Avoiding action

[squeaker]

Just a quick question, I was given avoiding action yesterday, to prevent a loss of separation more than a collision I think. The instruction was " avoiding action ,descend now 3000' ". This was followed by a TCAS TA "traffic traffic" as we started to descend, and we saw the other a/c pass down our left side at least 3 miles away. No problem, but I just wondered if you guys have been given any new instructions about giving avoiding action as a climb or descent, rather than a heading or turn, since the Swiss midair?
I am obviously referring to the potential for a conflict between a TCAS RA and an ATC instruction. I am in no way criticising the ATCO in my case.

[LostThePicture]

This little chestnut was discussed quite a bit on the last two pages of this thread. (http://www.pprune.org/forums/showthread.php?s=&threadid=98090&perpage=15&pagenumber=10) There are many, many ATCOs (including members of the training staff) who would maintain that vertical avoiding action is the way to go, because it's the quickest way to restore lost separation. I would disagree with this course of action, purely because of the confusion that could be caused if you are getting into a TCAS encounter (as is likely, if avoiding action is being given!). Turns are the way to go.

LTP

[Jerricho]

I agree with LTP. Especially in a busy terminal environment, where several SID's can interact with each other (or in a hold), "level" avoiding action can have a major knock on effect.

[FWA NATCA]

Squeaker,

Avoding action instructions are determined by what action will best resolve the situation, safest and quickest (based upon non TCAS aircraft). This could be an immediate descent, an immediate climb, or a combination with turns. It doesn't do anyone any good to give you an immediate instruction that resolves one confliction to only put you into another situation.

When you factor in TCAS, it gets more complicated because we don't know what instruction that your TCAS is going to tell you to do. There have been many documented cases of TCAS giving instructions that resolves the initial confliction to only put the pilot in a worse situation with another aircraft.

I've had pilots receive TCAS resolutions (on traffic that did not exist or was not a factor) that had they followed the instructions of TCAS they would had flown into the ground or into other traffic.

I support TCAS and I think that it is a great safety aid, but it scares me to have a pilot take an action that can potentially make the situation a whole lot worse.

Mike
NATCA FWA

[BALIX]

I support TCAS and I think that it is a great safety aid, but it scares me to have a pilot take an action that can potentially make the situation a whole lot worse.

It scares me that as a controller I can instruct a pilot to take an action that will make things a whole lot worse.

From reading various publications (don't ask me what they are) that we are going along the lines of lateral avoiding action being the preferred course of action. A lateral turn does not conflict with the vertical instructions given by TCAS and so should be less confusing for a pilot in what is a high stress situation.

[Spitoon]

Interesting doc (http://www.caa.co.uk/docs/33/CAP717.pdf) on the CAA website.

[West Coast]

Mike
A GPWS alert takes precedence over a RA. If they responded to a RA that sent them down, the subsequent GPWS would have them climbing again. I had a RA a few years ago in Reno. It had us climbing twords VFR traffic that also gave us a RA. Simply leveled out for a second. No big deal.

[LostThePicture]

Yes, an interesting document. There is a case for parts of it (the bits about vertical avoidance instructions possibly conflicting with TCAS RAs) to be included in the MATS Part 1, either as part of the TCAS Supplementary Instruction, or elsewhere.

Not a flawless document however. I had to gasp when I saw example 5.5. "Consider increasing the descent rate of the descending aircraft - inertia may be such that an attempt to level off could make things worse." What about the TCAS?! Conflicting vertical instructions can make things worse - recent history has shown that. And the diagram in the example is hardly complementary to the caveat stated in section 2.3; "This can result in the aircraft's position presented to the controller on the radar display being three quarters of a mile out of date immediately prior to the next update." :eek: :uhoh: :{

And the stuff trotted out from the Rules of the Air is meaningless as well. Totally inapplicable to a radar control environment. Common sense, together with an appreciation of the prevailing traffic situation, is what is required.

I find it shocking that, fifteen months after the Lake Constance mid-air, there are still no formal guidelines set out for British controllers on the use of avoiding action in conjunction with TCAS. The MATS 1 is updated every three months, so there have been at least four opportunities to have made the necessary amendments. Unfortunately, until the amendment occurs, there will still be a great number of controllers (and OJTIs) who will continue to issue (and teach their students to issue) potentially unsafe avoiding action.

