Airbus pitches pilotless jets -- at Le Bourget
Join Date: Feb 2001
Location: Europe
Posts: 1,109
Likes: 0
Received 0 Likes
on
0 Posts
Airbus is doing it quietly for more than 10 years now. It is the only possible breakthrough to improve flight safety further. And by the way AF447 case would have been happily resolved on such kind of new aircraft, just applying pre-progammed binary logic.
A terrorists dream perhaps?
Vulnerable to hacking and hence potentially utterly appalling consequences.
A real live person (preferably 2) at the helm when I fly please.
BSD.
Vulnerable to hacking and hence potentially utterly appalling consequences.
A real live person (preferably 2) at the helm when I fly please.
BSD.
Join Date: Feb 2001
Location: Europe
Posts: 1,109
Likes: 0
Received 0 Likes
on
0 Posts
Join Date: Jul 2014
Location: Harbour Master Place
Posts: 662
Likes: 0
Received 0 Likes
on
0 Posts
I'd like to know how a fully automated system would have dealt with the Cathay CX780 fuel contamination with one engine stuck at high thrust and the other at idle? How do you program that scenario? Ok, you don't, you have remote control. So it's not actually fully "automated", you have just moved the human decision maker to a different location. Still completely capable of making a Human Factor screw up. Humans have been covering for, and saving computers in aviation or a long time, the manufacturers may not even be aware of the extent and nature of this problem, see the Therac 25 report below and the operation of the fail safe mechanical interlock in opposition and protecting to the software command lethal dose. Two humans are the fail safe in aviation, we save a lot more than we kill.
The computer accidents history is replete with Human Factor screw-ups, they just occur in the coding cubicle, not the interface with the real world. A close read of the first documented computer accident, the Therac 25 and some of Nancy Leveson's other work on comparing the introduction of Software and the introduction of the High Pressure Steam powered era accidents and how to encourage public confidence should give anyone pause for thought about the future of aviation automation. Her contention is the software is the laggard and it's reliable operation is subject to enormous, drum roll please, Human Factors.
As Leveson sagely notes, the Steam business couldn't move forward until regulations had caught up with the boiler makers technological advancements and those regulations were driven by public outrage at the deaths and maiming caused by poor quality products.
We have already had an inkling of the public's tolerance for accidents in this sphere with the 737MAX - two. Two accidents of a pilotless aircraft and the entire effort will be put in jeopardy. There are enormous risks by actually proceeding toward a commercial product
The computer accidents history is replete with Human Factor screw-ups, they just occur in the coding cubicle, not the interface with the real world. A close read of the first documented computer accident, the Therac 25 and some of Nancy Leveson's other work on comparing the introduction of Software and the introduction of the High Pressure Steam powered era accidents and how to encourage public confidence should give anyone pause for thought about the future of aviation automation. Her contention is the software is the laggard and it's reliable operation is subject to enormous, drum roll please, Human Factors.
As Leveson sagely notes, the Steam business couldn't move forward until regulations had caught up with the boiler makers technological advancements and those regulations were driven by public outrage at the deaths and maiming caused by poor quality products.
Originally Posted by Nancy Leveson
A second reason for the number of accidents was that engineers had badly miscalculated the working environment of steam engines and the quality of the operators and maintainers. Most designs for engines and safety features were based on the assumption that owners and operators would behave rational, conscientiously and capably. But operators and maintainers were poorly trained, and economic incentives existed to override safety features in order to get more work done. Owners had little understanding of the workings of the engine and the limits of it's operation.
We have already had an inkling of the public's tolerance for accidents in this sphere with the 737MAX - two. Two accidents of a pilotless aircraft and the entire effort will be put in jeopardy. There are enormous risks by actually proceeding toward a commercial product
Last edited by CurtainTwitcher; 18th Jun 2019 at 22:14.
Psychophysiological entity
Absolute bollocks
I made a wisecrack years ago about full automation, with a couple of actors up the front . . . and then added, oh, there already is. Don't worry, just based on jealousy.
The reality is, aircraft of the future will set out with a computer driving that knows the details of every tall thing in the world, the weather everywhere, and a full memory, not only of every accident but the post accident analysis. It will be one smart cookie. Two, better make that two smart cookies, plus a spacesaver in a box somewhere.
Join Date: Aug 2017
Location: London
Posts: 94
Likes: 0
Received 0 Likes
on
0 Posts
Not quite. He does have a point. In some ways it's easier to write software to fly an aircraft than to drive a car. There is very little to crash into up in fresh air. There are endless things to crash into on your average street. And that lady with a pram isn't broadcasting ADS-B, so good luck programming your collision avoidance.
Psychophysiological entity
There is very little to crash into up in fresh air.
50 years ago there was the story of a transatlantic flight made with only one human - a press reporter. It was repeated for so long that I began to wonder if there was any truth in it, but I doubt it, they would have been using valves/tubes in that era. (of the flight)
Join Date: Mar 2014
Location: Antarctica
Posts: 123
Likes: 0
Received 0 Likes
on
0 Posts
Join Date: Jan 2019
Location: Wherever I need to be
Posts: 4
Likes: 0
Received 0 Likes
on
0 Posts
Maybe a ground based pilot alongside the single pilot, checking that everything is done correctly?
