great instrument cross checks!!!!!
 

http://youtube.com/watch?v=0qaDVaKQSkQ

Watch this for a full scale w
"we trust the aircraft"
No x checks untill below 1000 amsl @ 8 DME

They should have all gone back to ITS.

As a mere SLF with no technical knowledge, I'd love an explanation of why.

When I read the report, I'd got the impression that the crew had done a good job of being alive to clues that all was not well, resulting in the GA before things got critical. And I don't remember any criticism of their actions.

What did I miss?

Thank God they did notice even if it was later than they or others would idealy have liked. They then took positive action and put the aircraft in a safe place so good for them. I rather suspect that the lessons they took away will be more firmly imbedded than anything a flight school could impart.

For me its a reminder that any of us on a given day can make errors. I have seen mates I regarded as better operators than I kill themselves by making basic errors and if they could manage it then I am certainly not immune. The thing is we all want to make the evidence fit our model of what we think is happening and the visual cue of an ILS is a strong one. Would I have done any better on the day, don't know. I can only hope that the penny would have dropped and that I would have taken the same positive action.

It has made me think though, which I guess is the aim of the video so thanks for sharing it.

There but for the grace of God (can I still use that word?) go I. Easy to criticize but the whole purpose of this "internal" educational was to bring this deceptive chain of events to those drivers who feel they're above duffing it.

Here's the splurb:

4HP

.

great Instrument cross checks!!!!!
 

Hi Globaliser.
A standard ILS slope is 3 degrees, Faleola is a standard airport in Pacific Paradise.
one degree is 108 ft per Nautical mile.
so 3 degrees is 324 ft per mile.
At 8 miles they were 970 ft instead of 2464 ft above MSL
The pilot flying and even the Pilot not flying should be cross checking DME with altitude.
The Pilot flying and the pilots not flying should take a note of the time of capture and the distance of the capture.
Lets Say @ 10 miles so its going to take about 3 minutes at 180Kts to get to the runway.
A mental picture must be present.
Never rely on only one indication of what you are suppose to be doing if another method is available to cross check
These guys had two other sytems, Time over distance and altitude over distance but they all locked on to the IAS ( who looked at the VSI??)
It goes on and on.
They all needed a Base check(180Day check)
Thats Basic instrument flying:=Eh

One reason why crosschecks are so important with an ILS is that most people think that two narrowly focussed beams intersect to provide the guidance, but they don't!

The "beams" are created electronically by the equipment in the aircraft (that is, voltages are produced from the radio signals, which drive the instruments). This means you can get on-course or on-glidepath indications regardless of your position, as was found by an Air New Zealand 767 in July 2000, which got down to 400', 6 miles short of the runway - maybe that very aircraft?

Phil

Discussion regarding this incident by professional aircrews has been around for some time as the lessons in prevention are clear and valuable. Globaliser; - You are correct. This crew did not "screw up", but prevented a CFIT. Bbg is correct in stating the various parameters which need to make sense for all approaches but the tut-tutting is inappropriate.

From the CAA Report:

2. Analysis

2.1 Introduction

This analysis considers the principal areas where there were opportunities for the incident
to be prevented:


 Operational Factors


 Aircraft Instrumentation and autoflight system design


 ILS Ground facility design and maintenance


 Organisational Factors

Only factual information that was considered to have a bearing on the incident has been
included in the analysis.

2.2 Operational Factors

2.2.1 Preflight / Takeoff / Cruise

This was an appropriately cautious crew, displaying a high level of awareness of the
differences between operating to a major airport and operating to the Pacific Islands.
The preflight preparation was thorough and not rushed. The crew was communicating well
and functioning as a team before leaving Flight Dispatch. The crew environment ensured
all crew members were able to express any questions or concerns. The workload allocated
for each crew member was appropriate. With respect to the various NOTAMs on Apia, all
available resources were used to assess and evaluate operational matters pertinent to the
flight.

The preflight and takeoff proceeded normally. During the cruise a minor fault with an
automatic announcement tape was attended to with the assistance of Maintenance Watch
and the onboard Air New Zealand Service Engineer. The crew were unaware of the
presence of the Service Engineer onboard until informed by Maintenance Watch.
Of note is the individual attention to a self-brief by the crew for the approach into Apia and
the thorough and comprehensive descent and approach briefing given by the PF.

So why don't we emphasise a cross check of the Rad Alt? No mental arithmatic required. Rad alt displayed? Then we are 2500 ft AGL or less...is that where we should be?

