Airbus + Cathay working on Single Pilot during Cruise with A350
I have no doubt that fully autonomous passenger aircraft will be a reality at some time in the future. However, anything involving AI is going to need a fundamental re-think of all the certification processes as we will be replacing systems that have been engineered, tested and sometimes formally proven, with inscrutable black boxes. The “any day now” self-driving car of the last decade shows how non-trivial these kind of problems are, and that’s in something that can just stop if it can’t work out what’s going on and doesn’t have the huge potential for collateral damage that an airliner has...
Join Date: May 1999
Location: Runcorn,Cheshire,England
Posts: 165
Likes: 0
Received 0 Likes
on
0 Posts
FlightDetent
on my last flight on a two week old 321NEO (possibly one of the modest modern tech aircraft in the sky) my colleague picked up due to SOPs that I had forgotten to set QNH on the PFD. We were about to conduct an RNAV approach so if this hadn’t been picked up, the aircraft won’t auto set QNH, and consequently all the constraints on the STAR and the final approach path would have been dangerously below profile. Now granted, you could programme an auto GPWS pull up that an autonomous aircraft would be able to perform but how would it set the QNH or even know the QNH?
tech will be very good at mitigating but not at thinking ahead and making decisions that will avoid the need to mitigate.
on my last flight on a two week old 321NEO (possibly one of the modest modern tech aircraft in the sky) my colleague picked up due to SOPs that I had forgotten to set QNH on the PFD. We were about to conduct an RNAV approach so if this hadn’t been picked up, the aircraft won’t auto set QNH, and consequently all the constraints on the STAR and the final approach path would have been dangerously below profile. Now granted, you could programme an auto GPWS pull up that an autonomous aircraft would be able to perform but how would it set the QNH or even know the QNH?
tech will be very good at mitigating but not at thinking ahead and making decisions that will avoid the need to mitigate.
In your brave new world of autonomous aircraft with no pilots on board who is going to take on the responsibilities currently demanded of the PIC?
Australia CAR extract
(2) A pilot in command of an aircraft is responsible for:
(a) the start, continuation, diversion and end of a flight by the aircraft; and
(b) the operation and safety of the aircraft during flight time; and
(c) the safety of persons and cargo carried on the aircraft; and
(d) the conduct and safety of members of the crew on the aircraft.
(3) The pilot in command shall have final authority as to the disposition of the aircraft while he or she is in command and for the maintenance of discipline by all persons on board.
Australia CAR extract
(2) A pilot in command of an aircraft is responsible for:
(a) the start, continuation, diversion and end of a flight by the aircraft; and
(b) the operation and safety of the aircraft during flight time; and
(c) the safety of persons and cargo carried on the aircraft; and
(d) the conduct and safety of members of the crew on the aircraft.
(3) The pilot in command shall have final authority as to the disposition of the aircraft while he or she is in command and for the maintenance of discipline by all persons on board.
Join Date: Jan 2008
Location: Netherlands
Age: 45
Posts: 287
Likes: 0
Received 0 Likes
on
0 Posts
Where does it say the PIC should be on board and cannot be PIC of multiple aircraft at the same time?
Also, d is not applicable with no one on board, so it will definitely work with cargo only operations.
Also, d is not applicable with no one on board, so it will definitely work with cargo only operations.
Just seems to me that in the history of commercial airliner flying since Ernest Gann's "Fate is the Hunter" days, and after analysis of scores of accidents; the authorities, airlines and manufacturers gradually worked out how to fly safely.
But in recent years accidents - involving standard manoeuvres - seem to be on the increase because airlines appear to be reducing that safety layer by reducing pilots' and engineers' rest, training and testing. And some manufacturers have produced compromised aircraft modifications - all in the pursuit of ever cheaper passenger seat costs. As I say, who really benefits?
And now we are talking about single-pilot operations. We seem to be going backwards, safety-wise.
But in recent years accidents - involving standard manoeuvres - seem to be on the increase because airlines appear to be reducing that safety layer by reducing pilots' and engineers' rest, training and testing. And some manufacturers have produced compromised aircraft modifications - all in the pursuit of ever cheaper passenger seat costs. As I say, who really benefits?
And now we are talking about single-pilot operations. We seem to be going backwards, safety-wise.
I am amused by the illogical assertion that a two week old 321 is “modern” having first flown as the 320 in what… 1988? That’s what, 33 years ago? Designed and built for a two human crew environment.
It’s almost like saying a freshly rebuilt warbird is state of the art fighter tech because it was finished and flown first this year.
It’s almost like saying a freshly rebuilt warbird is state of the art fighter tech because it was finished and flown first this year.
Join Date: May 1999
Location: Runcorn,Cheshire,England
Posts: 165
Likes: 0
Received 0 Likes
on
0 Posts
BoeingDriver99
I guess the change from
the Original 320 to a 321 NEO is lost on you then hey. They are vastly differant
procede
mode s transmits selected Altitude. It does not transit the pressure setting set on the PFDs by the pilots.
I guess the change from
the Original 320 to a 321 NEO is lost on you then hey. They are vastly differant
procede
mode s transmits selected Altitude. It does not transit the pressure setting set on the PFDs by the pilots.
Enlighten me. The 321 NEO is more efficient and has been tweaked incrementally over time. But it’s based on 1980s design philosophy and regulations. Name three differences between a 2021 NEO and the 1988 A320 that have an impact on this discussion?
If only one pilot is left in the flight deck wouldn't we need like a camera link to the flight crew rest and system and nav/weather displays in the back to give the second guy the immediate big picture? What happens with a broken windscreen, cockpit fires and such?
