Human pilot's physical senses are far superior to computer programs!

[wsherriff]

Properly trained and experienced pilots with their natural physical sense inputs are completely capable of safe control of their aircraft! EXCEPT

EXCEPT When the computer programmer who is ignorant of the fact that a swept wing aircraft design will pitch-up in a strong up-draft!! With an increase in AOA! The program recognizes this as a possible stall and shoves the nose over into a steep dive attitude! e.g. Quantas in AU, (Two pitch overs),The Buffalo accident, the Marine F-18 into the house at Mira Mar Air Station and AF 447, are all tied to the same, "Flight envelope protection program"!

How many pilots voted for "side sticks" and "protection programs?

Watch out for those supposed "Experts"!

[ClippedCub]

With all due respect, there seems to be a misunderstanding that a computer programmer determines the behavior of software, particularly FBW and stall warning/pusher software. This couldn't be further from the truth. The engineers and test pilots determine the characteristics and transfer that desire in the form of a Software Specification to the software group. Then as the software is developed, it is checked, and rechecked, and rechecked... by both company test pilots and the engineers.

The Buffalo Dash-8 is a straight wing turboprop. The pusher bought itself on the airplane to satisfy FAA stall characeristics for aft cg, and I believe flaps down, power on. The pusher didn't nose the airplane into the ground, the pilot effectively did by exacerbating a stall due to poor training. The co-pilot also didn't recognise stall, again due to poor training.

Thought the F-18 lost the second engine on approach. Anybody have an accident report.

[CarltonBrowne the FO]

I think you're confusing two different functions, flight envelope protection and stick push.
A320 series flight protection does not pitch rapidly down when it detects an approaching stall- it maintains the AoA for maximum lift as long as I command it, and (except in the final phases of the landing sequence) if a stall is imminent it applies TOGA thrust to try and go up.
Stick push, on the other hand, does force a pitch down to prevent a stall- but stick push is not limited to fly-by-wire aeroplanes; in fact conventional stick push is not a feature of fly-by-wire systems, as the flight envelope protection prevents the stall being reached.

[Will Fraser]

Bear with me. The flight computer is not essentially different than an experienced pilot. What makes a computer 'Better'?? Speed. The human brain has sensors, and memory. A stimulus enters an initial 'port' to ascertain its importance. It then is shuffled in several directions to a more complicated and comprehensive 'memory'. The decision is made based on a rapid 'comparison' of stored data. If it is recognized in the bank, the decision is rapid. A 'new' stimulus involves a more complex computation, takes longer, and has more 'risk' in its solution.

Computers have chips with distances measured in microns, or nanometers between important terminals. Smaller is faster, the Brain may have dozens of feet or even yards involved in neurons and synapses. What makes the computer fast is what makes it so useful in cruise, or on landing in stink.

What makes the pilot occasionally superior is his ability to discern buckwheat from b***sh*t, and an ability to hesitate or 'innovate'

Will, not a computer

[lomapaseo]

Good subject for the Technical forums

my simplistic understanding is that computers work within a law, input predicts an output

man does not

[Will Fraser]

lomapaseo

'input predicts an output'. In other words, some human somewhere knows what the computer will spit out, given his knowledge of what went in.

Your explanation is the best in the fewest words I've seen.

[FrequentSLF]

'input predicts an output'.

Which in most of the computer failures shall be read

"garbage in garbage out"

Humans have the capability to evaluate if garbage in going in and eventually disregard it, even if they do not have sufficient backup systems.
A computer algorithm cannot do that, that's why AP and AT disconnect, because humans are much better in analyzing a situation when there are not sufficient reliable informations, and have a successful outcome.
This implies to be proper trained and alert, otherwise the outcome will be similar to the one of TK.

FSLF

[ZFT]

input predicts an output'. In other words, some human somewhere knows what the computer will spit out, given his knowledge of what went in.

Your explanation is the best in the fewest words I've seen.

Sorry, this is simply not the case. Ignoring aviation for a moment. An example. It took a computer a long, long time to find and verify the highest Prime Number (known to date). Your analogy indicates someone, somewhere already knew the answer.

[Will Fraser]

If you permit a drift into Lie Algebra, I would say you are wrong. In Lie, the answer predicts the solution. E8.

[vapilot2004]

Despite their blinding processing speeds and up to terabyte levels of programme and data memories, computers are profoundly stupid and require instruction on how to perform every task required of them.

Recipe for software:
Obtain computer(s).
Add code.
Run and debug.
Repeat if necessary.

;)