AF447 wreckage found

[MountainBear]

So the answer to too much automation is more automation?

Yes, that's correct.

[ap08]

The answer seems to be simple.... Pilots are no longer trained to interpret AoA indications.

You can't be serious. What is so difficult about interpreting an abnormally high angle of attack as a sign of an impending stall?

[Lonewolf_50]

The answer seems to be simple.... Pilots are no longer trained to interpret AoA indications.
Concorde had an AoA indicator.
I take it from several posts here: so do Navy (carrier-based) aircraft.

So too Navy training aircraft. (And IIRC, USAF T-38, but that may depeond on the mod)

I doubt the average A or B pilot would be able to get any usable info (SA or otherwise) from it.

ChristiaanJ, I respectfully disagree with you in general principle. AoA in certain flight conditions is a useful scan and cross check.

That said, as I mentioned a few posts back, choosing where on an instrument panel one places an AoA, and in what flight regimes you train and expect your pilots to use it, takes some thought with an eye toward the use of the aircraft in question.

Some pages back one of our contributors listed a whole bunch of commercial transport aircraft, one or two of which typically has an AoA gage.

Rather that asserting that "pilot won't get any usable info (SA or otherwise) from it," perhaps what's been done is a task analysis, and a scan analysis.

What I think happened was that a choice was made for the expected operating environment, and that choice was that AoA information is closely enough indicated indirectly via other information that it wasn't deemed of sufficient priority to add a display to limited real estate in the instrument panel. It appears that most air transport pilots get along fine without an AoA gage.

AoA is still measured and used by the robot ... it is useful information, isn't it, particularly on approach to a wet runway?

[J-Class]

[QUOTE]Given that AoA is still measured and used by the robot ... it is useful information, isn't it?[\QUOTE]

Surely AoA (and perhaps a GPS speed indication for good measure) would be very helpful in instances where airspeed indicators have gone doolally and the aircraft computers are assaulting with pilots with a bunch of not necessarily consistent error messages and alarms?

All modern aircraft rely on system redundancy, but can any system be deemed truly redundant if it relies on the same components on each of its legs? (I'm imagining that more than one pitot tube iced up). Given the problems of producing 'true' redundancy in a single measurement system, why not admit that visible workarounds should always be available to the pilot?

[Croqueteer]

Power (available) plus attitude (available) equals performance. Or is this too simplistic?

[Lonewolf_50]

Doesn't configuration factor in to that? (Is that implied in (available) in your equation?)

[Yankee Whisky]

Given that AoA is still measured and used by the robot ... it is useful information, isn't it?

Surely AoA (and perhaps a GPS speed indication for good measure) would be very helpful in instances where airspeed indicators have gone doolally and the aircraft computers are assaulting with pilots with a bunch of not necessarily consistent error messages and alarms?

All modern aircraft rely on system redundancy, but can any system be deemed truly redundant if it relies on the same components on each of its legs? (I'm imagining that more than one pitot tube iced up). Given the problems of producing 'true' redundancy in a single measurement system, why not admit that visible workarounds should always be available to the
pilot?

With respect to observations of the sort listed in quotes simillar to the above, I have the following comments;

1. In my flight training some 60 plus years ago, I was tought to use
"needle, ball airspeed" as the absolute minimum to get out of
a sticky situation. On todays panels, I notice more than one attitude,
direction and speed instrument all of which driven from different sources
i.e. main batteries , stand-by battery, electric gyro, air vaccuum, pitot
pressure etc.
There was recurrent training in the use of this. The "ball" function is
also still on panels, so is a magnetic compass etc..

2. I have not seen much, if any, mention of "pitot heat" ,but that should
be one of the first actions of any crew member seeing airspeed
deterioration etc. It would be odd if all sensors were not being heated
at the same time. I burned my fingers (once only !) when checking a
pitot tube for wasp intrusion (one may remember this is very common
at certain times of the year).

3. What I suspect is the the sophistication used in designing an all
electronic/electric airplane caused programmers to assume pilots
are "managers" instead of "stick and rudder" chaps who can fly an
airplane by hand and who would prefer to receive an airplane that can be readily flown by hand, even under some type of adverse conditions, such as
turbulance. Methinks, these compromises in the interst of payloads and speed have left pilots to accept situations where they have precious little leaway of reacting, but, statistically, taking a relatively small risk.After all
we can all use a paycheque from time to time, no?

4. Anyone who read stories (Thud Ridge etc) about fighter pilots, most recently in in Vietnam, hearing multiple warning sounds in their cockpits made the conscious decisions to switch the f'rs OFF! The overload being considered more dangerous than flying out of harm's way and retain situational awareness.


