AirAsia crash & automation failure

[LockyOK]

I have just finished reading nearly 3,000 posts on the Professional Pilot's Rumour Network (pprune.org) about the crash of an AirAsia Airbus A320 in which 162 lives were lost and am struck by a number of points.

The first and most relevant here is the suggestion that modern airline pilots have become overly reliant on software and automation which will inevitably not be 100% foolproof all of the time.

The A320 is a fly-by-wire aircraft which is VERY dependent on a lot of complex software and hardware. It seems that it is so reliable that pilots may have problems recognising that something may occasionally go wrong and they lack the training and/or practical experience to deal with such a situation.

This does raise the question of whether we are all in danger of becoming overly reliant on systems the complexity of which has surpassed our ability adequately to test?

Three other aspects that are noteworthy are:

1. The site seems to be dominated by American Boeing fanboys who clearly despise the French and are under the illusion that Airbus is an exclusively French manufacturer.
2. The number of "Jobsworths" who appear to fly civil airliners and are scared to take a stand when, as a result of flaky equipment they believe that the aircraft of which they are in command may not be airworthy.
3. I always believed that the computer industry was the trendsetter in terms of TLAs and ETLAs - it seems that the aviation industry wins hands down!

 

[Maccapacca]

I get these trash badly worded emails as well. Just delete.

 

[Destination]

You didn't think to add this to the existing thread?
Even after reading 3000 posts elsewhere, you couldn't be arsed to read the pages we have here?

 

[bitslice]

There is a rule that for every x number of lines of code there will be y bugs (forgot the numbers).

You can mathematically prove that something is bug free, but only if you don't make a mistake in spec'ing what it does and your development budget is infinite x 10

Then you'd also have to have a perfect testing schedule.
Then you'd have to assume that hardware will always behave in the same way, if it doesn't then that might make the software do something unexpected.

There was a military helicopter that crashed some twenty years ago, that was a suspected software bug, never proved though.


I wouldn't worry, flying is still safer than having sex with a drunk woman.

 

[Redgie]

There is a rule that for every x number of lines of code there will be y bugs (forgot the numbers).

Not sure this can really be called a "rule", perhaps an average mind you.

 

[bitslice]

It's a rule for my programs

 

[StriderX]

The pilots are by far the most chaotic element in a planes infrastructure, also ATC being people'd is just as nasty.

 

[cainer]

The pilots are by far the most chaotic element in a planes infrastructure.

would you rather get on a plane and hear (in a stephen hawkings voice) hello, my name is i7 920 overclocked to 4 giga hertz, and il be your captain today
?

 

[Jokester]

would you rather get on a plane and hear (in a stephen hawkings voice) hello, my name is i7 920 overclocked to 4 giga hertz, and il be your captain today
?

To be honest, yes! Most plane crashes are caused by the pilot **** ups, not the machine failing.

 

[StriderX]

would you rather get on a plane and hear (in a stephen hawkings voice) hello, my name is i7 920 overclocked to 4 giga hertz, and il be your captain today
?

Maybe Microsoft will outsource Cortana to such duties.

 

[Redgie]

My girlfriend's family were once advised over the tannoy that due to adverse weather their landing was being done by the autopilot, they've all agreed that it's the smoothest they've ever experienced under any circumstance, so it does make me think I wouldn't mind having a 100% automated flight.

 

[Russinating]

[*]I always believed that the computer industry was the trendsetter in terms of TLAs and ETLAs - it seems that the aviation industry wins hands down![/LIST]

You believed incorrectly!

I had to do a BNUC-S (instead of an RPQ) for SUA VLOS with EUSC before submitting our OM to the CAA for a PAW under the ANO.

 

[MrMoonX]

and this couldnt be posted in the massive thread already on this why?

 

[bitslice]

would you rather get on a plane and hear (in a stephen hawkings voice) hello, my name is i7 920 overclocked to 4 giga hertz, and il be your captain today
?

Clippy
"Hi, I see that you are trying to fly a plane, would you like some help with that ?"

 

[Destination]

My girlfriend's family were once advised over the tannoy that due to adverse weather their landing was being done by the autopilot, they've all agreed that it's the smoothest they've ever experienced under any circumstance, so it does make me think I wouldn't mind having a 100% automated flight.

Easyjet pilot once told me virtually all bad weather landings were done using autoland where the system was available. Not sure how true it is, but he was rather convincing.

 

[LockyOK]

Easyjet pilot once told me virtually all bad weather landings were done using autoland where the system was available. Not sure how true it is, but he was rather convincing.

I believe that the phrase "where the system is available" is pretty important here, how common is it? Have you ever seen any videos of landings at Funchal Airport?

The whole issue of automated systems also becomes increasingly an issue when you consider the modern automobile, e.g. the Google Self-Driving Car, the Volvo XC90, Volkswagen's DSG gearbox and Electronic Stabilisation Control, etc.

 

[Westyfield2]

Just have a read of the AF447 report if you want a good case study.

 

[Tefal]

I have just finished reading nearly 3,000 posts on the Professional Pilot's Rumour Network (pprune.org) about the crash of an AirAsia Airbus A320 in which 162 lives were lost and am struck by a number of points.

The first and most relevant here is the suggestion that modern airline pilots have become overly reliant on software and automation which will inevitably not be 100% foolproof all of the time.

The A320 is a fly-by-wire aircraft which is VERY dependent on a lot of complex software and hardware. It seems that it is so reliable that pilots may have problems recognising that something may occasionally go wrong and they lack the training and/or practical experience to deal with such a situation.

This does raise the question of whether we are all in danger of becoming overly reliant on systems the complexity of which has surpassed our ability adequately to test?

Three other aspects that are noteworthy are:

1. The site seems to be dominated by American Boeing fanboys who clearly despise the French and are under the illusion that Airbus is an exclusively French manufacturer.
2. The number of "Jobsworths" who appear to fly civil airliners and are scared to take a stand when, as a result of flaky equipment they believe that the aircraft of which they are in command may not be airworthy.
3. I always believed that the computer industry was the trendsetter in terms of TLAs and ETLAs - it seems that the aviation industry wins hands down!

not read the details but "failure" is probbaly an over statement.

if the sensors go down/are not effective ie pitot tubes freezing the computer says "i cant control the plane autmomaticaly any more heres ful lcontrol to you the trained pilots".


now normally an airbus will not let a pilot do anything that will crash the plane, you cant stall them you cant roll them etc.

but once it's given full control to the pilot in a sensor failure the pilot can do whatever they like, and we have seen in the past a poorly trained pilot will often slam the thing into the deck though pure incompetence.

its not the automation that fails its the pilot.


it's like taking someone whose only ever driven an automatic and putting them in a manual then blaming the automatic cars gearbox for them crunching your manual gears.

for the actual fly by wire stuff to fail, 3 separate independently designed computers running 3 separate independently designed software's mush all fail.

they use different software on each computer so there cant be a systemic fault across all three.

 

[LockyOK]

Just have a read of the AF447 report if you want a good case study.

LINK?

 

[[TW]Fox]

Rumour

Do you not think the official report might be a better place to start? All we have at the moment is speculation - some of it informed speculation by qualified people but it's still speculation unless they have first hand sight of the evidence which as yet nobody outside of the ongoing investigation does.