Airbus + Cathay working on Single Pilot during Cruise with A350
Join Date: Jun 2007
Location: Wanderlust
Posts: 3,406
Likes: 0
Received 0 Likes
on
0 Posts
UplinkerYou cannot take any aircraft and remove one pilot. Definitely it has to have better redundancies. A350 is moving towards that. Unreliable speed, TCAS, EMER DES is handled automatically. In dual engine flame out auto pilot is available. Something more may be in the offing. I repeat again accidents will happen. On another day with three sitting in front A350 incident in Paris would have resulted in a midair. I said before that all apprehensions are justified but it is happening because humans do not seem to guarantee safety. At worst with unusual problem aircraft will crash. It crashes even now without any problem.
vilas you are assuming that the technology required for this airline accountants wet dream is free from the contamination of human input. Software is developed and tested by people and they are subject to the same frailties that you think can be eliminated from the flight deck. It has already been tried and spectacularly failed in the 777. Boeing thought that they could save the airlines money by producing a fault tolerant ADIRU in the 777. It was supposed to be that reliable that an airline would never have to take the unit off the aircraft for repairs. It was tested against the requirements and found to be fully compliant with all the relevant standards. The only thing was it couldn't be tested against any scenarios that the developers could not think of. When something does happen that was not anticipated then you need some analogue biology to counter the digital stuff up.This is what was written when a 777 and its crew had that happen to them:
Thats just one specific component and you are suggesting that many more functions can be automated and perform flawlessly throughout the life of the aircraft! There is a very good reason why a major airline like Lufthansa do not consider the single pilot concept to be safe or worthwhile pursuing.
ADIRU design and checklist items
The B777 aircraft was designed to achieve a level of serviceability and system redundancy that would allow operators to reduce maintenance costs. The air data
inertial reference unit (ADIRU), with its fault-tolerant design and advances in software capability, was a significant part of that design philosophy. The built-in
redundancy was designed to allow for deferred maintenance on the ADIRU and to reduce the flight crew actions required in responding to any fault within the unit.
An internal failure would not be apparent to the flight crew during normal operations, other than through an Engine Indication and Crew Alerting System
(EICAS) status message. The B777 Quick Reference Handbook (QRH) did not, and could not, include ‘…checklists for all conceivable situations…’. Therefore, the
aircraft manufacturer did not include an AIRSPEED UNRELIABLE checklist in the B777 QRH.
When the upset event occurred and the primary flight display indicated an underspeed, then an overspeed condition, as well as the slip/skid indicator showing
full right deflection, the crew experienced a situation that had previously been considered not possible. The primary flight display pitch and roll indications, and
the standby instrument indications were not affected by the failure of the accelerometer within the ADIRU, but the crew were not sure which indications
were correct.
ADIRU operational program software The certification of the ADIRU operational program software (OPS) was dependent
on it being tested against the requirements specified in the initial design. The conditions involved in this event were not identified in the testing requirements, so
were not tested. The mitigating effects of the mid-value select and secondary attitude and air-data reference unit on the primary flight computer response to the erroneous
accelerometer outputs was not an intended function, but did prevent a more severe upset event from occurring.
The B777 aircraft was designed to achieve a level of serviceability and system redundancy that would allow operators to reduce maintenance costs. The air data
inertial reference unit (ADIRU), with its fault-tolerant design and advances in software capability, was a significant part of that design philosophy. The built-in
redundancy was designed to allow for deferred maintenance on the ADIRU and to reduce the flight crew actions required in responding to any fault within the unit.
An internal failure would not be apparent to the flight crew during normal operations, other than through an Engine Indication and Crew Alerting System
(EICAS) status message. The B777 Quick Reference Handbook (QRH) did not, and could not, include ‘…checklists for all conceivable situations…’. Therefore, the
aircraft manufacturer did not include an AIRSPEED UNRELIABLE checklist in the B777 QRH.
When the upset event occurred and the primary flight display indicated an underspeed, then an overspeed condition, as well as the slip/skid indicator showing
full right deflection, the crew experienced a situation that had previously been considered not possible. The primary flight display pitch and roll indications, and
the standby instrument indications were not affected by the failure of the accelerometer within the ADIRU, but the crew were not sure which indications
were correct.
ADIRU operational program software The certification of the ADIRU operational program software (OPS) was dependent
on it being tested against the requirements specified in the initial design. The conditions involved in this event were not identified in the testing requirements, so
were not tested. The mitigating effects of the mid-value select and secondary attitude and air-data reference unit on the primary flight computer response to the erroneous
accelerometer outputs was not an intended function, but did prevent a more severe upset event from occurring.
Hi vilas yes, obviously, aircraft will have to become semi-autonomous but no matter how much autonomy an aircraft might have, it could still suffer an engine turbine failure that takes out multiple other systems, such as the Qantas A380, or the Sioux City DC -10. Or a decompression. Or an avionics fire.
That's when you need two, well rested, well trained pilots on the flight deck. Having one pilot in the cockpit and another asleep in the cabin is not going to work well for such emergencies. That would be like shutting down an engine during the cruise to save money ! (in a twin). Very foolhardy.