LTP

[Fox3snapshot]

It would be nice to be working for a unit where TCAS resolution and ATC instructions are formally briefed and trained for.....sadly this is not the case at many units and the Swiss contol situation will be repeated....trust me!

:uhoh:

[squeaker]

Thanks for the replies, and a good link to the CAA doc.
I would say that our training tends towards "follow TCAS", and I'm pretty sure that the more we practice flying RAs on our base checks/LPCs, the more we will become conditioned into doing that automatically. Obviously, we also practice the call to ATC of "TCAS climb/descent", but I wonder how often this results in a crossed transmission i.e. the aircraft calling at the same time ATC is trying to give avoiding action instructions. Would it be possible to incorporate a system whereby the transponder automatically indicates to the ATC controller that an RA is in progress, and which aircraft are involved in the TCAS contract? I'm thinking along the lines of a similar display to ident, or maybe a colour change.

[Jerricho]

I'm sure in the not to distant future with the advent of Mode S and data-link, controllers will be able to see a vertical profile of jets, that may indicate a TCAS manoeuver. Or even highlight by flashing (kind of like our STCA) an aircraft receiving a TA or RA? I'm sure some of our Mode S guru-types (like Scott) would be able to shed more light on this.

But Sqeaker, you're right, sod's law dictates that while you're making your "TCAS climb/descent" call, the controller may be crossing with you with avoiding action. There has been a change to the phraseology for avoiding action, requiring the use of the callsign twice at the start of the transmission. The rationale I guess is to lessen the chance of the callsign being trod on.

[whowhenwhy]

I was always told that, if you can, give avoiding action laterally and vertically. 2 bites of the cherry and all that

[Roger Dodge]

As metioned in previous posts, turns are normally the best way to resolve the confliction in a TCAS environment. This, however, should not preclude the ATCO from using their judgement in initiating a climb/descent instruction to help the situation.

We all understand that if an a/c is responding to a TCAS RA then there is nowt we can do in the vertical plain. If you resolve the confliction before TCAS kicks in then using vertical is a useful tool.

[FWA NATCA]

West Coast,

Don't forget that when you are responding to an RA, I am NOT allowed to issue you any control instructions that could be considered as contrary to what the TCAS RA is telling you do do.

As a controller my only option to when you respond to an RA is to issue traffic, and hope that you don't wack another aircraft.

Mike
NATCA FWA

[squeaker]

Mike,
would a radar heading be deemed contrary to a TCAS RA, or a useful belt-and-braces means of ensuring a miss?

[FWA NATCA]

Sqeaker,

If you are responding to an RA, I'm not supposed to issue any turns, climbs, or descents. My only option is to issue traffic, then sort out what happened.

Another interesting issue with TCAS is when I issue traffic on non-transponder equiped traffic using the phraselogy, "traffic two o'clock 3 miles, north bound, altitude unknown." The pilots will sometimes say they see the traffic on TCAS (which is impossible since the traffic doesn't have a transponder), and quit looking for the traffic.

Mike

[Scott Voigt]

Squeeker;

One problem with issuing a vector that close in is one, the vector probably isn't going to work if you are close enough to get a TCAS RA. Two, you don't know which way to really go for the fastest escape manuever, radar is NOT that accurate in the enroute environement thus the reason that we need five miles of separation. We could give you a turn to the right and actually be taking you closer to the aircraft instead of further away in the final few seconds. Altitude is a VERY good thing. You don't need much of it and you can get it VERY quickly up or down.

regards

Scott

[jack-oh]

