PDA

View Full Version : Blind use of flight director could have led to crash


sheppey
2nd Jun 2020, 14:18
Flight International 5-11 May 2020 published a synopsis of the close call to a NEOS Boeing 737-800 at Bristol in June 2019.. Report here:

https://assets.publishing.service.gov.uk/media/5e8edca8e90e070778e2066f/Boeing_737-86N_I-NEOT_05-20.pdf

Yet another instance where blind following of flight director guidance during the process of an unstable approach led to a very close call. Flight Directors were initially designed to be an aid to instrument flying but it didn't take long for these instruments to become indispensable to some pilots even in good weather. Like moths to a flame is one apt description.

As David Davies the chief test pilot of the British Air Registration Board in 1949 wrote in his fine book Handling the Big Jets, "Finally do not become lazy in your professional lives. The autopilot is a great comfort, so are the flight director and approach coupler But do not get into the position where you need these devices to complete a flight."

Today we have rules which dictate when these devices are needed to start or complete a flight. Even so, blind following of flight director guidance without concomitant cross-checking of other flight path parameters as a matter of common sense, is not good airmanship.

Company SOP's that mandate use of the flight director at all times including visual approaches, eventually leads to pilot automation dependency. We see the inevitable result in so many accidents. The NEOS 737 flight director debacle at Bristol was nearly an accident rather than an incident.

FlyingStone
2nd Jun 2020, 14:44
Company SOP's that mandate use of the flight director at all times including visual approaches, eventually leads to pilot automation dependency.

So what's the solution if one flies for a company that prohibits raw data flying?

macdo
2nd Jun 2020, 14:59
FD's are removed for visual approaches on the Airbus family SOP. For good reason.

parkfell
2nd Jun 2020, 15:13
It is not until page 15 does it make clear that the PF flying the first approach was the FO whose experience is not known.

The Captain was very experienced and yet he allowed the shortened route to occur which was the start of the incident. It must have been at the back of his mind that it was going to be interesting?

The U/T approach radar controller also gets a mention (in dispatches) which I am sure has turned out to be a valuable lesson in aircraft performance.

Failing to set the MCP Alt window to the G/A platform on final approach was a critical event. The FO workload would be HIGH and Situational Awareness LOW. Understanding the FMA is critical. G/A.....TOGA displayed???
Speculation ~ the last time a two engine G/A was flown by the FO would have been on the type rating course??
Just why the Captain allowed the situation to deteriorate to the actual point where he took over is a matter for conjecture.....
Rich in CRM especially for MCC/APS courses and probably for junior *birdmen*on initial type rating courses.

The transcript of the CVR would have been most revealing.

The Aerodrome Controller instructed a Go Around iaw MATS part one instructions. He used his initiative as he did not like what he saw.
Is it to harsh to compare the ATCO actions at Bristol to the Aerodrome Controller lack of actions in Karachi?

Note: as gender has not been identified by the AAIB, I have used HE as *’generic gender’ * to refer to either gender or non binary, in these days of political correctness.

BitMoreRightRudder
2nd Jun 2020, 15:22
Handling the Big Jets is indeed a fine book and a must read for any pilot transitioning to jet aircraft. But the advice from D Davies, though fundamentally sound, harks back to a very different time. Automation (including flight directors) has come a long way since the book was written. A modern AFDS system when used correctly is invaluable. I think the updated version of the book, were it to be written, might include a section on the importance of appropriate use of automation and understanding FMAs and what different modes actually do for you. Frankly, the latest generation of jet transports are not designed to be flown manually "with everything turned off". That may leave some pilots aghast but it's true. Hence practising manual flying skills is a regular item in recurrent sim checks - arse about face for many of us, but that is the way the job is heading.

This particular incident has so much more to it than just blindly following the flight director.

jmmoric
2nd Jun 2020, 16:01
Seen from an ATCO's point of view, we are often told to "offer" shortcuts, in order to save fuel etc.... on the other hand, we are also adviced to be carefull to do so during the approach phase, since we're taking valuable time away from pilots.

So the best we can do is ask....

Then from the pilots point of view, some of us are a little "go" minded, feeling we're always ready for what ATC asks of us, we are always ready to help ATC.

