Interesting ILS failure mode

[DFC]

His reply is correct according to the rules, of course

Not quite.

Given conflicting information, the correct answer is to disregard both, get the aircraft to a safe level and position and then unless one is able to resolve the discrepency make an approach using a different aid either at the same aerodrome or at the alternate.

The the case being discussed that is exactly what the pilots did. They did not jump onto the VOR procedure and make an attempt to save the approach, they went round again, took time to think about the situation and then made a safe approach using a different aid.

In the private world of flying, pilots and owners are very slow to report radio faults because it is goping to cost money. Hence you hear "watch the ADF it is a bit weak or is a bit off". In the professional world that is unacceptable.

Regards,

DFC

[Fuji Abound]

I think in an interesting debate such as this it detracts from the quality of the discussion if remarks are made presumably purely with the intention of antagonising or provoking a response - as much as some might say they add to the pithiness of the debate.

[High Wing Drifter]

One day I asked him: if you were flying an NDB IAP and the ADF was saying you are OK, but a handheld GPS was telling you that you are about to die, what would you do?

The obvious answer is go-around, but what if the GPS says you are fine and the ADF says you are going to die? Why not use the method on the plate instead and avoid all confusion and don't die for either reason.

That has little relationship to the specific scenario in the film, in which case the answer would be to correlate position against the DME and height as defined on the approach plate. If you used the GPS distance in such a scenario then you might conceivably be just as dead. Of course you can always recalculate an unofficial crosscheck, but who and why!?!

[IO540]

Obviously the thread has got a bit sidetracked. What suprised me is how apparently easily, in the modern age and flying a modern airliner, the pilots allowed a gross positional error to remain undetected until somebody saw something strange out of the window. Presumably in IMC the result would have been different.

I think the sensible answer is that one should go-around if there is any discrepancy in instrument indications. This unfortunately is not very helpful in the "ADF" scenario where a huge error is pretty common.

[Fuji Abound]

Having watched the video and read the report I agree with the comment made in the video - before this incident some “pilots” assumed the ILS was always accurate unless otherwise flagged. Therefore, despite the NOTAM, this crew assumed that because it appeared the glide slope had been captured, it had been captured. The doubts came from the unusual pitch change and the external visual clues, not the DME, but were in them selves not sufficient to provoke the crew to react immediately to the lack of correlation in the information available to them because they were convinced they had captured the glide slope, which, in their experience was “infallible”.

DFC suggests that what is to be learnt is the importance of cross checks - never relying on a single instrument. He is clearly correct.

I0540 suggests a moving map GPS may have helped. In this instance the primary cross check had to be of range. Whilst this information may have been provided by almost any GPS, in many cases would the information be presented clearly, and notwithstanding the integrity of the GPS, would the range be the required range? For example, the range might be that to the center of the airport not the threshold. In other words if the pilot were cross checking with his GPS in the “amateur” world of GA would he necessarily know precisely what the GPS was telling him? The aircrafts position on the moving map would have been of little help.

In this particular case the correlation between the glide slope and the DME provides the best instrument clues that something is wrong and in themselves are sufficient to provoke the procedure to be aborted.

It is of course suggested that the localiser can never be susceptible to this type of error. Does that mean the localiser can be relied upon without any other cross checks? Is there any scenario in which the localiser might provide incorrect information without the crew being given adequate warning? In this instance a cross check with a moving map would provoke similiar concern that all was not well. If it did, and the moving map GPS happened to be one of DFCs three and six boxes on which would you rely?

In light aircraft I am becoming a great devotee of the G1000 displays. However, even with these displays this emphasises the importance of understanding the information with which you are provided, and the source from which the systems are deriving the information.

[LD Max]

Fly with 1/2 a dot fly down indication?

Fascinating video. I note the YouTube version is edited to fit the permissable download time, and the person who posted it admits to editing out a lot of the technical lessons learned in terms of the technology.

So if this has been suggested before, I apologise.

It occurs to me, that a modification of the intercept procedure would eliminate the possibility of this particular error. Before anyone jumps down my throat, I am NOT advocating this should be done outside SOPs, but only that we consider this as a rhetorical suggestion.