When providing avoiding action to a military fast jet it is often better to go for height separation rather than a turn as anyone who has seen a GR4 trying to turn at high speed will testify to. Getting at least 3000ft is sometimes far quicker than getting 5nm. Equally, providing avoiding action within a known traffic environment indicates that something has gone wrong with the system already. Outside of CAS avoiding action is aimed at, where possible, achieving 3000ft or 5nm from unknown ac. These ac may or may not have TCAS, which may or may not agree a safe course of action based on predicted profiles of both ac. Additionally, the conflicting ac may or may not have SSR at all. Furthermore, TCAS and radar separation are based on two entirely different principles, an ATCO may give avoiding action at 10nm knowing the turn or climb will achieve 3000ft or 5nm, TCAS however will only kick in based on time to collision and profile, this may or may not be at the same point the controller issues the instruction. I personally see no conflict between the two systems when TCAS it is used as a last ditch effort to prevent a collision. However, TCAS does fall down when RA's are generated between ac that are climbing and descending to cleared levels, which provide separation, but due to the ac's conflicting profiles and the systems inability to grasp this simple ATC concept, false or spurious RA's are generated. These RA’s may place the ac in question into far greater danger; the machine may be good but as with everything, people are involved as well.

[AirNoServicesAustralia]

Considering the lag factor on our radar information (ie. the altitude we have displayed is usually a few hundred feet behind what the aircraft is really at if they are on a fast descent or climb) we could quite easily make the situation worse by issuing a vertical avoiding instruction. So I agree with the guys who said a hard turn to both aircraft is all we can do (and cross our fingers and toes). Although have to admit that not having been in that situation I really don't know what I would do until it happened (and I sincerely hope it never does). It would be easy to throw everything at it, ie. climb/descend/turn left/turn right/ traffic is etc etc etc.

[West Coast]

Mike
I accept that you are being taken out of the loop for a matter of time as I respond to a RA. The limitations of radar make me believe the current policy of non intervention and traffic calls as the best response to an iffy situation

[radar707]

Just as an aside to this discussion,

As has been mentioned about the limitations of radar and non transponding aircraft.....

How many times have controllers issued avoiding action instructions against a weather return???

[av8boy]

WC,

Well, there's non-intervention and then there's non-intervention...

And I quote, "Once the responding aircraft has begun a maneuver in response to an RA, the controller is not responsible for providing standard separation between the aircraft that is responding to an RA and any other aircraft , airspace, terrain or obstructions." This is from the 7110.65. The AIM, while ostensibly saying the same thing (or one would think), seems to soften the language: "The serving IFR air traffic facility is not responsible to provide approved standard IFR separation to an aircraft after a TCAS II RA maneuver until [the end of the event]."

What the AIM is telling you is that the controller is under no obligation to provide IFR separation to the RA aircraft. What you need to read into it is the fact that it isn't just talking about the aircraft involved in the RA. It is talking about any and all IFR aircraft which may find themselves in the proximity of the RA aircraft.

So, now what? In a busy terminal environment tolerances can be pretty tight. Although it appears that I've got no legal requirement mandating that I separate this RA aircraft from others under my control (short of traffic advisories), is that going to eliminate the pucker factor? I think not.

When I think about this it always puts me in mind of an evening at what used to be Clark AFB in the Philippines. We landed, cleared ops, and were on our way downtown to the hotel. As we approached the front gate of the base, two USAF sky cops were running toward us waving their arms. "Get in the ditch! PAF is shooting!" they yelled. See, somebody was trying to sneak into the base by going through the fence, and the Philippine Air Force personnel were shooting at them. The USAF guys were just trying to limit the collateral damage to people who were not involved in the incident.

My RA fears run parallel to this experience. The aircraft involved in the RA are focused on resolving the conflict between themselves, and I'm left to try to get everybody else into the ditch (whether or not I am required to do so). What's more, I'm still going to try to help out in the RA to the extent I can. It makes my ulcer cry out.

Don't get me wrong. TCAS is an outstanding tool and it is needed. It's just going to take a long time to get these issues worked out.

As has been mentioned about the limitations of radar and non transponding aircraft.....How many times have controllers issued avoiding action instructions against a weather return?
I don't know about "avoiding action," but as far as calling traffic on something that turns out to be wx or ground clutter, not a problem. Far better to err on the side of caution.

Have at me. I want to flesh this out.