Well I think that is emphasized in both training and SOPs as well as just being solid airmanship for instrument pilots. Bbg mentioned the 1-in-3 rule and that ratio is a fundamental especially after the radio altimeter is alive. The assumption is obviously that one has a DME or other distance-determining equipment and is using it in concert with the rad-alt.

It works well any time...about 300ft over the 1000ft bar in an instrument lighting system is one of many examples.

It's been some time since I read the AN60 report but the comments regarding the crew's professionalism regarding what they knew was going to be a challenging night approach was what stood out.

If one is interested in comparisons and outcomes of similar initial circumstances, the KAL801 B747 into Guam is an example.

bbg,

Assuming your profile is correct, you are obviously an old has-been who now takes delight in sitting on the fence sniping at people who do what you can no longer do (though I wonder if you ever did!!).

These guys had a valid ILS indications, a valid Indent and no flags. These indications are sufficient for us to allow a CATIIIB approach in 50m Vis with no decision height!!
The situation got very rushed very quickly. A perfectly routine approach suddenly started to feel wrong, with the aircraft descending faster than normal.

Perhaps in your day (when accident rates were far higher) you might have decided this was a time to sit back and start trouble shooting. These days, believe it or not, we are taught to FLY THE AEROPLANE, which is exactly what the Captain did.

THEN faced with a confusing and contradictory set of data, the crew rightly concluded thing were not as they should be, went around, and prevented a CFIT. A professional crew doing what there trained to do.

How great is it that ANZ chose to put together a Video to highlight this particular problem, and to raise awareness of the limitations of ILS in general?

Now that any such video ends up on U-Tube, and is subjected to the half-arsed analysis of "Monday Morning Quarterbacks" such as yourself, not to mention scum-bag journalists, how likely is it that such efforts will be made in the future?:ugh:

Edited to add- We haven't done Base Checks for years (have you heard of a Simulator?) Get up to date or but out!!

While I appreciate the 'support' of the crew here, I have to be just a little picky.
They let the A/P fly them down to 400', before they executed the missed approach!
Now, when was the Alt check done?
At the IAF?
At the FAF?
No, it wasn't done in accordance with the very basic requirements of any instrument approach.
This is not an optional check guys.
Dont let the aircraft take you for a little ride

Speaking as an engineer (electronic, not flight), it appears that there's a fundamental design fault in how ILS works if it can lose the modulation and still convince an approaching aircraft that it's happily on the correct slope. You almost want to deliberately wobble up and down slightly on approach if it claims a perfect and instant capture just to prove it will tell you to get back on the appropriate slope. I've evolved a very "what if?" approach to things over the years to try and pick up obscure failure modes, and aviation is a rich source of things the designers didn't think of, partly because it's so public and well-documented after the event.

For those that didn't spot it, one of the contributors to the YouTube thread posted a longer version of the video that includes a simplified version of how ILS works and what was wrong in this case.

I agree entirely with PJ2. The crew did a good job and it is a shame that this thread is given a title suggesting a lack of airmanship on the part of the crew.
Moreover, Air New Zealand did the whole industry a favour by producing an excellent video of the event and allowing other airlines to use it free of charge.
In our airline we used the video as a basis for an ILS lecture during winter training. We have all learnt from the ANZ incident. Air New Zealand and the crew should be commended for their honesty and initiative in producing a video that could well prevent an accident in the future.

Perhaps ANZ could prevail upon Tube to remove the version which does not include the technical explanation.
Years ago, when flight checking ILS installations, one of the many checks we carried out was one for flyable false glideslopes; although, as indicated by the second video clip, this was not the problem on that occasion.

Wizofoz,
Hey, watch it, mate, there's a lot of us in here - just keeping an eye on you young guys'n gals ;)

You're very welcome,too, Bas, as long as you talk sense and not drivel like bbg!!

Outer marker/height check, not SOP?

No OM, it's over the sea, just an IAP - both the skipper & the jumpseater picked up the DME discrepancy inside the IAP.

Easy to sit in front of your PC & theorize - try doing it on a dark night at 180 knots when all other indications are what you're used to seeing & there's no flags.

These guys did a great job & we can all learn a lesson here.

4HP

llondell,

here's the link to the longer post.

Well worth a look.

http://youtube.com/watch?v=GelRBhJ4gmI

O

osbo,
Many thanks for the link!

All we need now is somebody coming up with a link to where we can buy the DVD, so we can refer to it even when this thread has dropped below the horizon.

