That skydeity apparently forgot that he had disconnected the autothrust and expected the engines to spool up! He was looking out visually for the runway, manipulated the side-stick but not the throttles. Mental fog, possibly.

This is very scary to say the least ... gimme a firefly bob up front please anytime ..against one of the new breed of p2f boy wonders...its reasonable to expect that the whole crew can actually fly the aircraft without an autopilot eh? or maybe not ...:=

The real problem is the current malaise of pilots' over-reliance on the automatics coupled with some carriers' refusal to allow their pilots to practice their manual flying skills - (at a suitable time and place of course.) Automation has become so good and reliable on all modern types that we naturally expect it to work all the time.

Pilots no longer scan the way they needed to while hand-flying and I've found it very informative to cover the engine instruments during an auto-approach and ask the other guy what the thrust setting was. 90% of the F/Os didn't know, which frightened the life out of me. Try it some day for yourself - you may be horrified too.

It's all in that excellent video about the children of the magenta line. It all works so well 99% of the time that we'd better really know our automatics when it all goes pear-shaped - regardless of whether it's an A or B a/c. Unfortunately, from 20 years of TRE experience, I've found that most pilots don't know it. Then add in the surprise factor and it can all go badly wrong in a very short time.

Hetfield,

Could you have a look to the Avherald incident description ?

My suggest is only to avoid non-sens comments in this topic.

So why don't you delete your post?

Air France is like korean air in the 90`s. Almost everyear they crash an airplane and have some incidents.

Better if they started recruiting koreans to improve their safety record.

Conclusions yet to be done, if you know all the circumstances of the event.:=

WHATTHEFUH is spot on and the training depts. of most Western Airlines also know it. Slowly we find more bits and pieces being added to LPC/OPC's which point to line pilots being found to have poor basic flying skills as well as only scant grasp of the autoflight system. We have recently had stall recovery, manual handling details, Unreliable speed drills, mishandled landings heavily covered, and you can sense from the briefings that the attitude is 'we want to make sure you know the basics'. Unfortunately, the elephant in the room is the massive growth in the number of inexperienced Captains and very inexperienced FO's being rostered together. Everyone knows its only a matter of time before one of the LoCos has a serious loss attributable to crew experience but seem curiously unwilling to address the problem for fear of offending those few who are getting hugely rich in the background.

I thought that the post by Gretchenfrage was of value. Knowing what input the other Pilot is making without asking them, would be a design feature.

However, Clandestino spent a long time dissecting that post and added;

''Alpha floor is generic term for automatic high power command when AoA gets way too high. Even 737 NG have it and on them it's called...Alpha floor''

I must have the old version of the handbook then! ;)

Clando, do your self a favour and please reread the post and try to understand it. :ok:

In the meanwhile there is an update on avherald.

AF7633 was recovered by alpha prot not alpha floor.


Incident: Air France A321 at Paris on Jul 20th 2012, speed drops to alpha prot on approach

:D:D:D

This is why I did not delete my post...:rolleyes:

A318

And your message is?

No matter if alpha floor or alpha prot, like you know there are only a few knots difference.

Anyhow this flight was about 30kts slow on final approach.

" Everyone knows its only a matter of time before one of the LoCos has a serious loss attributable to crew experience."

macdo, what has this to do with LoCo?

Is AF a LoCo?

Were the AF crew inexperienced?

To go back 4 or 5 pages, this got as bad as it did because of lack of awareness, which is just as likely to be caused by complacency and over reliance on automatics as on lack of experience.

Re generic alpha prot. #64, #73,
AFAIK the generic mechanism is to control alpha with elevator (pitch command); the application of power is normally via trigger to a fixed thrust setting.
The difference between a fixed power setting and control is important as there may be systems / situations where additional power is not available.

Is that the landing checklist was conducted by the PNF ??
The item autothrust is included in this list
So the PF must know !

Odd, I used to chastise my flight students when, on final approach during an instrument approach procedure, they were unable to keep their speed within + / - five knots of the briefed approach speed.

I would presume that professional pilots have slightly tighter standards, given how very important a stabilized approach is. Granted, on a gusty windy day, or with a lot of turb, you may get some variations that you don't normally get, but among professionals the pride of how darned well one flies the approach was once a given.

So the guy not flying in this case, on final, was doing the flying pilot what favors by not singing out "five knots slow" ... and upon not hearing "correcting" ... where then the standard follow up to the first alert that the plane is off of profile ... to make sure the man flying is in the game?

Automatics, all well and good, but if the aircraft is not flying the approach to spec, the pilot must. One needs to be mentally in the game. HAL may get a stray voltage, or a spurious input, or even a spurious human input, and then it's yours!