Dave

[West Coast]

Dave
Thanks for the post. I understand your point, but think of mine. There may be other aircraft out there, but the TCAS is telling me the one pointed at me is the one I may prang if I don't do something very quick. I respond to the RA and continue to search for other traffic. The EGPWS will hopefully keep us out of the rocks if the manuever takes me down. Your traffic calls and scanning along with additional TA/RA information from the TCAS will hopefully keep me away from the others. The procedures on our side have been refined over time to minimize the number of RA's we respond to. It used to be even if we saw the intruder we had to respond till clear of conflict was sounded. Now if we see the intruder prior to an RA and determine him not to be a threat we don't need to respond. If not in sight, we respond and if we get visual and determine it not a factor, then we can discontinue the manuever even if the box says otherwise. Happened at DFW a few weeks ago. Saw an eastbound departure climbing to 10 as we were at 11. His track would have in no way affected us. We recieved an RA, we both said in sight to each other and that was it. One of the places a TCAS is invaluable is in terminal areas keeping us aware of VFR traffic in the area. If for nothing else I like it for that option.

Always nice to hear a controllers views on things

[av8boy]

Trust me, I AM thinking of yours WC. That's why the post! There are simply some details left to work out when it comes to TCAS.

Just so we're clear, I am fully supportive of TCAS. You, me, and it, we're all after the same thing: avoiding metal to metal contact by really delightful margins. I only get up on this soapbox to make sure the conversation continues.

You ever read the obituaries in the New York or London Times? You ever think to yourself, "Self, I wish I'd known this about this chap before he went and died. Why don't they ever share this concise info before the chap expires?" Well, that's what we're doing here. I feel the need to flesh out as much as possible BEFORE somebody dies.

So, I want this to be the last time you ever say something like, "If for nothing else I like it for that option." You like it for all the things it does for all of us, and I like it to. Don't sell it short. Just be painfully critical when appropriate. We'll end up with a reasonable approach to this issue yet!

Dave

[Spuds McKenzie]

TCAS is a last resort tool. If not followed, then the result could be devastating.
However, as an ATCO, after Ueberlingen and Japan, I'm ever more in doubt, if the regulations concerning TCAS are clear enough.
As mentioned previously, what we know is, that when TCAS kicks in, ATC is out of the picture regarding separation of aircraft, BUT this requires the crews of the aircraft to notify ATC IMMEDIATELY, that they are following an RA. This DID NOT happen in the Ueberlingen case.
In fact it took the DHL crew 25 secs to do so, and, as we know, the TU crew didn't follow the RA at all. Now, within those 25 secs, the ATCO issued a descent clearance to the TU crew twice, since he was still in charge of separating the aircraft, as he didn't know of any RA being followed by one of the crews. Therefore his decision, based on the knowledge he had at the time, to let the TU descend, was APPROPRIATE.
So, as an ATCO, should I ever encounter a similar situation, how am I supposed to act? Tell the pilots to follow the TCAS RA, not knowing if they actually DID receive an RA as yet? Use lateral separation only, which can be tricky when you have a head on conflict (turn them right or left)?
To me, the solutions have to be as follows:

1. RA's have to be followed, no matter what (and most of all, ICAO has to implement regulaions accordingly, AT LAST!)

2. ATC has to be notified IMMEDIATELY

3. There has to be a link between TCAS RA indications in the cockpit and ATC Radar screens. If squawk, altitude and speed can be transmitted electronically, why not other information.

4. And last but not least, all of the above has to be implemented ASAP!

Remember, the Japan incident, which resulted in an extremely close encounter, and the Ueberlingen accident are 18 months apart.
Within those 18 months, ICAO did NOTHING to clarify the issue concerning TCAS RA's, although the Japanese authorities advised them to do so.

The whole issue remains, to a certain extent, a gray matter until something is being done about it.

[FWA NATCA]

West Coast,

Since I work at an Approach Control I rely on only one radar site for my radar information, where as an en route center relies on multiple radar sites (mosaic). In addition my radar spins faster than an en route radar site so my updates are more often thus more accurate.

I've had aircraft react to RA's when there wasn't any observed traffic, when this occurs, I tell the pilot no traffic observed, the majority of the time they still respond to the RA.

Being out of the loop (when a pilot responds to an invalid RA) when I'm working busy traffic, and running things tight tends to increase the stress level because the chances of the pilot climbing or descending into other traffic is great.

Mike
NATCA FWA

[West Coast]

Mike
I understand, but what other options are there out there that are viable?

[FWA NATCA]

West Coast,

At this time due to the airline requirements, and our regulations, I don't know of any other options. The pilots hands, and controllers hands are tied.

Mike