It's also interesting that for recovery from unusual/upset attitudes/positions, training emphasizes the need to trust the primary instruments and discount your own physiological sensations. In this case, although the other peripheral clues were present, the needles were locked solidly on to the ILS with no indication of failure, so other than a growing sense of discomfort, it was a natural instinct to believe the ILS all the way up to the point where the visual cues outside confirmed the underlying doubts about the approach. The VSI and power settings were disparate enough from the ILS indications to get the hairs on the back of neck up, but the error was not significant enough for the crew to immediately understand what was happening.

This was another good example that it is careful analysis of all the cockpit indications that presents the most likely picture of actual events, not just the one you would most like to believe.

I can see from a legal standpoint ANZ might not want this video to be available to all and sundry, but as a CRM teaching aid it is priceless. For all those here castigating the crew for their slipshod performance, I can only suggest you read more accident reports to identify exactly what kind of pilots find themselves in these situations - given time and the opportunity, all of us.

...or you could always just download the video from Youtube.

With your chosen video open and playing on Youtube, copy the URL from your address bar, open another browser window, go to www.keepvid.com and paste the URL in the green box. Click on "download", and when the link appears below the green box, right-click it and "save as..." or "save target as..."

For a free flv player (the format the video is in), go to:

http://www.rivavx.com/index.php?id=422&L=3

Have fun!

Probably the most educational thing I've ever seen on the web wrt flying.
So easy to get suckered in. Hats off to ANZ for making this.
20driver

Wizofoz,
Distance versus height is NOT rocket science, and should have been an essential part of the instrument x-check when the approach indications were being queried.
If the aircraft HAD impacted the island they so nearly did hit I'm quite certain that the resulting accident investigation would have severely criticized this lack of cross check.
The DME arc approach on to RWY 05 at Faleolo is NOT a simple one and requires full reference to ALL the electronic facilities available, i.e. Dme, ILS and VOR.
IF this had been done then it should have been readily obvious that the V Nav/ILS data was in error.
Check this 'Google Earth' location : 13° 49'07.14" S, 172° 09'10.17" W
Faleolo, Western Samoa.
HectorusRex

Wizofoz

Have to add...

No they aren't. You need quite a few other things as well, such as monitored aids and a pair of radalts to name but two. Also, because a CatIIIb is a monitored approach, it should have been immediately obvious if there was a height or vertical speed discrepancy. Of course this airfield could never be certified for CatIII so all that is irrelevant!

"Flying the airplane" includes having a very clear awareness of your position relative to the runway, and one of the most basic elements of that process is a constant "gross error check" to make sure that you are roughly in the right spot. These guys lost their SA, and that is not "flying the airplane" in any way, shape or form.

If they had really "done what they were trained to do", they would have recognised the problem a lot earlier.

Finally, "base Checks" were done in sims before they became LPC/OPCs (in Europe, where you apparently are), but then you are probably not old enough to remember that... so I suggest you pull your head in and listen to people with a little more experience.

What I don't understand is why the A/P kept descending at that high rate? The false G/S signal was centered - why didn't the A/P reduce the sink rate to the normal 750-800 fpm? Or was their groundspeed so high that the A/P calculated it would have to descend at that rate to keep the needle centered?

But they did not impact the island. Why, because they knew something was'nt right, kept asking the question and came up with the right answer and took the right action.

Some feel they should have done better, maybe some of us are right but can any of us say we would have done better on the night.

I can't say that myself but I can say that I have learnt from the video and it has made me think of how I conduct ILS approach's. That does not mean I won't get it wrong but perhaps it will stack the odds a little more in my favour.

The video was not published so we could take pot shots at the crew, it was published so we could learn. It was published on the assumption the target audience would be humble enough to recognise that it could be them.

In the bad old days when we were quick to blame,people covered up their errors so the lessons were lost. Result, higher accident rates. Thankfully things are now more open and acident rates have fallen.

Distance versus altitude ( not height ) is not all that straightforward. How quickly can you work out a 13nm check with a runway elevation of 763ft. If your outside the DME ranges on the chart its arithmatic time. By the time you've done it your 1-2 nm further on so thats no good so you have to anticipate the check. Add in taking flap, gear, speedbrake, checks, adjusting the MCP, cabin secure calls, frequency changes, complying with ATC speed and there is a-lot potentialy going on to get in the way.

Sure we should manage the approach so we don't allow ourselves to be rushed but in this case it all happened late on at G/S intercept and no amount of planning could have prevented that. So for me its a reminder to cross check and make use of as much info as I can but if I lose a bit of SA and start to get tunnel vision and things are not adding up then its time to follow this crews example and take positive action to put the aircraft in a safe place.