A normal GS provides a beam width of 1.4 degrees. 0.7 degrees above and below the GS, i.e. Full Scale Deflection of the GS indicator represents 0.7 degrees. Most HSIs show 2 dots above and 2 dots below the GS. Each dot represents 0.35 degrees and 1/2 a dot would represent 0.17 degrees. At a range of 1 mile this equates to around 17 feet and at 1/2 a mile, approximately 8 feet.

It seems to me, therefore, that pilots would be advised to fly the entire procedure with a 1/2 dot fly down indication. (Or recalibrate the display so that glideslope intercept occurs 0.17 degrees ABOVE the actual glideslope.)

This would mean that (especially) the Autopilot would continue to fly straight and level on the localiser until a 1/2 dot fly down indication is achieved. As we have seen this will have an insignificant effect on the profile or performance, but it WOULD provide an airborne failure indication of the GS lobe transmitter.

If no fly down indication is forthcoming, the aircraft would continue to fly straight and level and the autopilot would not commence the GS descent profile. It would then quickly become obvious that the aircraft had flown through the glideslope without descending - upon which a missed approach procedure can be commenced safely and without any particular hurry.

Just a thought.

Regards,

LD Max

[IO540]

I must be going thick, LD-Max, but I don't follow the meaning of the half-dot. Maybe I missed something in that movie.

Normally, one intercepts the localiser first, and one is in altitude hold (whether flying manually or on the AP).

Slightly before the GS intercept, the GS flag goes away and the GS indication is UP. At the GS intercept, the GS indication passes through the middle. At this point, one tracks the GS (whether manually or coupled).

If one is going to be receiving a false GS of any sort, I don't see how the half-dot would help, because the false GS could still look "right".

[dublinpilot]

I think his point is that if the g/s indication shows something other than perfectly on the g/s then you know it hasn't failed as per the video. Because if it had failed as per the video, then the g/s would be showing perfectly on the g/s. Hence his suggestion to fly slightly off the g/s.

Interesting video....thanks for posting.

dp

[IO540]

OK, I see it now. That wouldn't help. As a designer of electronic gear for some 30 years, I can say the system could fail with a half-dot indication almost as easily as with a perfectly zeroed indication. Especially if the half-dot offset was designed-in as the default indication when captured.

There is no way to solve this problem comprehensively, other than some kind of digital verification (with appropriate security) transmitted by a separate monitoring unit.

Do airline autopilots fly with the indicator (the HSI, or the flight director) perfectly centred? GA autopilots certainly don't; well not unless it is dead calm.

[englishal]

The thing with MM GPS is that it gives you an instant picture of where you are, where you should be, when the Loc / GS is going to come in - and if it doesn't you know instantly there is a problem. All this without twisting , tuning and trying to fly the aeroplane. Had this crew had MM GPS they would have known BEFORE they intercepted the loc that there was a problem.

Seems almost criminal to not include a few grand piece of kit in an aeroplane costing 10's of millions which could save a couple of hundred lives. There have been numerous serious accidents which WOULD have been avoided if this kind of kit was onboard. In fact a friend of mine killed himself, his co-pilot, his god children and their mother by flying into a mountain on an IAP. Had they had GPS it is 99.9% certain they would be here today.

[DFC]

In fact a friend of mine killed himself, his co-pilot, his god children and their mother by flying into a mountain on an IAP. Had they had GPS it is 99.9% certain they would be here today.

Two points regarding such statements.

1. They were not on the IAP if they hit a mountain. Perhaps one should ask why they were not on the IAP. Or if they were, why did the IAP fly them into the mountain.

2. Unless the mountain is marked on the GPS and flagged then there is little to be gained.

If extra or new equipment could have been used to solve the problems that this incident highlighted I would expect that it would have been highlighted by the investigators or even the people who made the video.

Never assume that adding more information into the situation will be the answer. It can be simply another distraction or the final bit of information that leads to overload.

Moving map displays for Large commercial aircraft are basic in terms of the information they provide and very well filtered to remove all unnecessary information. They are also very expensive.....never mind Garmin430 that would be peanuts in cost terms.

The larger type manufacturers are moving to the areas of terrain awareness and warning systems as well as HUDs and synthetic vision to improve safety.

Regards,

DFC

PS: Here is something to do - Put your handheld GPS in simulator mode and start flying round an airport a long way from base that you never visited. Eastern Europe or anywhere that you have absolutely no idea of the terrain or airspace. Then ask yourself if it improves your situational awareness or if in fact it takes so much playing with that you have no spare time to do anything else!

