Josh wrote:
Rwy20 – the runway threshold position was incorrectly coded in the database due to a confusion over planned work, leading to previous crews breaking out 4 Whites on the PAPIs. An incorrectly coded threshold position at 200ft minima would leave you dangerously positioned when becoming visual and with no option to go around and divert.
OK, I misread your statement to mean that the accident had been caused by this incorrect data, whereas you probably just meant that it brought it to light, as highlighted in the accident report:
The aircraft was equipped with a FMS database showing runway threshold 02 at 26 meters to the left of its actual position and 120 meters past the actual position. Nepal had issued an AIRAC AIP supplement on January 1st 2015 displacing the threshold by 120 meters due to planned runway extension work, the supplement valid until March 4th 2015 was cancelled by NOTAM past the AIRAC cut off date, the displacement thus remained active in the FMS database. The displaced threshold remained in the database and charts of the database provider even for the next AIRAC cycle.
The NAIC wrote: “Finally on 01 April 2015, an AIP Supplement was issued cancelling the AIP supplement S002/15 which was eventually never implemented. This means that the threshold of RWY02 was never officially and physically displaced but the NAV Database were modified.”
With respect to the 26 meters offset to the left the NAIC reported that the airline’s version used a resolution of degrees minutes and seconds to compute the runway threshold, while the AIP supplement provided the coordinates at 0.001 seconds resolution resulting in the position shift.
I still don’t understand how this would be a particular case of “all eggs in one basket”. On an ILS approach, you also put all your hopes into the correct functioning of the equipment until you break out and become visual with the runway. If you start to disregard crucial parts of procedures such as minima, you are always in uncharted territory.
Peter wrote:
So to fly an LPV you must have the current database cycle. If you have the previous cycle, but nothing has been changed, you are not legal to fly the approach. No such limitation on an ILS.Normally, for IFR GPS usage, you will have a current cycle but sometimes, especially on a long trip, it may not be possible to achieve that. The wording of your IFR GPS AFMS may authorise the use of an expired cycle if you have checked that nothing has been changed (mine is thus worded, for example). But in this case, IF this is an actual regulation, you won’t be legal to fly an LPV approach in such a case.
Actually, it is very easy to update the database. I do it regularly on the airplane (GNS430W) with my laptop and a mobile internet connection. Takes 10 minutes if that much. I would not see why it’s difficult to do on a longer trip, you have internet now everywhere.
I think the other bit they are on about is that you have to make sure, every time, that the procedure you want to fly or might want to fly is actually in the DB before you leave. There have been incidents when crews found that a procedure was not there when they tried to select it. Databases are all 3rd party products…
Re the ILS: If you fly the ILS with a normal radar line up, theoretically you don’t need the new DB. If you fly the ILS approach out of the DB, then obviously you need it. Personally, I felt that IFR with an IFR GPS ALWAYS means to have a current DB.
Josh wrote:
Their view I suspect is that they pay for ILSs through landing fees and Nav charges, and I would imagine Ryanair and Easyjet are making their views known to those airports that are thinking about binning their ILSs.
As I said before, this announcement could well be a shot across the bow waiting for the wailing to start and then offer them to keep their ILS’s if they pay for them.
Easyjet, on their A32x they most probably have LPV by now, at least RNAV/VNAV. The Airbus upgrade appears to be not that much of an issue, I recall from a company close to me here that it was done for all of their airplanes without even thinking about it. Especcially RYR and EZY will NEED to upgrade their systems, as their provincial airports they fly to are most likely to go this way. That is a totally different story than others such as ACARS, even though I must say Ryan are very annoying that they still have to block all the FIS frequencies getting weather. I once heard that Eurocontrol was about to come down on them on that regard.
Josh wrote:
All in all, they are well worth the hassle in the end, but at the moment we are at the “growing pains” area, where expensive investment is needed to meet the new standard, the goalposts on training requirements are moving all the time and national regulators in general are throwing spanners in the work with conservative and inconsistent approaches to certification.
True. And there are many who will refuse new stuff out of principle and “we’ve always done it that way”. Wonder if any of those clowns would be happy with a radio range approach these days :) Me for starters I am happy already that NDB approaches are no longer part of the exam curriculum in this country, for the lack of approaches to use.
Peter wrote:
There is no CAT 3 LPV. ILS is here to stay “for ever” for airlines.
No. Actually, I would not be surprised if CAT II/III would become reachable even for GA sooner or later BECAUSE of this technology.
That LPV CAT III will be available within a few years is not a question, it is a fact. It will go one by one, CAT II then III. Probably they’ll call it something else just to confuse everyone, but with the current accuracy of the ground aided augmentation, you could in theory even do auto-taxi these days.
I agree that the root cause of the accident was quite frankly what I would describe as criminal negligence on the part of the operating crew.
On the ILS one has outer markers or glideslope crosschecks plus idents and warning flags which all provide a check on the integrity of the system. A latent fault which places one aligned with the runway edge and at what would be 4 whites or full scale fly down at minima would end up in a go around every time under stable approach criteria for anything larger than an SEP and in limiting weather would be pretty rubbish.
On an LPV one has to take the integrity of the database on trust from LIDO, Jepp or Navtech and can verify only so much. One’s distance is taken from the same source as all other information. The only sourced crosscheck at the moment is RAIM, requiring satellite geometry to provide 6 visible and usable satellites, which is why locally based ground stations and GLS will be required to help reach the accuracy and reliability required from anything CAT I and below.
