Interesting lawsuit: Air Canada vs. Airbus

[KL-666]

On March 29, 2015 the AC flight 624 A320 landed short of the runway at Halifax in a bit of a snowstorm.

Now AC sues Airbus for not telling them to train their pilots flying skills. AC claims that the pilots did exactly as Airbus prescribed: "the flight crew correctly configured the aircraft for landing, including entering the correct flight path angle into the flight computer".

Surely AC must know that pilots can easily follow their track on the ILS indicators on their PFD. When trained to do so, of course.

Implicitly AC accuses Airbus of telling them that such expensive pilot skills are superfluous and that pilots do not need any training besides programming the plane.

This way of selling planes i have suspected Airbus already a long time of. They do not say it in writing, but i am pretty sure they orally use the selling argument: "No pilot training necessary". It comes forward in many accident reports where affected airlines state: "Airbus never told us that training the pilots any flight skills was necessary".

So on the surface AC's lawsuit may look silly. But in reality they are trying to open up a cesspool of Airbus lies. A very interesting test case, which unfortunately can hardly be won.

http://www.ctvnews.ca/business/air-canada-lawsuit-accuses-airbus-of-negligence-in-halifax-crash-landing-1.3347461

Kind regards, Vincent

[jwocky]

So well, people will always chose the cheapest flights and that means, the cheapest planes in the overall calculation. And one could also wonder why AC despite being an airline had allegedly no idea that a philosophy like that can't really work, regardless what the manufacturer tells them. But then, I read, Google is starting a new test series for their driver-less car project "because that is the future and will make everything so much safer". Kobayashi Maru, or what was the name of that test again?

[Octal450]

Why didn't they just see ze ILS?

CAN ZE NOT SEE ZE ILS? WHY NOT?

[HJ1an]

Google's driverless cars are scary. I have to say I have zero confidence in their work.

Tesla's made some sense initially, but then it also encourages drivers falling asleep at the wheel, figuratively speaking. And there are a handful of incidents that show this. This isn't new obviously, this was shown many times during the invention of autopilot in aviation, and may still be happening today.

Now the problem is, it seems like they (the automatic driving companies) are moving to full automation - that is, without drivers in them at all during the test - to force the development of the automation to become usable, because let's face it, putting engineers in the car will 3.cause them to fall asleep 2.risk their lives 1.still have to pay them. So, the logic goes, might as well go full driverless from the development sort of as an 'intensive development'.. a.k.a remove all humans.

I have 3 views on this...

1. since aviation autopilot is more developed than cars, plus it has an advantage maturity, and also in the Z axis. The car automation has neither, and is subject to real-world conditions far harsher and far more random. (Just driving through my town would probably cause a Google car to self destruct.)

2. the skies are still relatively uncrowded, there surely is a bit more margin of space to each other, plus regulations mean that you can't really do anything much anyway, so full automation to stick with the flight ways and traffic control.

3. so then if that is the case, they might as well development small, personal transports (like these that are fully automated, rather than have some random guy falling on top of my head because he or she decided training is for chumps. So then the 'flying car' would be a reality.. it would also be fully automated, probably secure (or not) and human factor would be eliminated (but not computer bugs, hacks or glitches - that's a story for another day) and my view is that THAT is where the automatic vehicle is - flying cars. (and maybe some transportation trucks (24/7 highway) or forklifts(warehouse) and things like that ) So these things will fall on your head because of computer glitches or whatnot, but not pilot, and will probably have to be extremely developed before is is certified to fly anyway, so such accidents are smaller chances of happening than when some teenager decided to joyride and crash (which is almost always).

However, not that I fully support automation for flying cars or automatic buses/trucks, but seems like the trend is an eventuality. Back to humans flying airplanes... so.. obviously there is still something wrong with this AC /Airbus picture, but I am always on the stance that pilots usually don't want to die or still somewhat trained to not fly into the ground (exceptions are always there), and the Perfect Storm (or Black Swan?) is always there to get them, usually at the weakest point - being human in that as humans we simply can't take in everything at once.

How much training on top of one another is too much? First you train them to fly. Then you train them to watch instruments. Then you train them to check things. Then train them to watch out for not so fixated on instruments. How to handle passengers (probably). Safety and fire and evacuations etc. Then show them how to fly again. etc. etc. This, on top of possibly asking them to switch planes suddenly and then retrain them, etc. Plus any additional pressure on the already pressured jobs (schedules, fuel, airline policies, new regulations etc). It's no wonder they fall back on automation and then you get the falling asleep thing at the worse times.

[Octal450]

And also flying you have a path to fly, whereas a car/bus you don't have a path. Anything can change you path suddenly.

[KL-666]

How about getting all the trains automated first? That should be relatively easy, as they stick to their tracks. As long as we can not do that, we better not try to automate "loose" objects like aircraft, cars, flying cars, etc...

Kind regards, Vincent

[Octal450]

Yeah exactly.

A computer can not react the way a human can at the moment.

[HJ1an]

How about getting all the trains automated first? That should be relatively easy, as they stick to their tracks. As long as we can not do that, we better not try to automate "loose" objects like aircraft, cars, flying cars, etc...

I thought they pretty much solved this. Since trains are 1 dimensional to the computer. Go. Stop.