So the best we can do, is learn to say "nope".....

safetypee
2nd Jun 2020, 16:20
Setting the altitude value in the mode controller:-

Aircraft type dependent, but the risks in using a procedure requiring Alt Sel to be used as a reminder / safety check at mins, were identified many years ago.

We, individually and organisationally forget.

Or do we change things believing that safety will improve. Management (training) not understanding the system operation in comparison for the desire for standardisation - SOP minded; or if known, then who decided that the risk was tolerable, what were the mitigations - knowledge, communication, training.

If - probably, crews had been caught before, but recovered (not reported), then how can we expect to improve safety.
Why bother to report the event if it is an SOP; a drift to pilot error, yet the root issues are in the technical, operational, and safety management mindsets.

Jwscud
2nd Jun 2020, 16:22
I think in this case, the root cause was being heroically high and fast and not binning it very early. SA was left behind way back up the approach leading to the lack of awareness of modes and aircraft state as BitMoreRightRudder correctly says.

601
2nd Jun 2020, 23:20
Frankly, the latest generation of jet transports are not designed to be flown manually "with everything turned off".

So why don't we do away with pilots and put computer programmers in the cockpit to monitor their software?

SaulGoodman
3rd Jun 2020, 09:39
So why don't we do away with pilots and put computer programmers in the cockpit to monitor their software?
computer programmers are too expensive ;)

stilton
3rd Jun 2020, 10:24
Handling the Big Jets is indeed a fine book and a must read for any pilot transitioning to jet aircraft. But the advice from D Davies, though fundamentally sound, harks back to a very different time. Automation (including flight directors) has come a long way since the book was written. A modern AFDS system when used correctly is invaluable. I think the updated version of the book, were it to be written, might include a section on the importance of appropriate use of automation and understanding FMAs and what different modes actually do for you. Frankly, the latest generation of jet transports are not designed to be flown manually "with everything turned off". That may leave some pilots aghast but it's true. Hence practising manual flying skills is a regular item in recurrent sim checks - arse about face for many of us, but that is the way the job is heading.

This particular incident has so much more to it than just blindly following the flight director. Heard this theory before and it’s just not true, although it can be self perpetuating if you allow it

Of course modern aircraft can still be flown manually, they are certified to do so, it is up to the individual operators and their pilots how much they want to encourage and practice this vital skill

FlyingStone
3rd Jun 2020, 10:56
Of course modern aircraft can still be flown manually, they are certified to do so, it is up to the individual operators and their pilots how much they want to encourage and practice this vital skill

It just depends on the environment you operate in. Mind you that when this book was published, they didn't have any of thte RNP AR and alike and I bet London and New York TMA were slightly less busy.

The aircraft I fly has a demonstrated RNP ability of 0.64nm for raw data using MAP display at 10nm range. You can't even handfly a normal RNAV approach like this.

Denti
3rd Jun 2020, 11:51
It just depends on the environment you operate in. Mind you that when this book was published, they didn't have any of thte RNP AR and alike and I bet London and New York TMA were slightly less busy.

The aircraft I fly has a demonstrated RNP ability of 0.64nm for raw data using MAP display at 10nm range. You can't even handfly a normal RNAV approach like this.
It is not only environment, but also aircraft. I flew the 737 with IAN enabled and with that it is easy enough to hand fly an RNP approach. It was approved and encouraged in the company i flew for and i did it a lot. On the bus it is theoretically possible, but not really pleasant at all and nor encouraged, and depending on operator, not approved, haven't flown the IAN look-a-like FLS, as that is not standard equipment on the A320 series.

This case and report is somewhat puzzling, and yes it sounds like FD fixation without realising where they were (aviate, yes, navigate, no). Dunno, for me the vertical situation display was always a great help on the 737, but apparently that is not standard either.

Meester proach
3rd Jun 2020, 11:56
This is just a switch pigs , but “ know your attitudes “. In other words, know where the aircraft should be , pitch wise for any config.

I’ ve given up trying to salvage iffy approaches , and pack it in on base now instead !

FlyingStone
3rd Jun 2020, 12:02
Denti, Both IAN and VSD are still option-only unfortunately.