I don't think I've said anything new here, but I do worry that what I discuss above about pride in precision may no longer be a foundational assumption.

Or maybe I am wrong.

On the bright side, the last few commercial flights I was on were pretty smoothely flown. This includes one where the flight deck crew initiated a go around because they were not happy with the aircraft on the ground encroaching to close to the runway.

Was that HAL, or were the pilots flying the aircraft?

It also includes a crew for American Eagle who were on final into DFW but waved off due to, as they explained over the PA system to us after they were established and getting sequenced in for our approach, that they had been put into an interval that was not sufficient. I got the impression that someone had cut in front of them, or was being sqeezed in, but as I wsan't in the cockpit, I am not quite sure what was the actual situation.

There are still good crews out there, and thanks for that. :ok:

Here it seems that none of both was flying the aircraft :)

Not true. This is why the ELACs, SECs and FACs are duplicated with functional redundancy across all six units. HAL is a *bad* analogy because it implies a centralised decision-making apparatus, when in fact it is distributed and redundant.

Could have been either. If autopilot and autothrust are engaged and the FMC is properly set up, a TOGA command will cause it to fly the escape/go-around procedure automatically. If the approach is manual the pilots can either do it off their own back or follow the FD.

In the case discussed .. it was not one .. or other
It was an approach mode "hybrid"
Part controlled by automation ... and other (autothrust) controlled ?? manually ...
A mixture of genres that was dangerous in this case
So .. not recommended

And I emphasize again :
Is that the landing checklist was conducted by the PNF ??
The item autothrust is included in this list
If yes (I hope that AF pilots comply checklists!) the PF must know !

Dozy, I don't care that you wish to nitpick the analogy. I am keenly aware that AFCS and AP systems have depth and range of features. I actually spent some years of my life flying. As noted early in this thread, one of the robotic features kicked in to remedy a well out of standards deviation. That is the core of my point, and it is my position that the maintaining of a professional level of standards is a HUMAN not ROBOTIC responsibility and task. It shall not be delegated.

HAL is a metaphor for letting the robot, of whatever complexity or advanced design, be in charge.

You are reaching here, and not making yourself look any sharper.

Lonewolf, save your breath.

It's as if these guys repeat the never ending " ..... resistance is futile ..... ", but in the end common sense prevailed, at least in fiction.

Let time heal their wounds, I'm inclined to say, but I start doubting common sense. Like even fiction is less cynical than the truth.

Maybe because in fiction no one pays the lobbyists ..... :cool:

That's a little harsh in my book. It's not a case of "nitpicking", it's a case of promoting understanding of how the systems are intended to work as I was taught. If I get something wrong through lack of specific knowledge or lack of line piloting experience, I'm happy to respect that input, put my hands up and apologise. But if people stay stuck in their own mental trench based on an opinion that doesn't reflect reality, then we'll always just keep going round in circles.

This isn't about depth and range of features though, it's about the ethos behind the systems design - which seems to have been significantly lost in translation, hence all the rubbish about beancounters and intent to sideline pilots, and as such bothers me greatly.

And the systems were designed with that in mind. The alpha prot and alpha floor features were designed to fulfil two primary requirements. Firstly to allow the pilot to make aggressive control inputs in an emergency while keeping the airframe stresses within safe limits, and secondly to provide one last line of defence when things turn CATFU'd

But that's not the case - the pilots are always in charge of the systems. Autopilot and autothrust can be disconnected just as they can in any other type. If a pilot or pilots come over time to rely on the automation too much, that's not a type-specific issue, it's an industry-wide one.

Do me a favor, Doze. Remember which person you are having which argument with. I do not appreciate being tarred with someone else's brush.

HAL isn't just an employee of Airbus.

Capisce?

That part was for general consumption - not directed at you personally. The point is that if people want to write what I'm saying off then that's fine - but I'll still happily respect the input of anyone - even those who drive me mad - because I'm always trying to learn something.

HAL isn't an employee of anyone or anything - he/it is a fictitious construct based on an extrapolation of technology that was state-of-the-art in the '60s (i.e. large mainframes). That way of doing things was on the road to becoming obsolete by the end of the '70s. Artificial Intelligence has not caught up to Clarke's vision, because it became apparent that the problem was far more complex than anyone realised at the time. However, the advent of processing and volatile storage on silicon meant that the level of redundancy and systems bandwidth available on microtechnology leapfrogged 2001's estimate by quite some way.

While we're on the subject of respect, as a software guy and systems engineer I find references to HAL regarding modern FMS/FMC and FBW technology about as upsetting as you'd find me assuming that all pilots were akin to a combination of Captain Queeg and Captain Demerest - and throwing that assertion into the mix every few posts.