Using incorrect or differing units is a big source of latent error – everyone should be working to the same numbers. Let us not forget that NASA lost a Mars orbiter due to a contractor not using the units specified in the contract (in this case pounds not Newtons of thrust) and this sort of error will become more common and more difficult to detect as we transition to RNAV based solely on supplied databases compiled from third party information.
There are significant differences between ILS and LPV in terms of what can fail and how it would be detected.
If the LPV data block is coded wrong in the database, the airport will never know about it, and arriving aircraft will track the wrong lateral and vertical trajectories, with no indication of anything being wrong. There is no airport-based monitoring equipment.
ILS is monitored by ground equipment and if this finds the radiation off-spec it (a) warns ATC and (b) switches to a backup transmitter(s) – AIUI.
Wrong LPV codings are very rare but clearly not impossible, and can get people especially if somebody is busting the minima 
And you need several holes in the cheese to line up for this to be discovered e.g.
whereas an ILS problem is likely to be picked up pretty fast because every old dog can fly an ILS, usually on autopilot, and will do so in anything resembling bad wx.
you have internet now everywhere.
Not so, which is one of probably several reasons why Jepp still sends out the updates on a DVD (optionally). I have been to many places where I had no useful internet connection, especially for the ~500MB to ~4GB download involved in updating Jeppview or a GPS database. And bizjet pilots have bigger issues with this, where they might travel.
Plus, sometimes, flashing a database card fails and buggers the card, which is why I have two of them so I can always revert to the previous cycle and fly with that.
If you fly the ILS with a normal radar line up, theoretically you don’t need the new DB
There is no “database” for an ILS. It is not a GPS procedure.
Peter wrote:
There is no airport-based monitoring equipment.
I have to ask my colleagues from Skyguide but I am quite sure there are. ILS is monitored with an antenna and other monitors, GPS performance can be monitored very similarly. We had a few of these things installed at ZRH very recently and were told they are monitor receivers for RNP verification. In fact, it should be even easier to monitor than an ILS installation.
Peter wrote:
how many planes will be flying the LPV (very few, for many years, in Europe)
I beg to differ. And I think that for once GA might be the ones going ahead here, because it is relatively easy to get equipment and installation. Almost everyone I know who has recently updated his airplane with Waas GPS will get the LPV paperwork, it would be stupid not to. Of course it will depend on the availability of the approaches but looking at that, they are popping up everywhere these days, certainly in Germany and very soon also in France.
Peter wrote:
how many will notice the error and bother to report it, perhaps after a flight to a foreign country
If there are errors, they must be reported ASAP. and they are. I recall vividly what happened when there was a simple SID error by Jeppesen in one of ZRH’s SID’s, after day one of the new cycle, it was very well known and published!
Peter wrote:
whereas an ILS problem is likely to be picked up pretty fast because every old dog can fly an ILS, usually on autopilot, and will do so in anything resembling bad wx.
In practice, flying an LPV is no different than flying an ILS. And those who have the capability WILL use it. My plane did several IFR trips this year and wherever the approaches were available, they were flown.
Peter wrote:
Not so, which is one of probably several reasons why Jepp still sends out the updates on a DVD (optionally).
As long as I have mobile internet (via my mobile phone) I can update the DB. So far I have not come across any place in Europe where I can’t do either W-LAN or 3/4G internet. Also, the updates don’t become available only on the day, they are available some days before. You can prepare the new cartridge before you leave. This ain’t the dark ages anymore…
Peter wrote:
There is no “database” for an ILS. It is not a GPS procedure.
Most ILS procedures including the line up are in the DB of IFR GPS’s. So the line up occurrs via GPS and is switched to ILS upon the last intercept course. If you fly one of these, obviously the DB must be current for that. If you are radar vectored or use conventional radio nav to get onto the ILS, then you don’t need it.
GBAS now has CAT II capability and is in service at a few US airports. They are working on CAT III and as I understand it, they are still having some technical issues on roll out. But this will be resolved in a matter of time. The airlines will equip with GBAS, but want to avoid SBAS and wait until the next generation of GPS satellites provide L5 and other enhancements.
There is checking of the LPV and LP final approach course. These procedures use a FAS data Block (Final Approach Segment). The database information for the FAS is provided directly by the approach designer, in the US, by the FAA. This data is protected with a 32 bit CRC code which is part of the record and is validated by the WAAS GPS on retrieval from the database. Before any pilot flies an approach with this data, it is comprehensively flight checked by the FAA. This provides end to end integrity for the database data for the final approach segment.
Peter wrote:
Wrong LPV codings are very rare but clearly not impossible
There are nasty ILS failure modes as well, if construction machinery cuts the cable to the difference antenna of the glide slope, you will get a perfect on glide indication with correct morse ident no matter where you are. Ask Air New Zealand how they know.
tomjnx wrote:
There are nasty ILS failure modes as well, if construction machinery cuts the cable to the difference antenna of the glide slope, you will get a perfect on glide indication with correct morse ident no matter where you are.
Isn’t the monitor receiver supposed to catch that?
Airborne_Again wrote:
Isn’t the monitor receiver supposed to catch that?
Yes it should… if it is operational.
There are other issues with the glideslope the monitoring receiver is unlikely to catch, such as (metallic) objects on the reflection surface, heck even snow alters the glide angle.
Which is why things like thrust settings, rate of descent and distance/altitude cross checks in IMC are important. They also used to have marker beacons for the similar reasons.
