Jeremy Vine Show - Pilotless Airliners
Join Date: Feb 2001
Location: The Winchester
Posts: 6,375
Likes: 0
Received 0 Likes
on
0 Posts
Tourist
I don't want to get into a spat but I did read what you wrote.
The question posed by Neville was:
"How does autonomous aircraft fly through thunderstorms?"
The passengers (and possibly Neville) probably don't want to know that the picture is synthetic, that it is clever, or that computers are already in the loop. What would be more relevant is how the left/right/through/turn back decision is going to be made, bearing in mind that there may be non-meteorological restrictions to your course of action.
OK, Badly worded, mea culpa, I was merely trying to ask/confirm how much weather avoidance experience you have had using modern weather radars in commercial ops, and if you thought they were up for the job of "advising" automated system downstream.
I'll reiterate what I have said before, I think you're right in claiming fully autonomous ops will happen, but from what I have heard (for the likes of my eldest who is working on an aspect of AI at a post grad level and others who have worked in the field for years) the various technologies involved are nowhere near enough mature or robust enough to support it in routine commercial ops for the foreseeable future, i.e. several decades.
I don't want to get into a spat but I did read what you wrote.
The question posed by Neville was:
"How does autonomous aircraft fly through thunderstorms?"
The passengers (and possibly Neville) probably don't want to know that the picture is synthetic, that it is clever, or that computers are already in the loop. What would be more relevant is how the left/right/through/turn back decision is going to be made, bearing in mind that there may be non-meteorological restrictions to your course of action.
No need to be snide about my airline time.
I'll reiterate what I have said before, I think you're right in claiming fully autonomous ops will happen, but from what I have heard (for the likes of my eldest who is working on an aspect of AI at a post grad level and others who have worked in the field for years) the various technologies involved are nowhere near enough mature or robust enough to support it in routine commercial ops for the foreseeable future, i.e. several decades.
Join Date: Dec 2006
Location: Florida and wherever my laptop is
Posts: 1,350
Likes: 0
Received 0 Likes
on
0 Posts
Neville-Nobody
Glad you are not saying it cannot be done - because it has been done. There are multitudes of autonomous aircraft, indeed almost all UAS are _required_ to be autonomous so that they can safely recover in case of a command link failure. Even the little octocopter toys will go back whence they came when they sense low battery.
Automating ATC?
Yes that is being done too - there are many simulators that will 'control' multiple simulated aircraft. Indeed they will happily run the entire FABEC or New York Metroplex at a few hundred times normal speed with 50% more traffic.
But you should always use the adage: Just because you can, doesn't mean you should.
The easiest to automate is the aircraft, mainly because much is already automated. The adherence to SOP to the extent that what used to be finely judged decisions are now 'no brainer' SOP box ticking makes automation even easier.
Air Traffic Management is a larger problem and while flying aircraft around a scripted pattern can work well, each 'sector' interfaces with other sectors, the multiple sectors form groups with centres/centers and interface with multiple airports. The Air Navigation Service Providers or Functional Airspace Blocks interface with others and decisions made in one area can rapidly impact others. So while in small areas ATC can be run by simulators it rapidly grows into an Np problem as more areas are added and the cross impact of multiple trajectories are taken into account. That does not prevent several areas of ATM being given 'Decision Support' tools especially in conflict detection and resolution.
Importantly, the new ATM systems use the aircraft intelligence so that the aircraft now provides earliest time/latest time of crossing of waypoints or arrival, and the trajectory description used by the ATM systems will in the future be the one generated by the FMC not one generated on the ground. As more airspace becomes 'free route' that is everyone flying their business or mission trajectory the airspace will become more efficient for the aircraft and the airspace utilization will increase.
So there will be a lot of automation, I expect UAS to be just other aircraft in the airspace and from the system point of view there will be no difference in handling between piloted, remotely piloted and autonomous. It is a commercial/safety decision not a technical decision whether the Airbus 390 or the Boeing 797 will be optionally manned or not. But there will be unmanned autonomous aircraft flying in normal class A airspace. There already are to some extent.