5. GPS was listed as unreliable because only groundspeed is shown. That
is so, but the modern GPS reading can be corrected by a pilot knowing
the windspeed on his flight level, which at around 40,000 feet is fairly
directionally constant. In cu-nim systems this may be of not much use
because of the cloud's influence on its surroundingding air.
In gliding, GPS can measure the wind because when a glider circles
to climb, the drift is measurable and the wind speed become known
giving a read-out of both GPS ground and air speeds.

6. Perhaps certain cultural behaviour with certain flightcrews could mean
that crew members could be/feel intimidated by captains and fail to
throw out the book and do what is needed in an extraordinary situation?
KLM/Tennerife comes to mind.

I cannot judge the reactions, or the lack thereof, displayed by the flight
deck occupants on flight AF447, but if too much confusing information lead them to make errors of oversight/ignorance, then the obvious answer is to simplify the management in these situations, rather than have the computer sensors throw multiple warnings in the air and on the panel ! Perhaps computer programmers should bear this in mind and start refining the systems they concocted. My two cents.......

[Safety Concerns]

As a result of the blocked static ports the basic flight instruments relayed false airspeed, altitude and vertical speed data. Because the failure was not in any of the instruments but rather in a common supporting system, thereby defeating redundancy, the altimeter also relayed the false altitude information to the Air Traffic Controller.

Although the pilots were quite cognizant of the possibility that all of the flight instruments were providing inaccurate data, the correlation between the altitude data given by ATC and that on the altimeter likely further compounded the confusion.

Also contributing to their difficulty were the numerous cockpit alarms that the computer system generated, which conflicted both with each other and with the instruments. This lack of situational awareness can be seen in the CVR transcript. The fact that the flight took place at night and over water thus not giving the pilots any visual references was also a major factor.

sound familiar? That was a Boeing 757.
Now get off the flawed airbus nonsense.

[SaturnV]

Safety concerns, false analog. Maintenance had taped over the static ports, and the 757 took off in that condition. Is Boeing to be faulted because its engineers failed to anticipate the possibility of incompetence and/or gross negligence, and thus did not design and install a static-ports-are-taped-over sensor.

[Jet Jockey A4]

SaturnV wrote...

"Safety concerns, false analog. Maintenance had taped over the static ports, and the 757 took off in that condition. Is Boeing to be faulted because its engineers failed to anticipate the possibility of incompetence and/or gross negligence, and thus did not design and install a static-ports-are-taped-over sensor."

If you are referring to the Birgenair Flight 301 that had taken off Puerto Plata's Gregorio Luperón International Airport in the Dominican Republic and crashed into the ocean, it was not as some say some "tape" that had covered the static port.

Investigators suspected that some kind of insect could have created a nest inside the pitot tube. The prime suspect is a species called the black and yellow mud dauber wasp, well-known by pilots flying in the Dominican Republic.

The aircraft had not flown in 25 days during which time the pitot tubes were not covered, giving the wasps an opportunity to build nests in the tubes.

[Teddy Robinson]

They would if they were trained to.

[edmundronald]

You get what you pay for.

So the answer to too much automation is more automation?

If the idea is to make a plane that thousands of people on every continent can fly, regardless of whether they buy their licenses or earn them, then yes I guess the only way to fix a problem with the automation is to improve the automation. Because the alternative is to improve the pilot's ability to hand fly their way out of trouble, and the training system for *civilian* pilots just isn't set up to do that. While Airbus etc are set up to improve the automation.

And before you dismiss my provocative post - please reflect: Why is it relevant whether the two FO's on that plane were less experienced than the captain? Were those FO's really so bad that they can be improved on by any reasonably priced training or recruitment short of several thousand hours on type? Were they really any worse than an average third world captain?

Edmund

[GlueBall]

Jet Jockey A4

That reference was NOT to Birgenair, but to AeroPeru departing Lima in which case the pitot & static ports were taped when the airframe was washed.

[MountainBear]

but if too much confusing information lead them to make errors of oversight/ignorance, then the obvious answer is to simplify the management in these situations, rather than have the computer sensors throw multiple warnings in the air and on the panel

This is a well understood problem. The answer, however, is not as simple as one might surmise. One major difficulty is that when the **** hits the fan the computer doesn't have any effective way to prioritize errors. One major reason why the pilot gets the situation dumped in his lap is based upon the expectation that, being the person on the spot, he can do what the computer cannot do: prioritize. The electronics which govern the flight systems is not a android or a cyborg; it can't think for itself. All it can do is what it was programmed to do. Well, by definition if the situation is thrown in the pilots lap it means the computer's programming capability has been exceeded. It's an oxymoron to insist that software be programmed for the unexpected.

In practical terms, "simplifying" things for the pilots means more complex software code, more automation, and more guesswork on the part of software engineers. The net result would be to 'simply' the pilot right out of the flight deck entirely.

