PDA

View Full Version : ACAS v ATC


StillDark&Hungry
4th Aug 2005, 08:16
Have just been reading a very interesting article in "Feedback" about following ATC instruction or ACAS, which is still a hot topic after the Swiss accident.

Being non-confidential I won't copy the whole article here, but, for those who didn't, haven't or can't get a copy it concerns an RT exchange between controller and pilot over Northern Europe during which an "Immediate" descent restriction was given followed by traffic information and an instruction to ignore any instructions given by ACAS!

Although it made me wince on first reading it the article does raise an interesting question - that being at what point, after noticing a potential loss of separation does the full responsibility remain with us to "save the day"?

In en route, for example, aircraft are travelling at upwards of 8 miles a minute and any restriction or instruction issued is going to take time to happen, the laws of physics state that.

Here's an example;
2 aircraft on crossing tracks, converging, at the same level. If the aircraft are, say, 30 miles from the crossing point when I notice the confliction I would have no hesitation in issueing an instruction to move one of the planes.
If the same planes were 20 miles from crossing I would still issue an instruction but may sound more urgent with my delivery!
If the planes are 15 miles apart, I hope I may even give traffic information and "avoiding action"
If the planes are 10 miles apart my instinct would be to do the above but now, in the back of my mind is the nagging suspicion that - as I can only talk to one of them at a time, if I do give avoiding action to the first, has the second began to respond to ACAS already?

It's a fine line but it's position is unclear. I fully understand that once issued with an RA a pilot must follow it, even if it contradicts ATC, but when exactly will that happen?

Seat1APlease
4th Aug 2005, 08:38
TCAS collision avoidance
ACAS industrial dispute/strike avoidance?

VectorLine
4th Aug 2005, 08:50
ACAS = Airborne Collision Avoidance System (Generic Term)

TCAS = Traffic Collision Avoidance System (Specific type of ACAS that provides traffic information)

Seat1APlease
4th Aug 2005, 08:57
Sorry, you learn something every day.

Now if I can only remember my eadi from my ehsi and mcp and gpws and cdu and pmc and gca metars etc.,

-- I think I will go for a lie down in a dark room!

Pierre Argh
4th Aug 2005, 09:48
stilldark&hungry
once issued with an RA a pilot must follow it, even if it contradicts ATC, but when exactly will that happen
Did I misread you... I don't think its difficult to imagine such a scenario? Controller notices two aircraft close together at the same level, acts by issuing a climb clearance to pilot A... but the airborne systems in both aircraft have "co-ordinated" their own avoiding action, giving pilot A a descent instruction, whilst Pilot B is told to climb.

Pilot A is compelled to take the descent.

There are interesting spin-offs too on the subject of TCAS, and at the risk of hi-jacking your point... Eurocontrol recently published a document advising pilots against using ACAS as a total avoidance system. It works vertically, but the display is not optimised for interpretation of the horizontal picture (it gives relative motion, with no indication of actual speed or heading)... There is concern that some pilots are using ACAS/TCAS inappropriately i.e. for separation not limiting it to a collision avoidance tool, and should rely on ATC for that.

DtyCln
4th Aug 2005, 15:11
Follow the TCAS Resolution, end of story.

Even if the resolution is opposite to what ATC instruct, go with it.

Advise us, ATC, you are following the resolution as soon as possible. As soon as you advise us you are following the TCAS we will only issue traffic information, we are prohibited from issuing lateral or vertical instructions once you advise us that you are following a TCAS.

:ok:

StillDark&Hungry
4th Aug 2005, 15:16
PA

Controller notices two aircraft close together at the same level, acts by issuing a climb clearance to pilot A... but the airborne systems in both aircraft have "co-ordinated" their own avoiding action, giving pilot A a descent instruction, whilst Pilot B is told to climb.

Pilot A is compelled to take the descent.


No. I don't think you misread me at all. That is exactly my point, except. I issue my climb instruction to aircraft A who, because he's yet to know any difference, initiates a climb. The scenario i'm dreading is, what happens during the next couple of seconds when, say, he's applied the power - just pulling back - when - Aircraft B gets's a TCAS climb instruction. A is told to descend even though he's started a climb!

It's complicated and probably the timing could never happen. But what I'm trying to get at is;

A) If ACAS gives an instruction the best ATC can do is give TI

B) If the aircraft aren't close enough yet for ACAS to give an RA then ATC give avoiding action.