For the record I'm not saying it can't be done I'm saying that the staus quo with some better techonology is probably the safest and most cost effective for now. I don't think full autonomy in pax aircraft will be happening anytime before 2090
Automating ATC?
Yes that is being done too - there are many simulators that will 'control' multiple simulated aircraft. Indeed they will happily run the entire FABEC or New York Metroplex at a few hundred times normal speed with 50% more traffic.
But you should always use the adage: Just because you can, doesn't mean you should.
The easiest to automate is the aircraft, mainly because much is already automated. The adherence to SOP to the extent that what used to be finely judged decisions are now 'no brainer' SOP box ticking makes automation even easier.
Air Traffic Management is a larger problem and while flying aircraft around a scripted pattern can work well, each 'sector' interfaces with other sectors, the multiple sectors form groups with centres/centers and interface with multiple airports. The Air Navigation Service Providers or Functional Airspace Blocks interface with others and decisions made in one area can rapidly impact others. So while in small areas ATC can be run by simulators it rapidly grows into an Np problem as more areas are added and the cross impact of multiple trajectories are taken into account. That does not prevent several areas of ATM being given 'Decision Support' tools especially in conflict detection and resolution.
Importantly, the new ATM systems use the aircraft intelligence so that the aircraft now provides earliest time/latest time of crossing of waypoints or arrival, and the trajectory description used by the ATM systems will in the future be the one generated by the FMC not one generated on the ground. As more airspace becomes 'free route' that is everyone flying their business or mission trajectory the airspace will become more efficient for the aircraft and the airspace utilization will increase.
So there will be a lot of automation, I expect UAS to be just other aircraft in the airspace and from the system point of view there will be no difference in handling between piloted, remotely piloted and autonomous. It is a commercial/safety decision not a technical decision whether the Airbus 390 or the Boeing 797 will be optionally manned or not. But there will be unmanned autonomous aircraft flying in normal class A airspace. There already are to some extent.
Join Date: Jul 2003
Location: Somewhere
Posts: 2,886
Likes: 0
Received 0 Likes
on
0 Posts
How does autonomous aircraft fly through thunderstorms?
Current weather radar pictures are synthetic in most modern airliners. This means that the computer has already looked at the raw data, analysed the information and presented what it believes to be useful info about it to the human pilot.
Now, we may well have opinions about how well it does this, but either way, the computer is already in the loop.
Current weather radar pictures are synthetic in most modern airliners. This means that the computer has already looked at the raw data, analysed the information and presented what it believes to be useful info about it to the human pilot.
Now, we may well have opinions about how well it does this, but either way, the computer is already in the loop.
How do autonomous aircraft land in gusty 40knot crosswinds?
How do autonomous aircraft land with shifting winds?
As previously discussed and agreed by John Farley, the answer is better than humans since the Comet.
How do autonomous aircraft land with shifting winds?
As previously discussed and agreed by John Farley, the answer is better than humans since the Comet.
Shifting winds on final is something else that current aircraft don't do well. Especially tailwind to landing in headwind scenario.
How do autonomous aircraft handle data input failures or data corruption?
(ie airspeed indications are no longer reliable or false sensory inputs ie stall warning goes off incorrectly)
Modern airliners are already totally dependent on software to fly. The computers are between the pilot and the controls whether they are autonomous or not. If data being corrupted is a problem, then the problem is already with us.
Added to this, it has unfortunately been shown that human pilots don't necessarily deal with such a scenario anyway.
(ie airspeed indications are no longer reliable or false sensory inputs ie stall warning goes off incorrectly)
Modern airliners are already totally dependent on software to fly. The computers are between the pilot and the controls whether they are autonomous or not. If data being corrupted is a problem, then the problem is already with us.
Added to this, it has unfortunately been shown that human pilots don't necessarily deal with such a scenario anyway.
So how will an autonomous aircraft figure out what is erroneous input and what is real? Computers are ultimately only able to compute what infomation they are given. No matter how advanced the old GIGO analogy still applies.
I guess similarly how will a computer know when another goes rogue? My assumption here is that two independant computers will be running the show but how will they know what is real and what isn't?