But then why not just have two pilots in the cockpit.
Psychophysiological entity
Of course, pilotless is just the start. Look how much we could save by going cabin-crewless too!
What is funny is the press leaping on a story like this, even to have a sketch of the folk all lined up holding onto straps.
I've posted variations of this many times - I have little doubt that we'll eventually end up with pilotless commercial aircraft. I also foresee a future where not only are fully autonomous cars common, driving a car yourself will be expressly banned aside from a few areas set aside for dinosaurs like me that actually enjoy a brisk drive through the countryside.
That being said, I also believe we are still decades away from that future - far enough I doubt I'll live to see it.
Airbus said a lot of stuff in that press conference that I have issues with - I wonder if he'd actually talked to his engineers before spouting off about going completely to hydrogen powered aircraft. I know people that have looked at that, and the problems are tremendous - especially where to put the fuel. In order to have a reasonable density of hydrogen, it needs to be liquid, which means really, really cold, and even then then density is so low that it takes a massive volume to fly even a few thousand miles - you're not just going to stick it in the wings... You're talking a massive volume that needs to be kept cryrogenically cold for long periods of time. Materials that can handle that level of cold for extended periods, and reusable thousands of times is also a problem. Not to mention all the inefficiency in using a cryogenic fuel as well as obtaining large quantities of hydrogen in an environmentally friendly manner (currently most hydrogen fuel is created by stripping the hydrogen off of hydrocarbon fuels).
That being said, I also believe we are still decades away from that future - far enough I doubt I'll live to see it.
Airbus said a lot of stuff in that press conference that I have issues with - I wonder if he'd actually talked to his engineers before spouting off about going completely to hydrogen powered aircraft. I know people that have looked at that, and the problems are tremendous - especially where to put the fuel. In order to have a reasonable density of hydrogen, it needs to be liquid, which means really, really cold, and even then then density is so low that it takes a massive volume to fly even a few thousand miles - you're not just going to stick it in the wings... You're talking a massive volume that needs to be kept cryrogenically cold for long periods of time. Materials that can handle that level of cold for extended periods, and reusable thousands of times is also a problem. Not to mention all the inefficiency in using a cryogenic fuel as well as obtaining large quantities of hydrogen in an environmentally friendly manner (currently most hydrogen fuel is created by stripping the hydrogen off of hydrocarbon fuels).
There’s been a few “automation” errors that were only saved by the crew. The confused computers could not have fixed the problem, indeed they caused it.....
So then what? You’ll Have yet another computer monitoring the other computers output?
So then what? You’ll Have yet another computer monitoring the other computers output?
As long as you can feed it with reliable data automation is already working very well and can only improve further. Of course absolute perfection will probably never be achieved but pilotless aircrafts will happen, and I'd wager to say sooner than most expect.
Thread Starter
Join Date: Mar 2019
Location: Canada
Posts: 72
Likes: 0
Received 0 Likes
on
0 Posts
Oh but we do. For trains, Fully Automated Operations (FAO) have been the reality since Kobe New Transit in Japan achieved Grade of Automation 4 -- the highest possible -- way back in 1981.
FAO trains have no drivers at all. Human attendants (if present) are mainly for customer service. Many of these trains operate 24/7 as part of the most complex systems in the world (typically large city metro / commuter light rails) and achieve the absolute highest safety records.
A prime example is the Copenhagen Metro, which was designed to be fully automated from day 1. At peak the Metro carries 12,000 passengers per hour. It was awarded the World's Best Metro for three years in a row (2009-2011).
In fact if you visit Airbus in Toulouse you might notice that all of the metro trains in the city are fully automated. Today FAO GoA 4 trains operate in 40+ cities / 20 countries around the world, including the Sydney Metro Northwest line that just opened last month.
FAO trains have no drivers at all. Human attendants (if present) are mainly for customer service. Many of these trains operate 24/7 as part of the most complex systems in the world (typically large city metro / commuter light rails) and achieve the absolute highest safety records.
A prime example is the Copenhagen Metro, which was designed to be fully automated from day 1. At peak the Metro carries 12,000 passengers per hour. It was awarded the World's Best Metro for three years in a row (2009-2011).
In fact if you visit Airbus in Toulouse you might notice that all of the metro trains in the city are fully automated. Today FAO GoA 4 trains operate in 40+ cities / 20 countries around the world, including the Sydney Metro Northwest line that just opened last month.
Originally Posted by ATC Watcher
at that time the boats and the trains will likely be automated too
Originally Posted by Herod
Yes, but they are not so dependent on the law of gravity
Originally Posted by Lord Farringdon
You've just described high speed trains carrying hundreds of passenger at a time. Although, if the driver of one of those decided to punch through a red signal, auto braking systems would engage and cant be over ridden. Everyone lives and the driver is carted off by Police. Anyone ever considered why we don't have driverless trains?
