Caution - Gross Navigation Error Cause

[JammedStab]

The ++++ Aviation Gulfstream G-4 (N++++) aircraft was en route from Nassau (MYNN) to Farnborough England (EGLF). While cleared to 45N 050W deviated north of track by 25 miles. The pilot confirmed the INS showed on course but Gander ACC corrected their course to 50W. Three other aircraft, a Swedish military G4 and two other US registered aircraft a G200 and GLF4 were all involved in similar events at the same location 050W. The two G4 crews reported that the FMS flight their plan entries 5050"N" were incorrectly entered as "N"5050. Apparently, this misplacement of the "N" caused the Gross Navigational Error.

[Intruder]

You don't say if these were downlinked routes or manually entered into the FMS. However, I find it difficult to believe they were downlinked, because the course & distance would have flagged the error. Besides, AFAIK, N5050 is not a valid format in the first place; the only valid formats I know of would be 50N50 (N50W150) or 5050N (N50W050).

It would have been better to enter 4550N (N45W050). It would have been even better to enter the longhand N45W050. Better yet would be to check the waypoint and route in the FMS before entering oceanic airspace.

I think there's a bit of a problem with that report...

[safelife]

SOP with my outfit is, after reception of the oceanic clearance, to compare it to the FMS flight plan including clicking on each waypoint to show the complete coordinates (45°00.0N/050°00.0W) are correctly coded.
Plus a check of average true track and milage between points of course.
Duly done every time...

[Intruder]

And here, from another forum, is the missing link. Indeed, Jepp added stuff to the databases recently, and likely caught people unaware:

Jeppesen NavData for cycle 1404, effective 3 April 2014, included a large number of new waypoints at half degree latitude/full degree longitude. These new waypoints are causing operational difficulties for customers who were not aware of the additions. The approximate geographic extent for the additions is from 22 degrees north latitude to 67 degrees north latitude and 050 degrees west longitude to 020 degrees west longitude.

The waypoints were added following the industry standard for database preparation, ARINC 424. That convention uses the positioning of an alpha character to denote full degree or half degree of latitude:

5040N – trailing character – full degree of latitude, i.e. 50 00 00N 040 00 00W
N5040 – leading character – half degree of latitude, i.e. 50 30 00N 040 00 00W

THEREFORE, FLIGHT CREWS ATTEMPTING TO ENTER WAYPOINTS USING LATITUDE AND LONGITUDE NEED TO BE AWARE OF THESE CONVENTIONS AND THE FACT THAT BOTH FULL DEGREE AND HALF DEGREE LATITUDE WAYPOINTS ARE INCLUDED IN THEIR NAVDATA FOR CYCLE 1404/1405.

[framer]

Wow, that's a real gotcha right there......still, checking tracks and distances as you go will keep you safe.

[DaveReidUK]

That convention uses the positioning of an alpha character to denote full degree or half degree of latitudehttps://en.wikipedia.org/wiki/Murphy's_law (https://en.wikipedia.org/wiki/Murphy%27s_law)

[PJ2]

Jeppesen will be removing the wpts on the May 29th cycle, link to Jeppesen announcement (http://ww1.jeppesen.com/documents/aviation/notices-alerts/navdata/Half_Degree_Lat_Long_Waypoint_1404.pdf)

The primary reason for adding intermediate waypoints is to put more airplanes in a smaller space. I wonder which group wanted this change and although thorough research would have been done but nevertheless how does a potential navigational mistake this simple make it through?

I think any airline pilot experienced in overseas ops could have told them what the immediate problem and possibilities would be.

[Intruder]

I saw nothing in the announcement indicating they will remove the waypoints. Where did you see that?

[PJ2]

On the Jeppesen site, at Aviation Notices & Alerts (http://ww1.jeppesen.com/company/alerts/aviation-alerts.jsp)

On the Jepp site under Notices and Alerts, select region, (Atlantic), sub-region (All) and publish date (Last 30 days). The notice is the same as the PDF except that at the bottom there is the statement:

"The waypoints will be removed from Jeppesen NavData for cycle 1406, effective 29 May 2014."

[Bearcat]

I guessed someone made a giant cockup.....last few days shanwick both directions have been advising crews to check shannon volmet re sigmets......in there it says check your navigation re waypoints. Causing a lot of extra chit chat on 123.45 and disturbing the crossword......

Not the first time N5040 and 5040N inputs have been screwed up and yes its a huge threat. Check your tracks and distances and your lats/longs !!

[Una Due Tfc]

More than one cockup I hear

[PJ2]

To start with I don't think all systems aloft are "self-loading", so the this is pretty fundamental human factors / organizational stuff.

A single, mis-typed/mis-placed character is a high probability, single point of failure event which is capable of producing a max nav error of 30nm on the Atlantic.

That is a serious failure somewhere in the organization. I'm surprised Jepp is leaving it until the next regular cycle to withdraw.

[bucket_and_spade]

Surely every operator's SOPs should pick this up!?

[Capn Bloggs]

Jammed Stab, could you please post a link to the original report.

[MD11forever]

Interesting, I already used this naming convention for entering waypoints quite some time ago.
In the Navdatabase I use these seem to be included for years.
All properly described in the manual.
However proper checking of the coordinates is very important, to check if you got what you wanted.

[PJ2]

The fact that the naming convention has existed "for years" is beside the point - clearly it hasn't been in widespread use by many others, and GNE's resulted immediately upon recent introduction.

Yes, SOPs should have caught all errors but they didn't.