This doesn't even consider the extreme failures over the past 30 yeas where humans have really saved the day.
I would imagine most of the above problems could be solved however the data corruption issue could be a bit of a achilles heal.
Last edited by neville_nobody; 25th Aug 2016 at 16:35.
Join Date: Feb 2007
Location: Central Scotland
Posts: 60
Likes: 0
Received 0 Likes
on
0 Posts
Try and write a convincing safety case for it.
Automation is fine....as an assistance; fantastic for work load reduction from mundane tasks, and for freeing mental capacity.
Unfortunately..... no mechanical /software systems can achive the sort of reliability figures required .
The best software had an error rate of approx 1-2 per million lines of code, and as for the systems integrated into aircraft,... well on the aircraft I fly, when the weather gets too rough, the autopilot disconnects itself. ("AUTOPILOT, AUTOPILOT...)
Unfortunately..... no mechanical /software systems can achive the sort of reliability figures required .
The best software had an error rate of approx 1-2 per million lines of code, and as for the systems integrated into aircraft,... well on the aircraft I fly, when the weather gets too rough, the autopilot disconnects itself. ("AUTOPILOT, AUTOPILOT...)
Join Date: Jan 2001
Location: Home
Posts: 3,399
Likes: 0
Received 0 Likes
on
0 Posts
FlyingApe
Does the fact that there are autonomous aircraft currently flying that achieve vastly more tricky tasks than that required of an airliner not cause you to doubt your opinion even slightly?
No, they are not carrying passengers, but that is not a technical difficulty.
Out of interest, the aircraft you fly, what decade of last century was it designed in?
I only ask, because I used to have a car that the ABS never worked on. This was because it was designed before ABS was invented.
Perhaps something similar with your aircraft and autonomous flight?
Does the fact that there are autonomous aircraft currently flying that achieve vastly more tricky tasks than that required of an airliner not cause you to doubt your opinion even slightly?
No, they are not carrying passengers, but that is not a technical difficulty.
Out of interest, the aircraft you fly, what decade of last century was it designed in?
I only ask, because I used to have a car that the ABS never worked on. This was because it was designed before ABS was invented.
Perhaps something similar with your aircraft and autonomous flight?
Join Date: Feb 2001
Location: The Winchester
Posts: 6,375
Likes: 0
Received 0 Likes
on
0 Posts
There are autonomous aircraft currently flying that achieve vastly more tricky tasks than that required of an airliner not cause you to doubt your opinion even slightly?
No, they are not carrying passengers, but that is not a technical difficulty.
No, they are not carrying passengers, but that is not a technical difficulty.
Join Date: Jan 2001
Location: Home
Posts: 3,399
Likes: 0
Received 0 Likes
on
0 Posts
Originally Posted by wiggy
Dare I ask - are you claiming/asserting that the technology to allow autonomous passenger ops working to the required level of safety expected by the traveling public and regulators is here right now? Today?
How can you prove a level of safety except with hindsight?
The US navy currently has an autonomous aircraft which all by itself does carrier launches and landings and in-flight refuelling amongst other things.
https://en.wikipedia.org/wiki/Northrop_Grumman_X-47B
It currently has 100% safety record.....
Now, this was never designed to be passenger carrying. If it was, I would guess they would add an extra engine and probably just as many redundant systems as an airliner.
The point is, that the tech is there, and redundancy can be added in as required.
Join Date: Feb 2001
Location: The Winchester
Posts: 6,375
Likes: 0
Received 0 Likes
on
0 Posts
I have no wish to continually argue but "disengenous" is a bit rich IMHO...
You did state "there are autonomous aircraft currently flying that achieve vastly more tricky tasks than that required of an airliner", and then went on to state "they are not carrying passengers, but that is not a technical difficulty."
Now maybe it's just me but TBH I'd say the question I asked was a fair one in the context of a thread about pilotless airliners.
You did state "there are autonomous aircraft currently flying that achieve vastly more tricky tasks than that required of an airliner", and then went on to state "they are not carrying passengers, but that is not a technical difficulty."
Now maybe it's just me but TBH I'd say the question I asked was a fair one in the context of a thread about pilotless airliners.
