Offsetting - time to make it official just about everywhere

[ARINC]

For ARINC - the big problem with our prehistoric 747-200s, is that they have no FMS... After initial positioning of the LTN-92 INS units, we XFILL waypoints to waypoints manually to the other 2 units. Works great, accurate and no problem for lateral NAV, update by GPS... In the past, before the GPS update, it was DME update... But VNAV functions are inexistant... So the autopilot will track with accuracy, but climbs or descents require human brains for planning. The 747-400s are an entirely different breed... You are lucky...

Thanks BeLArgUSA must be quite a performance and as you mentioned very easy to make a mistake.
For the record I only remove/refit the boxes, I don't fly by them...

[Basil]

GMS,

near encounter between a Turkish Airlines Airbus and a canadian Airbus on a NAT track

Don't recollect that one.
Do you know when it happened and which authority carried out the investigation?

With thanks, Bas

p.s. Used 1nm R offset for years - in appropriate circumstances.

[GlueBall]

Too many bored pilots getting creative is dangerous

And what's next. . . ? Offset your altitude? I've flown with one whacko copilot who thought it would be a good idea not only to fly a parallel track, but to offset our assigned flight level by 150 feet.

[MTOW]

It would seem you are indeed a "voice in the wilderness", 410, and not very many here agree with you that something should be done. As someone has already said, it would seem to suggest that most of us have total trust in the system we all operate in.

I wonder if the Brazilian crew were among that great majority?

[India Four Two]

MTOW, do I detect a slightly ironic tone in your post?

I have followed this discussion and the Brazilian crash thread with interest and have some observations about offsets. Before someone jumps all over me, I should state that I am not a commercial pilot, but I do have IFR/airways experience and I have a lot of experience using GPS not only for navigation, but also for precision surveying in the oil industry.

For all those people who worry about dialling in a 1nm offset, a 0.1 nm offset would take away ALL the risk of a head-on collision, even between two A380s, under worst case GPS accuracy.

Although I imagine seeing an A380 closing at 1000 kts and passing with a wingtip separation of 100 m or so might be unnerving to say the least, but you would miss each other.

[FullWings]

We're getting there but the pace of adoption is very slow. I've been flying offsets for a long while but with the widespread use of GPS I think the potential dangers of not offsetting are getting greater all the time. Gone are the days when you could count on random differences in track-keeping and altimetery to increase separation: we're all flying to within feet of each other now.

The NATS is probably one of the safer parts of the sky as it is closely monitored by what I would term "competent authorities" but even there, it's a good idea to use the SLOP as a backup, just in case. If I'm not on a direct routing or in a positive radar environment, I reach for the offset function on the FMC; there's even an argument for this when on a "direct" as others may be getting exactly the same coming the other way...

I operated down to Oz for the first time last year and applied an offset - informing ATC as the regs. suggest. They seemed slightly upset by this as if it somehow reflected on their controlling standards - I replied that there was no radar and they were relying on position reports, so I felt fully justified in going a bit to the right and it made me happier, which is what counts.

[410]

ATC ... seemed slightly upset by this (offsetting) as if it somehow reflected on their controlling standards

I've had the same reaction from ATC, once in Europe when I asked for an offset after the controller asked me to fly for some time at a non-standard level, (because he had too much *** traffic in his area!!!!!), and in Oz before the current procedures allowing SLOP were in force.

I wish we could get the message across that it's not meant as an insult to the ATCOs' competency or professionalism, but (if I may mix my metaphors rather horribly) just an attempt to put another slice of Swiss cheese in the way to help prevent an unbroken error chain.

As has already been said on this thread, I suspect 99.99% of us would never have even heard of the dreadful accident in Brazil if even one of the two crews involved had used this so simple procedure - a procedure that takes all of three seconds to implement (with the confirmation process included).

We wouldn't have heard about it because it wouldn't have happened. At the very worst it would have been an Air Incident Report - IF one of the crews had noticed the lights of the other aircraft as it flashed by. (Even in VMC, it's more than likely neither crew would have seen the other aircraft because if you're on a collision course, there's no relative movement, and relative movement is by far the easiest cue for the eye to pick up.) In IMC and with no TCAS, they probably wouldn't have even seen the other aircraft.

It's been said before by others, but it's worth saying again. I can't understand the resistance to the idea from (it would seem) the majority of line pilots. Are there so many of us out there with such incredible lack of imagination as to not be able to see that any one of us could one day find himself in the same situation as the Brazilian 737 crew?

- obeying every rule in the book
- flying exactly (there's that word again) to flight plan
- trusting ATC and everyone else out there is doing everything right?
- and someone else, over whom you have absolutely no control, makes one solitary mistake.

All it took in the case of the Brazilian 737 was for two holes in Professor Reason's Swiss cheese model to line up.

[Basil]

They seemed slightly upset by this as if it somehow reflected on their controlling standards

Don't ask, don't tell. (Military term - don't ask for the origin )
If you are under procedural control there should be no traffic within miles and if under radar then the radar controlled traffic should not be vectored into you.

[bubbers44]

To avoid nav errors it seems that the only way to avoid input errors for offset would be to have the software automatically put in an offset unless you override it to reduce workload. Missplacing a decimal point could cause a big problem. We don't need to increase workload to use offset, just make it automatic.