[englishal]

Clearly everything is black and white to you DFC.....The trouble is in the real world there are a million shades between black and white.

1. They were not on the IAP if they hit a mountain. Perhaps one should ask why they were not on the IAP. Or if they were, why did the IAP fly them into the mountain.

2. Unless the mountain is marked on the GPS and flagged then there is little to be gained.

They failed to go missed at the MAP. It was a VOR DME approach and although we don't know what happened for whatever reason they never went missed. Which meant they hit the mountain 15' below the summit. HAD they had a GPS, it would have been painfully obvious that they had passed the MAP, and if it was more spohisticated with TAWS (which many / most have these days) they would have seen their course ending in a big red splat. Surely even for the most stupidest pilot out there that would start getting the alarm bells ringing.

Oh and for the record, this wasn't some silly little PPL CFIT we're talking about, this person was a JAA ATPL, FAA ATP, CFI, CFII and MEI. "Co-pilot" was ATP.

I have been in a plane (the back of) attempting to intercept an ILS....except there was a nearby VOR dialed in (very similar frequency, 112.4 instead of 114.2 or something like that). HAD we continued and flown a LOC approach (because GS wasn't coming in) we'd have hit the hill the VOR was on. The GPS showed us flying towards the VOR and 2&2 clicked and the situation was "saved".......

[Fuji Abound]

Firstly, this is a GA forum - that is why it is called private flying.

The discussion arose out of an incident in a commercial aircraft, with three crew. That environment is totally different from the average SPPPLIR ops.

Clearly in a commercial aircraft there is no possibility of operating with a three and six unapproved GPS unit - end of discussion.

In GA however, a three and six unapproved unit is possible, as is a panel mounted Skyforce, a panel mounted G1000 or an Avidyne unit.

If you have used a G1000 you would be totally convinced that the situational awareness this gives, coupled with the seamless integration with the automatics, procedures, weather and terrain information and avionics is a huge leap forward in the improvements it brings to situational awareness, cockpit management and workload. Could a professional pilot perform as well in SPIR ops with “traditional” equipment in all circumstances - of course not. However, it is not that simple. There is a great deal of information available, and the smooth operation of the kit in a potentially very high work load environment demands familiarity. Without that familiarity it is another distraction, the consequences are potentially disastrous. It is not a panacea.

At the other extreme are the typical moving map non approved units used by many.

They were not on the IAP if they hit a mountain. Perhaps one should ask why they were not on the IAP.

No - that is not the question to ask. If every pilot was always on the IAP this type of accident would never occur. Yet the accident record is “littered” with pilots of considerable experience who find themselves in just such situations. In short pilots are trained with all the expertise and experience accumulated over the years by instructors not to let this happen, but it still does. Would a moving map of any type provide a secondary reference that the aircraft was not correctly established on the IAP or might it add to additional confusion, where none had existed before?

In my opinion IF the pilot is familiar with the unit so that it does not become another distraction then it can only aid him in situational awareness. If only once it leads him to doubt that he is correctly established on the IAP and enables him to reposition it has proved its worth.

Unless the mountain is marked on the GPS and flagged then there is little to be gained.

I don’t follow your argument. Many incidents of this type are caused by the pilot missing a clue that they are at some other point in the procedure. In Englishal’s case we can all conject why the pilots did not go missed, however in addition to the information they should have had from their DME or from the markers they would have seen on a moving map that they had passed the threshold.

In short I am not suggesting relying on a moving map, particularly in an unapproved fit, but it aint half a bad backup which will give you a clue that what the primary instruments are telling you might just not add up!

Englishal - sorry to hear about your freind.

[LD Max]

1/2 a dot

Originally Posted by IO540

OK, I see it now. That wouldn't help. As a designer of electronic gear for some 30 years, I can say the system could fail with a half-dot indication almost as easily as with a perfectly zeroed indication. Especially if the half-dot offset was designed-in as the default indication when captured.

There is no way to solve this problem comprehensively, other than some kind of digital verification (with appropriate security) transmitted by a separate monitoring unit.

Do airline autopilots fly with the indicator (the HSI, or the flight director) perfectly centred? GA autopilots certainly don't; well not unless it is dead calm.

Thanks for the reply. Apologies for not getting back sooner... been busy.
I think what you are saying is that if the lobe transmitter was re-calibrated as I suggested then a failure would still show half a dot fly down. Actually that's not what I meant.