We know that it does no good to simply implore that SOPs be "followed" because people make mistakes. Otherwise we wouldn't have AF447 and a ton of other accidents which the adherence to SOPs may have prevented.

One of the easiest mistakes to make when pressing the FMC keys is a transposition of characters. Usually that is caught right away either because the FMC doesn't accept the sequence or the other pilot is watching. If not caught there, checking the flight plan tracks and distances at each waypoint is the SOP that should catch it, followed by the standard checks at each waypoint passage while enroute.

As your experience illustrates, it isn't the naming convention that is the problem. It is familiarity, which perhaps means that this particular change was too quick, and/or not properly described in manuals?

[deefer dog]

It is astonishing that a change of such magnitude would not be advertised, promulgated or posted far and wide before implementation.

Yes of course we should check entries for validity etc etc, but by failing to tell operators of such a drastic change is tantamount to negligence, and surely served to line up 90% of the holes in advance of three incidents. (Three instances AT LEAST so far, probably more.)

PS. They should make the volmet announcement more clear also. We had to listen to it four times before understanding it.

[Una Due Tfc]

11 GNEs in the first week.

[oceancrosser]

It is astonishing that a change of such magnitude would not be advertised, promulgated or posted far and wide before implementation.

Yes of course we should check entries for validity etc etc, but by failing to tell operators of such a drastic change is tantamount to negligence, and surely served to line up 90% of the holes in advance of three incidents. (Three instances AT LEAST so far, probably more.)

PS. They should make the volmet announcement more clear also. We had to listen to it four times before understanding it.

As PJ pointed out, this naming convention has existed for years. That does not make it any better though. Last year I had a new F/O that punched in
N6310 (which is N6330W010) instead of 6310N. It was caught though and an embarrasment (or worse) avoided.

Volmet? Not used those in years, but that would be very subject to HF conditions. How do you suggest to improve those? ACARS is the biggest improvement.

[deefer dog]

Oceancrosser:

I agree the convention has been in place for some time, but entering N6030 has for all that time flown me to 60N30W.

If it was suddenly decided to action the convention, would it not have been sensible to promulgate it? I think we are all agreed on that, and 11 GNE's in the first week surely endorses the sentiment.

How many reminders do we get about largely trivial matters such as 270 kts approaching the London TMA, endless repetitions of how to avoid level busts and runway incursions? A few people in a few authorities did not have their eyes on the ball, and frankly they were negligent in my opinion for not to highlighting the introduction date of, albeit, an old convention.

PS. Volmet was on VHF, and we were told to listen to it AFTER we had crossed the pond!

[Astra driver]

I happen to know the Captain of one of the aircraft involved, who is a highly experienced ex-Air Force pilot. He told me how he inadvertently inserted the waypoints incorrectly and then failed to miss the error when checking the Lat/Longs on the way point list, basically he saw what he expected to see, something like, N53 00.0 W050 00.0 . Instead he got N53 30.0 W050 00.0
I could see if you're not expecting to see 1/2 degree waypoint it would be possible to overlook it.

The crew were cleared direct to the first waypoint coasting out from Canada, bypassing the "Gate" waypoint which meant that a GNE chart plot would not have clearly shown the error (or was required, since they were still in radar contact) . Luckily the Gander controller caught the error and it was corrected before reaching the first fix, he informed them they were the 7th crew in a week to make that error. No violation was filed, but they got to have "Tea and biscuits" with the CP and it resulted in a company wide memo being distributed.

[Ian W]

Interesting, I already used this naming convention for entering waypoints quite some time ago.
In the Navdatabase I use these seem to be included for years.
All properly described in the manual.
However proper checking of the coordinates is very important, to check if you got what you wanted.

There have been many conventions used for years that were found to be unsafe and altered. This particular one will become extremely difficult when Europe becomes 'business trajectory only' and multiple waypoints have to be entered. There is a demonstrated risk that someone has and will get it wrong in non-radar airspace. Who will take responsibility now the risk is known, if there is a mid air or even just an RA solved NMAC in transatlantic airspace?

[Capt Groper]

If a new NAT track or routing is issued I recommend to enter the full LAT and LONG and let the FMS determine the alphanumeric code for the waypoint. For example enter N5030.0/E03000.0 as a new waypoint and the FMS should abbreviate appropriately. This avoids incorrect data input i.e. 5030N instead of N5030.

[Astra driver]

Not sure if entering the full Lat/Longs is the solution as this now gives you more digits to potentially screw up, such as Capt Groper posting his waypoint of N5030.0/E03000.0 which is E5030 (Apologies if you intended to post a East waypoint)

The lesson I take away from this is to slow down, be careful and double check.

Jeppesen and the associated avionics manufactures should have also gone to greater lengths to better disseminate this information ahead of its implementation.

[misd-agin]

Capt Groper and safelife are bang on. Which is why, and how, SOP's get created.


Type in waypoints the long way. Line select auto loaded waypoints and confirm the raw lat/long data. Compare distance and heading(usually 2 degrees off on the N. Atlantic) between waypoints.

[underfire]

Sounds reasonable. Simply enter the raw LAT/LONG.

As shown by this thread, multitudes of incidents have occurred. You enter numbers a certain way, and there is a pretty good chance you will mess it up, and the aircraft will drive in the wrong direction. :eek:

So, rather than just simply enter the LAT/LONG, there needs to be training and SOP how to do the shortcut the right way?

In reality, why is this even an issue at this point?