[JCviggen]

The net result would be to 'simply' the pilot right out of the flight deck entirely.

I don't think that's going to happen any time soon, although a pilot may become less of a "pilot" nonetheless.

Automation is the way forward unfortunately, especially with safety in mind. Statistically the computers are making considerably less mistakes than the pilots who are human, which is to be expected. Obviously the incident rate will never drop to zero entirely, but I think it would be very difficult to defend the notion that planes would be inherently safer with less automation.

There is still room for improvement in automation for sure,and the unfortunate AF447 will eventually lead to improvements in future airplanes. Hopefully it has shown among other things that you still depend on humans with their common sense to sort things out when the machines throw in the towel. So it's in one's best interest to give them as clear a situation as possible, not throw a bazillion error messages onto their screen. And fill in the gaps in training for the situations that cannot possibly happen.

Going back 20+ years in technology is not the answer.

[MountainBear]

but I think it would be very difficult to defend the notion that planes would be inherently safer with less automation.

I don't think that's true. "Automation' is a nice word but behind every piece of machine and every bit of software code is a human being. The issue is often portrayed as "machine vs man" but it's best understood as "human vs human". Aviation engineers, software programmers, etc are as a class no less capable of error than pilots. I think the trend towards automation will continue not because it's safer but because it's cheaper.

People often conceive of risk as a glass from which one drinks until there is nothing left on the bottom but the dregs. That view is wrong. After a certain point risk in complex systems becomes a tug of war. A decrease in risk in one area leads to an increase in risk in another. Often this increase in risk in another area is unforeseen, until it bites you on the ass. The net effect is that "improvements" are more often than not simply rearranging the furniture. We might just be at that point in aviation today.

Obviously the incident rate will never drop to zero entirely

Exactly. Look at the bright side. If the flight deck becomes completely automated that's two less possible deaths avoided right off the bat.

[XX621]

Originally Posted by MountainBear

I think the trend towards automation will continue not because it's safer but because it's cheaper.
.....
The net effect is that "improvements" are more often than not simply rearranging the furniture. We might just be at that point in aviation today.
...

.

Disagree.

The next generation of flight control systems may not just be rigidly programmed, but will contain increasing amounts of machine-learning / pattern-recognition systems - although this technology is very much in it's infancy. There is probably scope in applying such systems to failure workload analysis and prioritisation, providing flight-crews with real-time assistance. For example, providing flight-crews with a situtation analysis and possible root causes with an indication of the degree of recognition between the observed inputs and the trained pattern:

e.g
95% match, given current ADC inputs, of an airspeed probe failure.
62% {some other related pattern}

At the very least such indications would take the flight crew direct from a collection of "low-level" warnings to a small number of possible explanations.

As a software engineer myself, I am amazed we don't see more accidents directly related to FBW; given the inherent complexity in building safety-critical software. The fact we don't is a credit to the specialists involved in these highly complex projects.In terms of reliability and correctness of function the field is light years ahead of other applications of computer programming.


NASA - NASA Dryden Fact Sheet - Intelligent Flight Control System

[Safety Concerns]

the reference was to pilots losing control when surrounded by alarms and unreliable indications.

The fact is, analogue or digital, lose basic readings and most of us are in trouble regardless of the flight control system as highlighted by several analogue accidents.

[arcniz]

So the answer to too much automation is more automation?

"Better automation" is more to the point. Relative to the sophistication of airframes, turbines, and other aircraft electromechanical and aerodynamic systems, the Automation is roughly in the era of the DC-3: clunky and sparse in function and comfort, but moderately tough.

All modern aircraft rely on system redundancy, but can any system be deemed truly redundant if it relies on the same components on each of its legs?

This is really a core issue -- which applies to the entire present global aircraft fleet: reliance on a single method for airspeed sensing, no matter how tried and true, means that single-mode perturbances will be able to cause accidents leading to the loss of aircraft and all aboard.

At present there are at least a dozen methods known that could be used to provide precision airspeed indications (especially in specific situations, such as very high, very fast, very slow) to complement pitot speed data, and/or replace it during icing and similar fault events.

To not have such redundancy in measuring this absolutely critical flight parameter on large passenger aircraft, even with the inevitable increment of cost for the devices and associated logistics, is to disregard a range of undesirable possibilities that probably should not be dismissed as unimportant.

[Safety Concerns]

what rubbish. There are usually 3 completely independent systems and history of operating aircraft for over 100 years would indicate this may be the FIRST time ever ALL 3 pitot systems failed.

I am shocked at he complete lack of reality and perspective here. The A330 global fleet has performed superbly since introduction and has an enviable record. The answer in my opinion is very very simple. Don't go near storms.