[Wiley]

Are there so many of us out there with such incredible lack of imagination as to not be able to see that any one of us could one day find himself in the same situation as the Brazilian 737 crew?

It would seem the answer to that is a resounding 'yes'.

I'm an unapologetic offsetter and have been for years. However, from what my FOs tell me, I'm one of the very few captains in the company who does use the procedure, even though it's authorised for particular areas in our SOPs. I've even had one FO take the offset out while I was out of the cockpit on a toilet break because he believed we should be 'on centreline'. (And for 411A's benefit, I suppose I should add that that particular FO is unlikely to do so again, at least with me.)

[Semaphore Sam]

Besides the Brazilian tragedy, can people think of others that might have been avoided with the offset technique? Possibly BA's climbout into an aircraft over Eastern Europe (Yugoslavia?) in the early '90's, or the Saudia/Russian Delhi accident? This idea seems inherently sound, especially in 126.9 territory. Sam

[Andu]

A few that come immediately to mind, (I think there are others).

A couple of corporate and commuter aircraft accidents in the US, one, I think, in the Grand Canyon area, all resulting in death to all on board.

The USAF C141 and the Luftwaffe Tu154 off the west coast of Africa. All killed.

The Saudia 747 head on with the Il76 just west of Delhi. 400+ killed. (Probably not a good example, as neither aircraft was equipped with IRS or GPS, but it was the accident that started the campaign for offsetting here on Pprune way back when.)

I know of three very near misses head on with opposite direction traffic in my own company, all due to altutude busts, (one out of Dhaka, one near Muscat and the third between Hong Kong and Bangkok) that would have been a lot less exciting for all concerned if the opposing aircraft had not been absolutely aligned. In all three cases, the miss was very near indeed.

[forget]

......... near encounter between a Turkish Airlines Airbus and a canadian Airbus on a NAT track. ------ Don't recollect that one. Do you know when it happened and which authority carried out the investigation?

The A330 commander's report. Both aircraft were in clear air as the A330 was slowly overtaking the A340 below it. The A330 commander stated that his aircraft was slightly to the right of the A340 and almost abeam it when he saw the A340's wings start to flex. At about that time he felt a bump, which he described as similar to entering a mountain wave. Five to ten seconds later there was another bump during which the A330's altimeter reading descreased by 200 feet. Immediately thereafter, the A330
commander heard a TCAS "climb climb" warning and he noted that the A340 TCAS symbol had changed colour to red on his navigation display. He looked out and down at the A340 which was some 200 to 300 feet to his left in a nose-up attitude and climbing steeply. The A340 passed through the A330's level before the commander had time to react to the TCAS warning and the
TCAS was still issuing a "climb" instruction for a short while after the A340 had climbed above the A330. The commander continued to monitor the A340 visually and on TCAS. It appeared to reach an apogee above FL380 although by this time it had fallen behind the A330. Nevertheless, it was still laterally quite close to the A330's track so the commander altered course to the right to make space for the A340 to descend back to FL 360.

UKAAIB Report here -

http://www.aaib.dft.gov.uk/cms_resou...pdf_501275.pdf

[3rd_ear]

That's really quite an interesting report, not least the CAA recommendations. Of course, if both aircraft had previously offset by the same distance, the situation would have been unimproved. Would it not make more sense to take action much like on the road, and for the overtaking aircraft to change offset in relation to the one ahead prior to overtaking? IANAP, but surely this could also be done safely without visibility?

[fox niner]

If we all offset with the same amount, nothing will change.

In my B777 FMS I can only enter 1NM, 2NM, 3NM etc offset distances.

I would like to be able to enter 0.1, 0.2, 0.3 NM etc but I can't.

[Wiley]

What about opposite direction traffic, fox_niner - which was the main point 410 was covering in his original post?

[fox niner]

Ah, wiley, sorry about that:

Let me rephrase....

In Brazil and Africa I always use 1 NM right of track. Perfect.

However, last weekend I was on track X-ray (NAT) like everyone else and literally EVERY flightlevel was occupied, within a distance of 10 miles. If I were able to enter say 0.5 NM right/left, and the others used some other convenient 0.1 step-distance, everything would work out better on a busy NAT. But unfortunately we have to wait till a software update Blockpoint such-and-such for an improvement.

[Green Guard]

I think you'll find that 410 is indeed saying that everyone should offset 1/2 NM so that (not in the usually one way NAT airspace, but in the rest of the world), no one will be on the centreline on a two way airway.
And what the 'ell is wrong with that ?
Are you also worried that nobody drives OVER the center line, between two opposite lanes of your road?

[MTOW]

That was a bit too cryptic for me, Green Guard. Are you saying you don't think it's a good idea?

[Engineer]

You have read the script so why not look at the film

The idea is to introduce randomisation that was built in with poorer quality INS system of bygone days. With today's accurate GPS systems you could find your self flying straight down the middle of the NAT track every time.

Since the North Atlantic is a busy place it seems sensible to introduce a little randomisation by electing to fly 1nm offset one day, down the middle the next and then may be 2nm offset the day after. If every one approaches it this way then may be it will reduce the possibility of an incident