In Singapore afaik their transits are fully automatic. In Perth Australia, though, oddly enough they are manned. Even though they seem to be the same make / model of transit systems.

The reason cars can't be fully automated, or at least not so quickly, is that one use in my town and there are inadequately warned roadwork hazards, faded lines, stupid lane merging, inadequate signs, stupid drivers, illegal parkers, random stray animals or even the person crossing and the fast, zig-zagging, wrong-side overtaking motorcyclists thrown into the mix.

Put it in the air, aside from the occasional thunderstorm cloud, high winds, and other air traffic with a separation of x feet, it's pretty much clear space. Make it so that they can't fly or forced to make a landing during sudden storms or predicted-to-be-worse-weather, then it would solve that. Even mountains are solved as long as someone doesn't draw a problematic flightpath for the computer into a mountain, and even so provided its programmed to avoid obstacles its mitigated somewhat.. of course, we all know it's not really THAT idealistic surely. Still, if there are flying cars, I know I probably want it to be fully automated. I don't want Joe Joyride at the controls.

[jwocky]

I am always so weary of this. This entirely wide-eyed believe into technology without even thinking about the extreme scenarios. I mean, we had that a lot ... after dynamite was developed, people thought, they had the use of explosives in mining finally under control. All over the world miners died because it was an illusion. The explosives were easier to handle, alright, but rocks still crumbled under the force of explosions in unexpected ways.

At the end of the 19th and beginning of 20th century, we believed several times in the idea, we had it now all under control with seafare. Of course, we know today, we hadn't. The Titanic was only one disaster that followed. Some fifteen or twenty years earlier, the fire on the Atlantic was born out of the same mindset creating an extreme situation. Many experts think, the disappearance of the Waratah in 1909 could have been such a scenario. So yes, it took a long time till all the rules stemming from the reactions on those disasters were "reworked". We could do so because opposite to those stupids back then, we had the technology now totally under control. Yeah, right ... Derbyshire, Munich, Herald of Free Enterprise, Estonia, Moby Prince, just to name some of the worst disasters that resulted from that kind of thinking.

Here is another good example: Nuclear power plants! There are two extreme safety philosophies, one says, human should have control, don't trust automation. That was the one, that made Chernobyl so bad. And then, there was one that says, computers, you just can't trust humans. That was what made Three-Mile-Island so bad. Maybe someone should wonder why as of yet, not one serious incident, I don't even speak about fully blown accidents, happened in any nuclear plant that has safety based on a symbiosis of well-trained humans with sophisticated technology? And funny thing, the same can be observed in plane disasters. Every time, well-trained humans got all the information needed from technological systems, people get away with a black eye, even in the worst situations. Whenever technology doesn't deliver information, people make wrong decisions and things end in disaster. Whenever people are insufficiently trained to handle the complex technologies around them, things end in disaster. The mindless belief in technology without considering the extreme scenarios brings people in the delusion that learning and training is unnecessary now, the technology does it all for us.

There is not really a thing like a black swan. Well, there are of course black swans as in animals. But the whole idea is nothing else but a very twisted and logically failing excuse for not having thought through all possibilities. Think 9/11, the textbook example for a black swan event. Who could have guessed, terrorists would hijack planes and fly them into high rises? Well, actually a lot of people! Because it was a recurring theme in thriller novels since twenty years before 9/11. And because all that is thought can be in one form or the other also done at some point, actually, the security forces of many countries had plans for attacks like that. Israel for example. So it was not "unthinkable". Only after it happened because nobody did the thinking work ahead of time, there was afterwards a need for a good and well worded excuse. Yeah, it was a "black swan", something, nobody could predict or even think about.

So, shall we really believe, that
a.) this time the technology is under control
and
b.) all eventualities are thought through
????

Here is another question, just food for thought. If the computer in an automated car comes in a situation, where the car will either drive into a group of people or crash frontal in a semi-truck, which will kill the occupants of the car for sure, what are the decision patterns programmed in the software? What will the computer decide? Just ponder a bit about it. Star Trek fans, think Kobayashi Maru here.

[HJ1an]

So, shall we really believe, that

a.) this time the technology is under control
and
b.) all eventualities are thought through
????

Here is another question, just food for thought. If the computer in an automated car comes in a situation, where the car will either drive into a group of people or crash frontal in a semi-truck, which will kill the occupants of the car for sure, what are the decision patterns programmed in the software? What will the computer decide? Just ponder a bit about it. Star Trek fans, think Kobayashi Maru here.

Well, you and I both know that a. & b. isn't even remotely close...

And this is why I say automatic (land) cars won't fly (hahar, I made a pun). The decision making is driven by the computer - who was programmed by a human (at least as of right now). I've always joked that what if the programmer was animal lover and unsocial person and in such a conundrum made the computer avoid animals and plow into a bus full of school children? Or that how would a computer know that it had a flat, assuming the sensor was defective? I certainly won't buy a car where I have to sign a clause where the car will sacrifice me.

They may have their place in sports though, I think it's interesting to race a robot car around a track against each other - but like racing, automatic driving cars should be off limits to public roads.

However as personal flying taxis become a reality and it's a less regular thing (like taxis right now) then it is probably a bit more acceptable to take an certified automatic flying machine (on a good day) that had its software derived from aviation industries. Not that I would jump on the first one, I will let them work out the bugs first.