Yes, pilots make mistakes, and, frankly, there have been far too many basic piloting mistakes in recent years - I am as appalled about that as you probably are.
But that is because training and examining have become diluted, so that marginal, and under-skilled pilots are now on the flight deck, in charge of passenger aircraft.
The simple solution is not to spend countless billions on developing autonomous aircraft, but spend some millions on better training and examining of pilots - to make sure that only competent skilled pilots get onto the flight deck in the first place.
Obviously no airline or aircraft manufacturer is going to listen to me though
That's when you need two, well rested, well trained pilots on the flight deck. Having one pilot in the cockpit and another asleep in the cabin is not going to work well for such emergencies. That would be like shutting down an engine during the cruise to save money ! (in a twin). Very foolhardy.
Yes, pilots make mistakes, and, frankly, there have been far too many basic piloting mistakes in recent years - I am as appalled about that as you probably are.
But that is because training and examining have become diluted, so that marginal, and under-skilled pilots are now on the flight deck, in charge of passenger aircraft.
The simple solution is not to spend countless billions on developing autonomous aircraft, but spend some millions on better training and examining of pilots - to make sure that only competent skilled pilots get onto the flight deck in the first place.
Obviously no airline or aircraft manufacturer is going to listen to me though
Last edited by Uplinker; 29th Aug 2021 at 10:23. Reason: Clarification about which point of vilas' I was referring to, since quote boxes keep disappearing
Uplinker, shared reality;
The human contribution in safety approaches its limit; best to focus on maintaining the current level with training.
Humans cannot be considered as machines; not every training scenario can be foreseen or practiced, and that which is, can never guarantee to have well rested, well trained pilots with reliably memory, requisite situation assessment, and decision making.
Who pays; operators have difficulty in balancing current training budgets, increases unlikely to be passed on to the customer.
For technical solitions, manufactures should be able to absorb some cost, if there are advantages from reduced cost of operation, particularly with new aircraft - more sales. Similarly operators with reduced operating cost.
There are two distinctly separate concepts:- single pilot throughout the flight for commercial operations is distant, requiring new designs and certification standards.
Limited single pilot operations should be feasible with today's technology and certification; proof of concept required - AI & operators.
The certification of modern aircraft includes single pilot operation in considering incapacitation - continued flight and landing.
Thence as per slast.
Defeatist hypothetical scenarios are easy to imagine, but first assess these against certification requirements which are probability based; a realisation that not everything can be avoided or managed, but that which is, provides an acceptable level of safety in an uncertain world. Any new concept has to heed this.
A better understanding of the issues could be gained by considering how these new concepts can be achieved; how can systems be designed, tested, and certificated. Use the expertise of current operations and pilots to solve problems, not create flimsy barriers to inevitable progress.
The human contribution in safety approaches its limit; best to focus on maintaining the current level with training.
Humans cannot be considered as machines; not every training scenario can be foreseen or practiced, and that which is, can never guarantee to have well rested, well trained pilots with reliably memory, requisite situation assessment, and decision making.
Who pays; operators have difficulty in balancing current training budgets, increases unlikely to be passed on to the customer.
For technical solitions, manufactures should be able to absorb some cost, if there are advantages from reduced cost of operation, particularly with new aircraft - more sales. Similarly operators with reduced operating cost.
There are two distinctly separate concepts:- single pilot throughout the flight for commercial operations is distant, requiring new designs and certification standards.
Limited single pilot operations should be feasible with today's technology and certification; proof of concept required - AI & operators.
The certification of modern aircraft includes single pilot operation in considering incapacitation - continued flight and landing.
Thence as per slast.
Defeatist hypothetical scenarios are easy to imagine, but first assess these against certification requirements which are probability based; a realisation that not everything can be avoided or managed, but that which is, provides an acceptable level of safety in an uncertain world. Any new concept has to heed this.
A better understanding of the issues could be gained by considering how these new concepts can be achieved; how can systems be designed, tested, and certificated. Use the expertise of current operations and pilots to solve problems, not create flimsy barriers to inevitable progress.
Safetypee, yes you are right, I was really only referring to the idea of autonomy: the extract from uplinker I wanted to quote seemed to have gone AWOL, which rathe defeated the point! I've put it back now.
As far as single pilot cruise is concerned, how much is "controlled napping" allowed now?
As far as single pilot cruise is concerned, how much is "controlled napping" allowed now?
Flimsy barriers? Inevitable "progress"? Defeatist !
Good luck with it - you're going to need a huge amount of good luck.
Good luck with it - you're going to need a huge amount of good luck.
Last edited by Uplinker; 29th Aug 2021 at 10:36.
Join Date: May 2000
Location: SV Marie Celeste
Posts: 655
Likes: 0
Received 0 Likes
on
0 Posts
It is easy to look at an specific accident, point out the limitations of the pilot on that occasion and conclude the best option is to eliminate the pilot. What is missing in that logic is the times the machine malfunctions and it is the pilot intervention that saves the day. It is the relationship between the number of those two types of occurrences that should determine if the pilot is removed or not. With current equipment there must be ten of thousands of human interventions that save the day for each one that makes the situation worse. We have a loooooong way to go before we get to break even.
calypso: 
Re ‘What is missing in that logic is the times the machine malfunctions and it is the pilot intervention that saves the day.’ Yes; …
Reinterpreting this as the need for ‘understanding’ - knowledge of these issues and interventions, requires looking at and recording normal work, human adaptation, how often things go right.
