ensure positiv separation by creating line of sight to the other aircraft.

Some guys seem to have loss the will to survive. It's like my wife; I issue an order camouflaged as a suggestion; she thinks about it and tries to find a reason not to agree or come up with an absurd alternative. This takes quite a few moments, and my response is pointing out her counter suggestion is not practical. After a few ping-pong comments we agree, finally, that my original idea will work as the best option and off we go.
Now have this discussion with a closing speed of 16km's per minute. There comes a time for someone to make a decision PDQ. If you want to think about it afterwards, feel free. At least you will have the luxury. If you don't...............then you won't.

Peekay4;

Just to watch how the system handled multiple TAs/RAs, we used the simulator to put several aircraft above & below our "aircraft". The TCAS handled normal RAs & reversal RAs well.

Indeed, one must be prepared for additional RAs.

And, as the literature states, the manoeuvres do not require brisk or high 'g' loadings.

I'm with F16GUY on this. TCAS RA's should be generally followed, no one is arguing this. However, if the situation is obvious (eg. VMC, one target only and one can unambiguously determine it is not a threat), are we supposed to follow, just because the computer says so?!

Same for eg. spurious GPWS warning in VMC (had it 2 times), "long flare" warning that got the EK777 into trouble at DXB, or the instance wher the EK A380 couldn't land at MAN and diverted, because the ROPS system said "no", even though the rwy length was more than sufficient. I feel sometimes we are doing things contrary to common sense, just to make the computer happy.

Sidestick...

Again, we're back to data, which is the basis for operations. The exceptions cited are rare.

Spurious warnings mandate a report so it can be fixed. If a ROPS system is indicating "no" when the runway is deemed sufficient, then fix the ROPS system so that it provides accurate information. (I was under the impression, perhaps false?, that the EK777 fell back to the runway because the PF/crew had not applied go-around thrust)

Yes, these systems generally work well and I am not trying to advocate ignoring them.

The exceptions cited are quite rare, but far from remote. Every now and then it's plainly obvious the computer has gone haywire, or its output is not relevant under prevailing circumstances. What I'm against is following the computer slavishly, even when it obviously doesn't make sense. The FAA stance on TCAS quoted earlier seems to be in accordance with that.

As for the EK777 at DXB. Yes, it was a mis-handled g/a. However, the whole event was triggered by "long flare" warning issued at ~1000m deep into a 4000m runway. Was it a cause of the accident? No. A contributing factor? Yes, in my opinion at least. What would have happened, had the crew continued the landing? Nothing.