Get that microchip off of your shoulder, boy.

OK, so in the AF447 thread HAL is the enemy because, er, he didn't take control, as the rules said. In this thread, HAL did, as the rules said, and it's his fault.

And your last point is meant to be an argument, apparently.

Sorry, poultry man,
1) you don't get it, and
2) no, you are wrong in your attempt at a summary.

Try reading what I have posted in this thread, to include the following statement. It's only a few posts up, but I guess that's too much work for you.
There is probably a class at your local college on how to improve your reading comprehension. Suggest you enroll.

Seriously, why so combative? Have I somehow been discourteous to you in a way of which I am currently unaware? Sparring with CONF iture and Lyman/bearfoil, I will occasionally unleash a little sarcasm in jest because otherwise the conversation would be downright depressing - but I don't think I've deliberately said anything to you that could be considered a slight...

Regarding your post #111, where you and I diverge is that you seem to consider the existence of the technology itself to be the reason for the apparent decline in handflying and problem-solving skills. The manufacturers have little say in how their customers apply that technology to training - all they can provide is guidance on how their product is operated. The technology in and of itself is nothing more than a tool. It was not intended to sideline or replace the human pilot, it was intended to assist them.

If the airlines have decided to use technological advances as an excuse to cut back on training costs to the extent that safety has been affected, then the onus is on them to correct it. If the industry wants to continue to expand to the point where automation is mandatory in certain types of airspace, then they must increase hands-on training for when the fit hits the shan to make up for it. If a plurality of pilots feel that the technology has been misused by the airlines in such a fashion, then they should band together and do something about it instead of railing against what is, at the end of the day, nothing more than an inanimate object.

Is that the airlines or the regulators who set the trainings required (mandatory) and the rules for apply them ?
Who decide ? .. airlines or regulators
No laws for training ? .. just curious :)

The regulators specify general minimum standards, but they're just that - minimum standards.

Spoken like a true Techy guy Dozy, manufacturers have an ENORMOUS amount of influence with respect to regulators and operators.

Again Dozy, easy to say, and rather nieve.
Firstly, airlines have been sold a concept that more automated and envelope protected aircraft are inherently safer, which based on past experience is statistically correct, the trouble is it has created a raft of other issues, some of which include a continual erosion of piloting skills and organisationally a loss of respect in the profession as well as a new set of complex unforseen failure modes that overly clever designers and systems engineers never even contemplated.....after all the technology is so advanced and aeroplanes can fly themselves don't they?
You can inform us as eloquently as you like regarding the logic and inherent redundancies built into a system..but the fact remains that aviation is still the cruel mistress when things go wrong....just as it was recently where on a final vector to the ILS at our destination with the A/P engaged we flew through a localised area of turbulence that was powerful enough to roll the aircraft to 30 AOB and disconnect the A/P....pilot intervention prevented any further roll deviation and a departure towards the terrain 2000ft below us.

So that's what I figured ... regulators are to criticize if there is a lack of training
Their responsibility is total
If you want a better workout .. regulators can only regulate and impose
Suggests that airlines will invest in training for pleasure more than has rules and laws is a sweet dream :)

Isn't "close to Stall" where this filly plays her top cards?

"where you and I diverge is that you seem to consider the existence of the technology itself to be the reason for the apparent decline in handflying and problem-solving skills"

Dozy, are you talkin' to me, or the fifteen or so posters whom you joust with on a regular basis?

What you say there is blatantly false.
Not.
Even.
Close.

Try not to attribute to me that which you have made up in your own imagination.

I'll try to be a bit less snarky.

Deal?

AutoThrust/Manual
 

Having read six pages of expert comment on near stalls due to Auto Thrust/Throttle failure, please tell me gentlemen, what does a captain do with his right hand when flying an approach/landing on any Airbus?. Just askin!.

Well, it depends..., in my outfit it was

- A300/310 keep right hand on thrust levers/throttles (!?), even flying with A/THR
- A320/340 hold your coffee, because SOP wants you to fly with A/THR on, even flying manually. But this has changed many times, yes, after incidents/accidents.

To be fair, I included the qualifier "seem to" very deliberately for that very reason. Also because I wasn't 100% sure, but that was the vibe I was picking up from your posts. That way I hoped to make clear that you were more than welcome to set me straight if I was wrong.

Done and done. :ok:

I disagree. This is commercial aviation. Certainly since the end of the '60s, no manufacturer has built an airframe solely on the basis of their own wishlist, because developing and selling a product no-one wants to buy would be commercial suicide.