Rio Tinto is moving some of the heaviest trains in the world thousands of miles daily, with no driver's within cooee. They are not without their teething problems, but the system appears to be proving itself reliable enough to continue with thus far. If it was that bad or unreliable that it was hurting their bottom line, they would get rid of it in a heartbeat and reintroduce manned trains.
On the flight deck one human, one dog.
Dog bites human if he touches anything
Dog bites human if he touches anything
Join Date: Feb 2019
Location: shiny side up
Posts: 431
Likes: 0
Received 0 Likes
on
0 Posts
Interesting thoughts, but really, think it though...
On a CATIII autoland, what exactly does the crew have to do? drop the gear?
I watch landings all the time, who cant tell when the ac lands on auto vs a pilot?
On DEP, once weight off wheels....again, retract the gear?
There are fully autonomous aircraft, some sizable ones, including helos, fuelers, and armed with ****, that fly sorties 24/7.....
Looking at that AB concept, with the driver in the front baggage area, kicking back, surrounded by screens, I would love to kick back in that space with that view...
We learn to fly in the sim, but cannot possibly fly by large screens in front of us...really, sign me up!
Really, I am telling you, especially with drivers of experience, forget the whole BS of getting to the airport, whatever hours, and the rest of the bull****, sign on remote, and see the same thing, well, even more, and fly it remote...
(why do you think there is an ADSB-In port?)
Look at the positives...
On a CATIII autoland, what exactly does the crew have to do? drop the gear?
I watch landings all the time, who cant tell when the ac lands on auto vs a pilot?
On DEP, once weight off wheels....again, retract the gear?
There are fully autonomous aircraft, some sizable ones, including helos, fuelers, and armed with ****, that fly sorties 24/7.....
Looking at that AB concept, with the driver in the front baggage area, kicking back, surrounded by screens, I would love to kick back in that space with that view...
We learn to fly in the sim, but cannot possibly fly by large screens in front of us...really, sign me up!
Really, I am telling you, especially with drivers of experience, forget the whole BS of getting to the airport, whatever hours, and the rest of the bull****, sign on remote, and see the same thing, well, even more, and fly it remote...
(why do you think there is an ADSB-In port?)
Look at the positives...
Join Date: Oct 2017
Location: Rosterabuseland
Posts: 64
Likes: 0
Received 0 Likes
on
0 Posts
Not a good test. An autonomous aircraft facing the exact same situation might simply return to La Guardia without drama.
Not to say Sully that didn't do a superb job -- he absolutely did -- but computers are much better than humans in such situations. I.e., problems with clear constraints where an immediate solution can be computed. There would be no hesitation, no need to query ATC about possible options, no "can we make it to Teterboro?" back-and-forth, etc., all wasting precious seconds.
Not to say Sully that didn't do a superb job -- he absolutely did -- but computers are much better than humans in such situations. I.e., problems with clear constraints where an immediate solution can be computed. There would be no hesitation, no need to query ATC about possible options, no "can we make it to Teterboro?" back-and-forth, etc., all wasting precious seconds.
Thread Starter
Join Date: Mar 2019
Location: Canada
Posts: 72
Likes: 0
Received 0 Likes
on
0 Posts
An "immediate decision" for a human being is an eternity for a computer. By the time Sully noticed something was amiss, an autonomous system could have considered thousands of scenarios, computed the best course of action (having the highest probability of success), and initiated a safe turn back to La Guardia, all without much drama. This is exactly what computers are good for.
Join Date: Jul 2014
Location: Harbour Master Place
Posts: 662
Likes: 0
Received 0 Likes
on
0 Posts
What exactly are they good at? The leading type of AI, neural nets are exceptional of taking a closed problem and solving it eg AlphaGo and AlphaGo Zero
(AGZ). However, these problems were not solved on the fly, they took extensive computational resources to generate it's own training data in the case of AGZ it was a 40 day process simulating playing itself to generate the dataset.
However, neural nets need extensive clean training data, either from the real world or by simulation. There was no extensive corpus of air returns with double engine damage for A320's unlike the millions of recorded games of Go for Alpha Go training. Even Sully's effort represents a single instance that is effectively useless for future algorithmic training. Billions of simulations would be necessary just to replicate & solve this exact scenario on this day. For self driving cars they actually model intersections and do billions of simulations to generate the training data to enable self driving: Inside Waymo's Secret World for Training Self-Driving Cars.
There are many things that computers can do exceptionally better than humans, but solving novelty is not one of them using the current leading AI technology.
(AGZ). However, these problems were not solved on the fly, they took extensive computational resources to generate it's own training data in the case of AGZ it was a 40 day process simulating playing itself to generate the dataset.
However, neural nets need extensive clean training data, either from the real world or by simulation. There was no extensive corpus of air returns with double engine damage for A320's unlike the millions of recorded games of Go for Alpha Go training. Even Sully's effort represents a single instance that is effectively useless for future algorithmic training. Billions of simulations would be necessary just to replicate & solve this exact scenario on this day. For self driving cars they actually model intersections and do billions of simulations to generate the training data to enable self driving: Inside Waymo's Secret World for Training Self-Driving Cars.
There are many things that computers can do exceptionally better than humans, but solving novelty is not one of them using the current leading AI technology.