Instrument cross check
 

I do not withdraw my explanation of the cross check requirements but having seen the second unabridged video,
I soften my critical remarks about the PF. He did do a DME alt x check( edited out in 1st video)
However, the aim of the exercise is to learn something that just might save your arse one nite when tired, pissed off, or just in a plain old bad humour.
I do not and have not set my self up as a critical whiz boy, I was a poor instrument pilot who suffered from the leans even after 30 years Instrument flying in Heavy and little jets.
I had a similar experience back in 1965 at NFFN, where a cross check that was ratified at 2400ft established that the machine was on a false glideslope.
The initial response was to revert to a twin NDB( for you new smart arsed pilots that was the only system for several decades and is still a valid sytem where it exists)
I am still in the loop, I am a curent and practising Flight Instructor, with a rating that has been valid and active since 1961.
I will add I am still learning, are you?
BBG

bbg,I think I saw it mentioned that the full DVD is 38 minutes, so the second video is also edited down to the YouTube 10 minutes limit, even if it has more info about the ILS failure.

Knowing where to get the DVD would still be useful.....

The incident report explores this in detail.

The carrier signal had a very slightly biased DDM, displaying the aircraft ever so slightly above GS, and the aircraft apparently performs zero sanity checks on vertical speed, slope, etc. So the slight bias caused the aircraft to continually push the nose down until the crew went around.

As llondel notes, there appears to be a design fault in a system that can give a "correct" indication when the system integrity has failed. It also appears silly to me that the avionics don't make better use of all the data available to check the sanity of an approach. Isn't all the data to complete a DME check already in the FMS? The humans take a while to complete those sanity checks...but thankfully they did.

Ashling

Try this in a lot of other locations (PGUM foer one) and you'll be pushing up daisies a lot sooner than you would like.

I ask three simple questions. When you answer them to yourself, consider if they were fulfilled by this crew.

1) what is the minimum glideslope intercept altitude on the approach in question. (hint: it's on the chart)

2) At what distance should this occur. (hint: it's on the chart)

3) How is this distance determined. (hint: it's on the chart)

Anyone who descends below the minimum glideslope intercept altitude without verifying by distance X check, has no place in this industry. That was fundemental when I did my CIR 37 years ago and is equally valid today.

Prepared to cop the comments. But it needs to be said. Unhealthy amounts of dick licking and ego stroking here.

Maui

ChristiaanJ, you might find this link useful then :P


http://www.airnz.co.nz/aboutus/fligh..._resources.htm

13x3(00) = 3900
3900 - 750 = 3150

altitude/pressure correction at your discretion.

Yeah really hard math that... rounded figures being fine for a gross error check... :rolleyes::rolleyes::rolleyes:

One mistake we usually do is looking out of the window ,if feeling something is wrong.It's an IFR flight ,just to DH ,so looking out will only waste valuable time.
Beside the vey important distance/altitude check and the OM altitude check (which btw should be done as an altitude-position check and not the reverse ,getting you in a situation where you'll wait for the OM and land before it), one could always check the vnav path,the vertical speed ,the gs capture point on the lnav route also, the vertical display on the Boeing is a big plus.Also the egpws will add a lot to safety.

Remoak,

I was assuming the crew were operating on QNH, maybe I'm wrong about that. Sorry if I am. If they operated QFE then it really is simple math I would agree.

If it is QNH then your math is wrong, with a runway elevation of 763' that is what the altimeter will read on touchdown, give or take a smidge. Add 3900 for 13nm times 300 and you get 4663. In this case you will perhaps see my point as you appear to have made the wrong calculation while in a very benign non time critical environment. Again sorry if I am mistaken re QNH/QFE.

In Europe we do operate QNH. At Geneva on the Vor/Dme they only publish check altitudes on the plate every 2nm so you have to interpolate somewhat. I write the check altitudes out longhand on a seperate peace of paper so I have a profile check every nm. That is after personal experience when under pressure the old grey matter started to fail me. I've seen the same thing happen to lots of people with the resultant poor tracking of the vertical profile.

I do agree that the most effective way of preventing this kind of incident is effective crosschecking and crew co-ordination. Currently in my company the only formal profile check required on an ILS is the approach fix which usually occurs at around 4 Dme. We pactise constant descent approach's so rarly intercept the glide from platform. That said lots of 3 times table involved in manageing that profile so generaly good awareness of profile at glideslope intercept. Once capture occurs I would surmise most people trust the kit and indications subject to either a visual or instrument crosscheck at the appropriate point. I would like to hope that I would notice an unusually high rate of descent coupled with speed runaway and either check at that point or go-around but having flown and instructed on a wide variety of aircraft in both the military and commercialy people do not always react as they might hope they would or think they would or as quickly as they or I would like. Hence my desire to learn from the shared experience rather than see people tut tut at the crew.