Dan Dare
3rd Jun 2020, 12:07
The assistant seems to be a hero of this story being the first to draw attention to the low approach and missed approach, but the role is going the way of the runway-caravan leaving a controller with ever more ancillary tasks and less chance to continuously observe safety essentials such as runway and gear. The first approach wasn’t far off the energy profile being slightly fast and slightly low, but correcting? Wouldn’t this have been a successful, if messy landing without ATC intervention and we’d have heard no more about it? Most ATCOs have zero flying experience and often never see anything other than 3 degree glide path approaches so don’t really have the required knowledge to say when a flight is dangerously positioned. If every missed approach adds a degree of risk it could be argued than an ATCO could lead to a more dangerous situation with an unnecessary go-around instruction than by leaving the pilots to fly the plane.

parkfell
3rd Jun 2020, 14:33
The assistant seems to be a hero of this story being the first to draw attention to the low approach and missed approach, but the role is going the way of the runway-caravan leaving a controller with ever more ancillary tasks and less chance to continuously observe safety essentials such as runway and gear. The first approach wasn’t far off the energy profile being slightly fast and slightly low, but correcting? Wouldn’t this have been a successful, if messy landing without ATC intervention and we’d have heard no more about it? Most ATCOs have zero flying experience and often never see anything other than 3 degree glide path approaches so don’t really have the required knowledge to say when a flight is dangerously positioned. If every missed approach adds a degree of risk it could be argued than an ATCO could lead to a more dangerous situation with an unnecessary go-around instruction than by leaving the pilots to fly the plane.

This is always the dilemma faced by the Aerodrome Controller. Clearly the ac position in terms of azimuth, and range (in terms of glidepath) is a judgement call, based on experience at their present (& previous) airfield/s.
Those who have experience at airfields where significant light ac training occurs will have seen some ‘interesting manoeuvres’ and are well placed to know just what “does not look pretty”.

What has not been published are the thoughts of the aircraft commander and the CRM aspects. The CVR transcript might well be very revealing.

The ropey G/A is hardly due to the ATCO. That is simply down to a FO overwhelmed by the occasion and failed by pitch to +15° as the opening bid. Was the Capt as PM at that point MONITORING carefully?
Blindly following the FDs is not recommended. Disconnect the FD, and then rebuild the automatics carefully.

Overall the ATCO gets my vote.

jmmoric
3rd Jun 2020, 15:49
Most ATCOs have zero flying experience and often never see anything other than 3 degree glide path approaches so don’t really have the required knowledge to say when a flight is dangerously positioned. If every missed approach adds a degree of risk it could be argued than an ATCO could lead to a more dangerous situation with an unnecessary go-around instruction than by leaving the pilots to fly the plane.

The pilot has the final word when it comes to the safety of the aircraft, and it is in his right to land the damn plane if he deem it the safest?!

Stating that controllers have not seen anything else than 3 degree glide paths, that's bull**** (okay, I'm not sure about it, but I'm pretty sure most have seen other types of approaches when they're checked out in a tower position), but looking out seing someone approaching the ground very fast, short of the runway, would spark a reaction from controllers.... especially if it's an airliner that would normally not operate like that. And it hard to come up with a reaction that'll get an aircraft to stay in the air, other that calling for a missed approach (which would normally give a lot of thrust and immidiate climb..... except thrust was forgotten in this case)

But you're right, a missed approach is a risk, and I don't recall any ICAO documents stating a requirement of calling for a missed approach in the same way as stated in the report?

lederhosen
3rd Jun 2020, 16:31
I heard from one of our copilots who flew for NEOS that they had some interesting ways of flying the the 737. The one I remember vividly was them being strongly discouraged from using vertical speed mode. I doubt that dive and drive using LVL Change would be most airline's SOP for this kind of approach if VNAV would not capture. It does seem otherwise a pretty straightforward example of rushing a non precision approach and when startled by the order to go-around by the tower getting out of sequence and not setting the missed approach altitude. Our procedures would have had the PNF do this when commanded by the PF, who in the excitement of the approach going wrong does not seem to have noticed this omission. But the usual mantra about poorly flown missed approaches and lack of pitch power awareness also holds true as mentioned by various other posters.