Join Date: Jan 2001
Location: Home
Posts: 3,399
Likes: 0
Received 0 Likes
on
0 Posts
You asked a question which as you well knew had no answer since there are no regulatory standards yet.
In terms of technological rather than regulatory challenges, there are autonomous aircraft flying today that are achieving vastly more tricky aspects of flight than airliner ops.
If you can successfully and consistently land on an aircraft carrier, then a runway is zero challenge.
If you can air-to-air refuel, then most other challenges pale into relative insignificance.
As I see it the tricky part will be the handover to an automated ATC, but it is not beyond the wit of man.
If you rephrase the question to exclude the unachievable bit re non-existent regulations, then yes I absolutely claim that the tech required already exists. There are no new technologies required, merely the will to place them all in a suitable airframe.
In terms of technological rather than regulatory challenges, there are autonomous aircraft flying today that are achieving vastly more tricky aspects of flight than airliner ops.
If you can successfully and consistently land on an aircraft carrier, then a runway is zero challenge.
If you can air-to-air refuel, then most other challenges pale into relative insignificance.
As I see it the tricky part will be the handover to an automated ATC, but it is not beyond the wit of man.
If you rephrase the question to exclude the unachievable bit re non-existent regulations, then yes I absolutely claim that the tech required already exists. There are no new technologies required, merely the will to place them all in a suitable airframe.
Join Date: Feb 2001
Location: The Winchester
Posts: 6,375
Likes: 0
Received 0 Likes
on
0 Posts
Look, I think I get your POV, I really do.
You seem to think airliner ops are trivial. Many here would disagree with you.
You seem to think the technology for fully automated/autonomous ops is pretty much here, but perhaps not quite ready for pax ops just yet, though all it needs is some unspecified technical advance.
If anyone fields an an objection as to why autonomous ops might not be a player at the moment you come up with a vague or trivial fix ( add another engine - really? , redundancy - yes but what/where/at what level), or failing that throw in comments about peoples mobile phones or ABS..which doesn't move the argument anywhere.
I can promise you many others (some working in AI) think we're decades away from autonomous ops.
Anyhow this is just a debate, the answer is at least 20-30 years down the road..I'm out.
You seem to think airliner ops are trivial. Many here would disagree with you.
You seem to think the technology for fully automated/autonomous ops is pretty much here, but perhaps not quite ready for pax ops just yet, though all it needs is some unspecified technical advance.
If anyone fields an an objection as to why autonomous ops might not be a player at the moment you come up with a vague or trivial fix ( add another engine - really? , redundancy - yes but what/where/at what level), or failing that throw in comments about peoples mobile phones or ABS..which doesn't move the argument anywhere.
I can promise you many others (some working in AI) think we're decades away from autonomous ops.
Anyhow this is just a debate, the answer is at least 20-30 years down the road..I'm out.
Join Date: Feb 2002
Location: uk
Posts: 301
Likes: 0
Received 0 Likes
on
0 Posts
What Huck said. And back on track.
What has to be asked, and what I said last time this appeared as a thread:
A: How long will development and certification take?
B: How many years of cost saving will it take to recoup (A)?
C: How long will the oil last?
If A + B > C it's a dead duck.
If it helps let's remember that even BP (surely the definition of fossil fuel optimists) are saying just over 40 years
What has to be asked, and what I said last time this appeared as a thread:
A: How long will development and certification take?
B: How many years of cost saving will it take to recoup (A)?
C: How long will the oil last?
If A + B > C it's a dead duck.
If it helps let's remember that even BP (surely the definition of fossil fuel optimists) are saying just over 40 years
Join Date: Jan 2011
Location: England
Posts: 652
Likes: 0
Received 0 Likes
on
0 Posts
16024
Your post appears to be based upon the premise that running out of oil will cause commercial aviation to stop. This is not a realistic idea, we will be running aircraft on "vegieburger oil" or some such substance by then.
Your post appears to be based upon the premise that running out of oil will cause commercial aviation to stop. This is not a realistic idea, we will be running aircraft on "vegieburger oil" or some such substance by then.