Question;

At what point does B) happen, and, what happens during that little gap between A) & B)?

Okay it's been a long day! Perhaps what I've just written will make sense in the morning!

Pierre Argh
4th Aug 2005, 15:46
SD&H...

I think TCAS works faster than you or I can... and the RA is only issued when collision is imminent. If you have given a climb instruction and the pilot has a TI he should act on your instruction... If its closer and the pilot has started to act on your instruction, modern systems should take that into account in the RA and issue a corresponding manoeuvre to the other aircraft (and probably an increase climb to yours)

If the pilot is still reacting to your instruction but not made any control input, the system might generate an opposite RA. Lerts have a look at an account of the Uberlingen accident immediately prior to collision...

Within seconds of getting their initial TCAS RA to climb, the Swiss Skyguide controller instructed them (the Tu-154 crew) to “descend Flight Level 350, expedite, I have crossing traffic.”

(They) did not respond and 14 seconds later Skyguide told them to “…descend Level 350, expedite descent” at which time (they) acknowledged the instruction and, contrary to the onboard RA, complied with the ATC instruction to descend. Thirty-one seconds later the Tu-154 hit the DHL aircraft

This was not the result of slow reaction... I believe the Tu-154 pilot's action was the result of lack of standardisation of response to RAs... (their national procedures was obey ATC instructions...) so they continued to descend against the RA advice. This loophole has since been resolved by Intl agreement.

The fitted equipment at Uberlingen was also early generation kit.
Later systems will monitor the pilots' actions and even if they act to the contrary, will continue to give RAs designed to avoid the collision, based on the updated positions of the aircraft.

In the account you'll note there is time after the initial RA for the controller to pass two sets of instructions to the Tu-154 crew before they initiated the manoeuvre... and that over 45secs elapsed from first trigger to collision. I don't think there is much chance of you issuing a simultaneous, conflicting action that cannot be countermanded by ACAS and that does not allow the pilot time to react?

A Google search comes up with many different sites that discuss the how and why of ACAS/TCAS that you may find interesting?

BOAC
4th Aug 2005, 16:01
I think you can 'relax' a little, SD&H, as Pierre says. TCAS will respond very quickly and should issue the approriate manoeuvre instructions. Please carry on trying to deconflict us! I would think that if it has got to a TCAS resolution the time has long since passed when you can deconflict with your head spin rates etc. I would suggest the best you can do in your 'sitiuation' is to offer traffic advice - clock position etc?

A I
4th Aug 2005, 16:38
Don't forget the time lag between the moment that the radar return is made to its display on the radar console. With modern radar data processing systems this can be significant. I believe it can be up to 15 seconds meaning that the aircraft have travelled quite a distance. Radar is an historic picture not an actual one. ACAS is much much faster than this.

A I

atcea.com
4th Aug 2005, 17:13
In the U.S. it's quite clear:
7110.65 (http://www.faa.gov/ATPubs/ATC/Chp2/atc0201.html#2-1-27)
"When an aircraft under your control jurisdiction informs you that it is responding to a TCAS Resolution Advisory (RA), do not issue control instructions that are contrary to the RA procedure that a crew member has advised you that they are executing."
---

ATC Chat (http://atcea.com)

Lon More
5th Aug 2005, 09:35
SD&H I couldn't find the report. Have you got a link?

VectorLine
5th Aug 2005, 10:26
Lon More

Here is a link (http://www.chirp.co.uk/new/Downloads/html/FB75.htm#ACAS_vs_ATC_-_A_Reminder) to the report in CHIRP Feedback.

First time I've been to the CHIRP website and I must say, I like being able to look back through all the issues. They are available in three different formats too!

StillDark&Hungry
6th Aug 2005, 19:27
Didn't realise myself that CHIRP had gone electronic!

atcea

Don't disagree with your quote, totally the opposite in fact - but that's not the point I was trying to make.

A1

In a way the time lag problem just adds to my point.

BOAC

Wilco! chiilling out as I speak

cresslime
7th Aug 2005, 01:29
Just another link about mid air collision at Uberlingen, it's from spanish ATC magazine. ( english version available).

http://www.atcmagazine.net/35-seguridad.pdf

looks like interesting.Regards.