I merely suggest that if you intercept the G/S as normal, but delay the descent until you get a 1/2 dot fly down indication then this would provide a failure indication of the G/S. The reason being that if the lobe transmitter had failed, you would always get a centered needle on the G/S indication even after you had flown right through the glideslope.

My suggestion to re-calibrate the display is simply to do with the aircraft. If the display were recalibrated so that a centred indication was given when the aircraft was, in fact, flying 0.17 degrees above the glideslope, then pilots (and autopilots) would not have to modify their normal procedures. In this case if the lobe transmitter failed, then the pilot would see a permanent 1/2 dot FLY UP indication.

Either way, no descent could occur because the aircraft would continue to fly straight and level trying to intercept, until someone twigged the DME was indicating they had flown through the glideslope. Much safer, I think, than starting a descent before the DME starts screaming at you.

In terms of failure indication, you are quite right that separate monitoring is the only sure way to know. But in the case of this incident, the ILS was notamed as "unmonitored" - meaning that the normal fail-safes were out of the loop. Since the pilot has no failure indication other than the flags, this particular mode of failure was impossible to identify in the cockpit.

My suggestion to fly the procedure with a 1/2 dot fly down indication would provide an additional means of identifying such a failure.

You're also right of course that it would take a very skillful pilot to fly a perfect profile and perhaps the "perfection" of a "perfectly centred" G/S should be treated with suspicion... However this particular crew were coupled to autopilot and would be used to flying a very accurate profile, which in itself did not indicate anything unusual.

Regards

[S-Works]

I had an interesting experiance flying the ILS into Cambridge yesterday. GS was indicating normal and then about 2 miles from the MAP it started fluctuating up and down at half scale in either direction. Now I have been known to have bad days but not a half scale in both directions.

It turned out that an aicraft had joined a tight base leg and flew accross the GS. Had I not been monitoring the DME/Height profile it might have ended in tears!

[LD Max]

Originally Posted by englishal

I have been in a plane (the back of) attempting to intercept an ILS....except there was a nearby VOR dialed in (very similar frequency, 112.4 instead of 114.2 or something like that). HAD we continued and flown a LOC approach (because GS wasn't coming in) we'd have hit the hill the VOR was on. The GPS showed us flying towards the VOR and 2&2 clicked and the situation was "saved".......

Sorry to hear about your friend too. Any accident is tragic and it's all very well to treat such events as lessons to the rest of us, but I do think it important to learn the RIGHT lessons.

The right lesson to be learned from your own anecdote is the importance of "identifying" the radio aid.

Firstly, note that neither of the two frequencies you mention (and I realise they are arbitary examples), are ILS frequencies. The LOC transmitter operates in the “normal” airband VHF frequency range between 108.1MHz and 111.95MHz. Note the frequencies are all assigned odd-tenth decimals.

VORs on the other hand operate between 108.0MHz to 117.95MHz. The band between 108.0 and 112.0 is in even-tenth decimals to preclude any conflict with localiser frequency assignment.

It is therefore possible to mis-dial something like 110.5MHz (ILS) and dial in 115.0MHz (VOR). If you know the rules, this should become obvious when cross-checking the radios during approach checks.

Secondly as you know, all LOC transmitters identify themselves by Morse code. Localisers always start with the letter "I" (dit dit). I believe there are no VOR idents which start with the letter I, (bit more research needed to be sure). In any case, a navaid needs to be positively identified before use - but the lack of a letter "I" when identifying an ILS, should be enough to alert you to listen more carefully.

Finally, I cannot agree with your point about GPS being used as a "backup". It may have saved your skin this time but just a week ago, a buddy of mine was flying with a GARMIN GPS 430 in a twin, just south of Ocala in Florida. Just for fun, he wrote down the GPS coordinates of where the GARMIN unit said he was. Later, on the ground, he looked up the coordinates on a map. To his astonishment, the coordinates placed him a few mile South East of Kissimmee - about 100 miles away from where he actually was.

We could put this down to a "glitch", "finger trouble" or simply picking the wrong numbers from the display. Either way GPS is probably one of the best ways of being wrong about your position, without any clear indication of the error. There is a saying that GPS can be very precisely inaccurate - and that is the danger.

At the end of the day IMHO, there is no substitute for good IFR procedures.

Regards,

LD Max