I Have Control
Join Date: May 2004
Location: North-West England
Posts: 49
Likes: 0
Received 0 Likes
on
0 Posts
Plenty of people on this thread are in favour of pilotless flight now or in the future. Not a single one of them is a qualified and experienced and current Airline Transport Pilot, I strongly suspect. Their views are therefore worthless, IMHO.
Until you've gone around late due to windshear/squall for example, and then made an immediate low level turn to avoid a Cb, and then decided upon a practical level-out and a quick required course of action/navigation based upon fuel status, alternate weather, other diverting air traffic volume, and the unforecast weather, at destination, then you know squat. (I was part of a 2-crew op that faced this precise scenario at CUN last year, fortunately in a serviceable highly automated aircraft. What followed successfully was in no way a series of automatic reflexes, way off what could ever be programmed in the next hundred years)
Ignorance is bliss for so many of you armchair pilots. Enjoy.
Until you've gone around late due to windshear/squall for example, and then made an immediate low level turn to avoid a Cb, and then decided upon a practical level-out and a quick required course of action/navigation based upon fuel status, alternate weather, other diverting air traffic volume, and the unforecast weather, at destination, then you know squat. (I was part of a 2-crew op that faced this precise scenario at CUN last year, fortunately in a serviceable highly automated aircraft. What followed successfully was in no way a series of automatic reflexes, way off what could ever be programmed in the next hundred years)
Ignorance is bliss for so many of you armchair pilots. Enjoy.
Thread Starter
Join Date: Nov 2007
Location: Brexitland
Posts: 1,141
Likes: 0
Received 0 Likes
on
0 Posts
Exactly what Roy Hudd said.
And John Farley - ie there is a huge difference between "steering" and "operating".
And John Farley - ie there is a huge difference between "steering" and "operating".
Last edited by Arfur Dent; 27th Aug 2016 at 21:43. Reason: sp
Join Date: Jan 2011
Location: England
Posts: 652
Likes: 0
Received 0 Likes
on
0 Posts
RoyHudd
I have not read all of the posts in this thread, and with 8 pages I do not intend to do so. But from what I have read, it is not so much a case of people being in favour of pilotless aircraft or against them. One side appears to be arguing that they will eventually happen, and the other side is arguing that they will not. A number of posters also appear be accept that they will happen one day, but not for several decades.
Sadly like most PPRuNe threads, it will achieve nothing much. Each side will never convince the other, and debate will end only when the two sides become exhausted. None of this will have influence on what the future will actually bring.
I have not read all of the posts in this thread, and with 8 pages I do not intend to do so. But from what I have read, it is not so much a case of people being in favour of pilotless aircraft or against them. One side appears to be arguing that they will eventually happen, and the other side is arguing that they will not. A number of posters also appear be accept that they will happen one day, but not for several decades.
Sadly like most PPRuNe threads, it will achieve nothing much. Each side will never convince the other, and debate will end only when the two sides become exhausted. None of this will have influence on what the future will actually bring.
Guest
Posts: n/a
Originally Posted by rideforever
The big change will be when not only the pilots but the passengers too are robots.
Flight fantastic | The Economist
Join Date: Apr 2006
Location: uk
Posts: 416
Likes: 0
Received 0 Likes
on
0 Posts
The idea that there can be a sensible coloration between current UAVs and future Pilotless airliners is a little weak.
UAVs are not pilotless per se, they are remotely piloted. In addition carrying only hardware rather than "software" people, the loss of a UAV is only monetary.
At the moment there is no such thing as 100% safe software, indeed all our FBW aircraft have redundancy in spades but still need the Mk1 eyeball and Human interface in some situations.
When the Military go pilotless for combat aircraft the civil world will follow probably a decade later. IMHO.
UAVs are not pilotless per se, they are remotely piloted. In addition carrying only hardware rather than "software" people, the loss of a UAV is only monetary.
At the moment there is no such thing as 100% safe software, indeed all our FBW aircraft have redundancy in spades but still need the Mk1 eyeball and Human interface in some situations.
When the Military go pilotless for combat aircraft the civil world will follow probably a decade later. IMHO.