Even AF447 with a full crew left not enough time for the captain to regain situational awareness/control after he had to be called back from the crewrest.
I think it is generally wrong to use a cockpit layout made for two with only one person at work.
Even AF447 with a full crew left not enough time for the captain to regain situational awareness/control after he had to be called back from the crewrest.
I think it is generally wrong to use a cockpit layout made for two with only one person at work.
Join Date: Aug 2006
Location: Lemonia. Best Greek in the world
Posts: 1,754
Received 0 Likes
on
0 Posts
I suspect that plane drivers might do a little more than Stan from On the Buses. Plus, given the issues with autonomous driving of cars etc in 2D, one suspects that 3 D might be a little more difficult.
Pegase Driver
Join Date: May 1997
Location: Europe
Age: 73
Posts: 3,419
Likes: 0
Received 0 Likes
on
0 Posts
BoeingDriver99 : I think 3greens mean cockpit automation rather than aircraft design , but I let him or her reply .
I will make myself the devil's advocate here . We had these heated discussions a few times before internet existed : first in the early 70s when the Navigator was replaced by the Inertial platforms and in the 80s where Airbus removed the FE from the A310.
For me as ATC the number of incidents I saw in my career caused by crews that would not had happened if the automation was switched on and/or had been followed is quite remarkable. I include EGPWS alerts here . Of course the number of incidents caused by faulty automation and prevented by crews is not recorded in the same way.
One remark however : we had and still have a very high number of TCAS incidents caused by crews not reacting or following TCAS RAs correctly . But none involving the A380 and 350s. why? because they have the RAs linked to the FD and those are flown automatically. One of the most tragic collision in recent time, Ueberlingen would not have happened if automation had been available.
But as I said ,am the devil's advocate here, I prefer and support the 4 eyes principle , both in the air and on the ground (ATC)
not really as the drone industry clearly demonstrates, but we are not talking about autonomous flight here , just reducing the number of the crew.
I will make myself the devil's advocate here . We had these heated discussions a few times before internet existed : first in the early 70s when the Navigator was replaced by the Inertial platforms and in the 80s where Airbus removed the FE from the A310.
For me as ATC the number of incidents I saw in my career caused by crews that would not had happened if the automation was switched on and/or had been followed is quite remarkable. I include EGPWS alerts here . Of course the number of incidents caused by faulty automation and prevented by crews is not recorded in the same way.
One remark however : we had and still have a very high number of TCAS incidents caused by crews not reacting or following TCAS RAs correctly . But none involving the A380 and 350s. why? because they have the RAs linked to the FD and those are flown automatically. One of the most tragic collision in recent time, Ueberlingen would not have happened if automation had been available.
But as I said ,am the devil's advocate here, I prefer and support the 4 eyes principle , both in the air and on the ground (ATC)
given the issues with autonomous driving of cars etc in 2D, one suspects that 3 D might be a little more difficult.
zlin77
The driver assist features have systems that monitor the driver to make sure they are still paying attention - and if they are not alerts them and disconnects.
It should be relatively easy to incorporate some sort of system that monitors the PF for signs of subtle incapacitation - and if sensed would alert the resting pilot that they better get back to the flight deck ASAP.
The automotive automation is leaning heavily on what's been learned from aviation automation/pilot interactions. As automotive automation advances and overtakes aviation (which it will), that should become a two way street.
The driver assist features have systems that monitor the driver to make sure they are still paying attention - and if they are not alerts them and disconnects.
It should be relatively easy to incorporate some sort of system that monitors the PF for signs of subtle incapacitation - and if sensed would alert the resting pilot that they better get back to the flight deck ASAP.
The automotive automation is leaning heavily on what's been learned from aviation automation/pilot interactions. As automotive automation advances and overtakes aviation (which it will), that should become a two way street.
Join Date: May 2003
Location: USA
Posts: 83
Likes: 0
Received 0 Likes
on
0 Posts
Good points ATC Watcher thank you. As just a lowly private pilot and self loading freight I don't think I would ever board a long distance commercial flight with one pilot in the front, no matter what the level of "electronic assistance" was available. I'm ok with solo bush piloting, solo short helicopter trips etc, but with GermanWings and now of course the anniversary of 09/11 coming up a solo pilot on a typical commercial flight is just inviting terrorism.
More than one of everything is needed for redundancy. From airliner engines to hydraulic systems to pilots. Going from three to two pilots might be okay somehow but below two is not enough anymore.
Using a cockpit environment made for two pilots with only one pilot in the room is just dangerous. If a single pilot environment is wanted for some reason the entire cockpit must be made for it and certified from the beginning. I can't imagine to see that happen in an airliner. Realistic failure scenarios have to include electrical failures, physical damage and fires with no automation and no datalinks left. This is when you need more than one pilot. In the cockpit not just on the plane. We have seen with the first lion MAX event and with Qantas A380 how much better some enhanced crew could handle unexpected events.
Saving costs by just cutting layers of redundancy would be a quite a paradigm change in aviation.
Using a cockpit environment made for two pilots with only one pilot in the room is just dangerous. If a single pilot environment is wanted for some reason the entire cockpit must be made for it and certified from the beginning. I can't imagine to see that happen in an airliner. Realistic failure scenarios have to include electrical failures, physical damage and fires with no automation and no datalinks left. This is when you need more than one pilot. In the cockpit not just on the plane. We have seen with the first lion MAX event and with Qantas A380 how much better some enhanced crew could handle unexpected events.
Saving costs by just cutting layers of redundancy would be a quite a paradigm change in aviation.