For the last 3 to 4 decades, manufacturers have developed products by polling customers at every level (meaning management, finance, crew and maintenance) to find out what the priorities should be for the next-generation product. So - breaking down your next section:

Let's start by separating automation from envelope protection and FBW, because they're completely different concepts.

With automation the improvements in reliability and safety come from the fact that machines, whether electromechanical or electronic, tend to be better at doing mundane, repetitive tasks than humans. From the earliest wing-levellers through the alt/hdg hold autopilots to the modern FMS-driven systems it is primarily this technology that has encroached on hand-flying practice while on the line - and as you say, statistically every generation of the technology has led to improvements. The INS- and radio-driven analogue units that preceded the digital ones were the first to effectively allow autoflight from wheels-up to top of descent, or even autolanding (which was developed primarily by us limeys because of the operational problems caused by the infamous pea-soupers at Heathrow). The move to digital didn't really alter this practice as far as I've read and it was certainly never intended to be used as a way of de-skilling pilots (though I'm sure it was probably perceived that way in some quarters even decades ago).

Digital FBW and envelope protection is a newer concept in civil aviation, although it's been on the line now for 24 years (36 years if you count analogue FBW as used by Concorde). There's been a lot of rubbish talked about it, largely through lazy or intentionally inflammatory journalism.

Digital flight control in general provides a cost saving via weight reduction (which pleases the management and accountants) and reduces points of mechanical failure (intended to please pilots and crew) by simplifying mechanical implementation through fewer moving parts (hopefully making maintenance easier).

Envelope protection is a side benefit of digital control - particularly with regard to airliner operation and like FBW it is a very misunderstood beast. The press liked to play up the aspect in which it can help cover pilots if they have a bad day at the office, but the other primary benefit - equally important if not more so - is that when the systems are functioning (i.e. well over 99.9% of the time), it can allow a pilot to make emergency manoeuvres without having to worry about overstressing and damaging the airframe. The press tend to ignore that aspect because it has less chance of generating controversy of the kind which sells newspapers.

Then surely the onus is on the airlines to counter that problem, and on pilot unions and regulators to make the airlines do it.

What you're talking about there is not restricted to pilots - it is the advent and primacy of managerialism in every aspect of business. Search PJ2's posts for some very eloquent analysis of the problem.

The systems (FBW and envelope protection in particular) were specified with pilot input, and none of the systems features were implemented without sign-off from pilot engineers. The idea that the digital systems were cooked up in a closed room by techies and beancounters simply isn't true.

And while grumbling about clever-dick techies may be cathartic, I can assure you that they, like you, are front-line staff and as such, just as subject to the capricious whims of management and markets as airline crew.

Most technology-related accidents have tended to be INS/FMS/autopilot related rather than flight control issues. In engineering terms, the failure modes of digital FBW systems are actually relatively simple in nature because they were designed in such a way as to be an interconnected set of very simple systems which, as a whole, can perform complex tasks. Even then, what the systems are doing is not especially "clever" or "complex". As for "unforeseen", the recent example of AF447 stemmed from a blockage of all 3 pitot tubes - which was not merely unforeseen by the designers and aero/systems engineers, but also by the pilot engineers. Preceding hull-losses stemming from loss of air data had only one or (exceptionally rarely) two pitot tubes blocked at once.

Another press-originated fallacy. Don't get me wrong, some industry characters did suffer foot-in-mouth syndrome, but what was printed for general consumption (i.e. outside aviation specialist publications) was shorn of a lot of important context and therefore misconstrued.

I think it's very important to understand that in Airbus's case, the training cost savings to the airlines from going to digital FBW technology were primarily related to having all models in the range from the short-haul narrowbodies to the long-haul widebodies having an almost-identical flight deck layout and feel - i.e. conversion training would be cheaper. *Never* was it suggested that the technology was so good that airlines should feel free to cut back on existing pilot training.

As an aside, the first time I heard that phrase was in the movie Airport 1975 in reference to the Classic 747, which featured no digital technology at all.

As far as I'm aware there have been no hull-loss accidents related to the failure of a FBW system in the 24 years they've been flying commercially.

Well done (and I mean that sincerely). I don't see how it would have been any different on a previous generation airliner though. I'd be interested to know what type it happened on.

[Apologies for the essay everyone!]

No apologies needed, that was a good one Dozy. :D

hmmmmmm... and that from me.....:confused:

AF have acknowledged “malfunctions” with Pitot tubes on A320s. Do you think one could be on the list of culprits?

Why?, a pitot tube is such a simple device and preventing them freezing (about the only thing that stops them functioning once in flight), is equally straightforward. Its fully developed, known technology, they simply shouldnt be a cause of problems. all the crap they plug into them, well thats a different matter