At many airfields you can encounter strong tailwinds down the approach. Alicante Rwy 10 and Malage Rwy 13 spring to mind. At glideslope intercept the aircraft pitchs to follow the glideslope. Often the autothrottle does not quite move quickly enough so with a tailwind the speed can start to runaway a bit. People solve this in different ways, speedbrake, intermediate flap to fully deploy the leading edges, gear down or a combination thereof. This of course leads to a higher than normal rate of descent while the aircraft sorts itself out. Something akin to this is maybe what initially misled the f/o/crew ?.

A few years ago we gave crews an unflagged failure of the PF's attitude indicator while in the sim. It would lock at say 15 degrees bank angle a bit nose up in the climbout, IMC of course, just as he rolled out of a turn. All very amusing until one crew crashed and several had to be rescused by the PM. If we re-ran this incident as a sim scenario I wonder how many crews would perform as well as this one ?





I

Maui, sorry missed your post

Of course you should follow procedures and observe step downs, platforms, protected areas etc etc, you should also cross check using available aids as you and others have suggested. If someone sets out to ignore those restrictions and refuses to be corrected then you are quite right they have no place in the industry.

This crew did not set out to be gash or reckless, they prepped thoroughly and briefed thoroughly. Could they have executed the approach better and realised that the glideslope was giving duff gen earlier ? Yes they could of. Hindsight though is 20/20 and I can also see how this could have ended in a CFIT. The crew broke that accident chain and took appropriate action and diagnosed the problem correctly too.

So I would rather pat them on the back for what they got right and look to understand were they could and perhaps should have done better rather than tut tut at them. Always assuming of course that its an isolated event. If thats dick licking and ego stroking then I guess I'm guilty. If we sacked every crew or pilot that at some point made a basic error there would be no one left flying.

Ashling

When I said "altitude/pressure correction at your discretion" I meant adjusting for QNH, and, if you want to get really picky, pressure correction. Sorry that I wasn't clearer.

The math is still simple - 13 x 3(00) gives you height above the airfield, if operating QNH just add the elevation and voila! If using QFE it's very simple, if you are still at a FL at 13 miles (which you could be in parts of Europe) the problem is a little harder, but not much.

More to the point, every pilot should know the basic heights, ie 10D/3000' AAL, 5D/1500' AAL etc. These are the most basic tools in the pilot's bag, as are a standard set of descent rates, these are even printed on the plates so no excuses there.

Nobody - certainly not me - is suggesting that these guys should be villified or punished. They made a mistake, I have done the same after some six-sector days, we all do. However, when I make a mistake, I don't try to cover it up or excuse it, or make out that even though I screwed up, it's actually OK because I didn't hit anything. Nobody learns anything that way, so all power to Air NZ for approaching the incident as they have. However, it doesn't excuse the crew for not exercising the most basic aspect of airmanship - maintaining SA.

As far as tailwinds and descent rates go, I too have experienced Malaga and Alicante (and let's not forget Funchal). We all learn how to deal with that, so it should be no great surprise unless the captain concerned is either very new on type, or very inexperienced. This one wasn't. If it is so bad that you get that "slightly anxious" feeling, you can either try and ride it out and stabilise later in the approach (possibly risky), or chicken out and try again. It shouldn't be a problem if you are ready for it, and you should be ready for it, because your nav display has that handy little wind vector to tell you what is going on (or just the DME if you are flying something steam-driven). SA again.

There are really no excuses, only lessons.

Remoak

I agree. But I still get 4650 on the alt at 13 miles ! the way I read your math you got 3150.

Haven't had the pleasure of Funchal yet ! but I'll look forward to it when I do.

Certainly no excuses but reasons and lessons its important to try to understand both.

Cheers

Didn't this happen for real in South America a few years ago? It was dark, intermittent connection on the attitude indicator(s) and they flipped the plane over. I thought that was a design fault as well, if one of your instruments loses its input signal you don't want it to just freeze, you want it to clearly tell you that it's not to be trusted so you can try something else.

I must admit I am very surprised that an airport supporting the type of operation involving fully loaded 767's lacks some form of approach radar.
I would have thought operations to such islands would require some sort of surveilance at least during the intercept stage prior to the FAF.

F900EX,
Do you actually FLY aircraft?