Old, but interesting topic

On Mar 31, 4:45 am, Mxsmanic wrote:
Sammy writes:
Calibration off by a small tolerance will do that.
How much of a tolerance, exactly?

You tell me. You're making the claim that it's good enough. If you're
wrong and people were to do what you say people would die. If I'm
wrong they're inconvenienced. So the burden of proof is on you.

Great enough that it's considered an unacceptable risk to auto land
unless the equipment is certified.
How great is that, exactly?

You tell me. You're making the claim that it's good enough. Show me
proof that it's an acceptable risk. You might think you know better
than your local air safety regulators but you don't.

I'm sorry. The A380 has landed in Sydney airport. That landing
required the runway to be lengthened. I was unaware that it was a
fictitious aircraft and that everyone is hallucinating it.
It is not in regular service yet, nor has it been delivered in quantity.

So what you're saying is that you don't believe it will ever be in
service or delivered in any quantity? Quite a bold claim you're
making. Care to tell me based on what evidence/knowledge you're making
that claim?

That's rubbish. You said the FMC is programmed from the start
(Incorrect, the approach is usually input en-route, but never mind).
The approach can be entered into the FMC at any time. It is also optional.

....and since the approach changes due to conditions at the airport
including weather and traffic, there is no reason for a pilot to
program in an approach so far in advance.

Now we know that planes do use these notorious approaches.
But notorious appraoches are rare, and it's unlikely one would be programmed
in the scenario under discussion. If it were, and if this risked being a
problem, another one could replace it.

As I've said there's no reason to believe any approach would have been
programmed in yet. Regardless you can just choose another approach.
Unfortunately that means programming it in, which completely
contradicts your little scenario that the entire flight to autoland
has been laid into the computer before takeoff. That just doesn't
happen.

Yes but it only takes one to kill everyone on board.
One of that gravity is statistically unlikely.
Absolute rubbish. Take a look at a few air crash reports some time.
The crash reports logically address only the cases in which it was enough to
kill everyone on board. The other cases in which it was not do not cause
crashes, and thus do not generate crash reports.

Yes they get filed as incidents. I seem to remember at least one well
documented case of takeoff thrust being calculated incorrectly and
causing a plane (a 747 freighter I think) to crash because the pilot
neglected to double check the weight entered into the FMC. It does
only take one mistake, and it has happened. That's what makes check
lists critical.

They still fail, particularly when you have many thousands of flights
every day and an aging fleet of aircraft.
They rarely fail.

Rarely is enough when you're talking about tens thousands of flights a
day. That's an excellent argument for giving those pilots all that
extra training you seem to think is unnecessary.

Landing isn't usually automated.
But it can be, in the scenario under discussion. It would be safer than a
landing flown by hand.

Possibly but it wouldn't have been programmed in from before takeoff.
There's no reason to believe that it would have. Therefore your
inexperienced non-pilot has to learn to follow instructions to enter
the autolanding into the automated systems. Over a radio link there's
a good chance a mistake could be made.

There is no reason to assume that an autoland is programmed in
if a pilot becomes incapacitated.
Since that cannot be done in advance, it certainly would not be programmed in,
but a few presses of a few buttons would fix that.

How many presses and what buttons? Does your PMDG addon simulate
autolanding? How accurately does it claim to? How many button presses
are required to select a STAR and enter in an autolanding? What
assumptions must be made about what's already been entered?

Most? That's a statistical statement.
Yes. Good.

Want to provide me with stats on
what proportion of declared emergencies end in loss of life for large
aircraft? Any idea at all what the number is?
There were 91 accidents investigated by the NTSB in February. Of these, 23
involved a loss of life. Only a very small fraction of declared emergencies
lead to an accident that is investigated by the NTSB, but essentially all
accidents resulting in loss of life are investigated. Therefore the
percentage of declared emergencies that result in loss of life is very small
indeed.

Wow your grasp of stats is even worse than your grasp of aviation.
Incidents in February may not have been fully investigated. One
month's data from one country isn't a large enough sample size to draw
any conclusion. Your data must include a lot of GA aircraft where you
keep harassing me to stick to a very specific scenario dealing only
with a large automated jet.

Your statement doesn't even attempt to answer the question, but
instead answers another. I asked how many declared emergencies ended
in loss of life (since you claim most don't end in anything
catastrophic). Instead of telling me you give me some statistic about
how few fatalities were caused when an emergency had been declared
which. This is a completely irrelevant statistic to our discussion. We
were not discussing anything about fatalities that occurred when no
emergency had been declared. Do you know anything about set theory?
You're dealing with two sets of data with the only overlap being
emergencies declared that did end in loss of life. This is typical of
your style of argument. Complete misdirection, and failure to actually
address questions or counter claims in any meaningful way.

And when you consider that none of the fatalities occurred on airliners, the
percentage diminishes by at least another order of magnitude.

An extrapolation based on a small data set not even relevant to the
discussion. Again typical of your severely flawed logic.

Look you hold a bunch of truly bizzare opinions and once stated try to
tell people you've provided facts. They're not facts they're
unsubstantiated supposition.
Only if you consider the NTSB reports to be unsubstantiated.

Lets look at the facts shall we? You've taken a small incomplete (too
new to be complete) data set, presented a flawed argument that fails
to address the issue, and tried to force that into fitting your
argument because there is a lack of data. Your analysis is rubbish. If
you do this for a living in any capacity I wonder how many jobs you've
been sacked from.


Your world view does not tally with the
majority of accepted evidence, or with what experts report.
I got it from the experts.

No, you've demonstrated that you can unconvincingly twist incomplete
data in a flawed an indefensible way to support a nonsense argument.
Pathetic really.

If you want me to take you at all seriously you need to provide solid
reference material.
I don't care if you take me seriously or not. And others can look up the data
for themselves.

Yes they can. Others will probably be more sensible about drawing
conclusions from it that you will. They'll probably take a more
complete less recent data set and actually make an analysis that's
logical and consistent. Something you seem incapable of.