This is widely discussed in ‘Safety II’ and ‘Resilience’; complementary ways of thinking about safety and everyday operations - view the human as an asset.
A significant point is the number of situations which might be encountered in modern aircraft (probability) and how many actually require two crew - need vs nice; and if two; what is the residual risk - time available for the resting pilot, vs a non-pilot in the other seat, getting back to the cockpit.
Re ‘What is missing in that logic is the times the machine malfunctions and it is the pilot intervention that saves the day.’ Yes; …
Reinterpreting this as the need for ‘understanding’ - knowledge of these issues and interventions, requires looking at and recording normal work, human adaptation, how often things go right.
This is widely discussed in ‘Safety II’ and ‘Resilience’; complementary ways of thinking about safety and everyday operations - view the human as an asset.
A significant point is the number of situations which might be encountered in modern aircraft (probability) and how many actually require two crew - need vs nice; and if two; what is the residual risk - time available for the resting pilot, vs a non-pilot in the other seat, getting back to the cockpit.
We had GEN 1, (A321), go intermittent while in the cruise, about an hour out of our UK base - full output, no output, full output, no output etc. As quickly as it takes to say that. The shock loading on the engine made a sound like back-firing - bang bang bang.
To start with, we didn't know what had gone wrong and the ECAM and the cockpit warning lights and the single chime were chaotically going on and off like a Christmas tree, making no sense, and we could not follow the ECAM instructions at all - everything kept changing so rapidly. The aircraft could not cope, Cap lost all his instruments, and the aircraft spat out the A/P and the A/THR. So I took control and flew manually, with manual thrust.
Cap was trying to make sense of what was happening, but he couldn't, because it was changing so rapidly. Out of the corner of my eye, I noticed - during the fractions of a second that the ELEC page was displayed - that there seemed to be something amber in the region of GEN 1. With my electronics background, I realised what might be the problem, and I suggested to the Cap that we switch off GEN 1 completely.
This he did and all became calm and we could finally understand what had happened. We started the APU, restored the A/P etc, and continued home.
Whilst it was very disturbing, we were never in immediate danger and were over land, near many airfields, not over the middle of an ocean. But the fault rendered the on-board diagnostic ECAM useless, and it took two pilots to sort out the problem: One to hand-fly, the other to reconfigure the aircraft, and both to work out what the problem was.
Would a semi-autonomous aircraft cope if its electrical supply became compromised?
To start with, we didn't know what had gone wrong and the ECAM and the cockpit warning lights and the single chime were chaotically going on and off like a Christmas tree, making no sense, and we could not follow the ECAM instructions at all - everything kept changing so rapidly. The aircraft could not cope, Cap lost all his instruments, and the aircraft spat out the A/P and the A/THR. So I took control and flew manually, with manual thrust.
Cap was trying to make sense of what was happening, but he couldn't, because it was changing so rapidly. Out of the corner of my eye, I noticed - during the fractions of a second that the ELEC page was displayed - that there seemed to be something amber in the region of GEN 1. With my electronics background, I realised what might be the problem, and I suggested to the Cap that we switch off GEN 1 completely.
This he did and all became calm and we could finally understand what had happened. We started the APU, restored the A/P etc, and continued home.
Whilst it was very disturbing, we were never in immediate danger and were over land, near many airfields, not over the middle of an ocean. But the fault rendered the on-board diagnostic ECAM useless, and it took two pilots to sort out the problem: One to hand-fly, the other to reconfigure the aircraft, and both to work out what the problem was.
Would a semi-autonomous aircraft cope if its electrical supply became compromised?
Exactly the sort of event I referred to in a post which seems to have been deleted.
Did this get investigated as a serious incident?
Is there any investigation report?
Did this get investigated as a serious incident?
Is there any investigation report?
And here, ladies and gentleman, is an excellent reason to retain two qualified pilots on the flightdeck.
Incident: Biman B738 near Nagpur on Aug 27th 2021, captain incapacitated and later died
Incident: Biman B738 near Nagpur on Aug 27th 2021, captain incapacitated and later died
Incident: Biman B738 near Nagpur on Aug 27th 2021, captain incapacitated and later died
Incident: Biman B738 near Nagpur on Aug 27th 2021, captain incapacitated and later died
Only half a speed-brake
Yes, present day aircraft require 2 pilots. No surprise.
Present day technology, not installed on present day aircraft, is already capable to overcome the challenges.
All the cases listed above included, the HF kills more people these days than it saves. Problem with humans is we cannot be elevated to a higher level of sentient beings. Less prone to error, more emotionally stable, easier to replicate (cross-generation training), repairable.
The future aircraft case to consider:
- autonomously capable
- remotely piloted
- locally supervised on-board by 1 ex-pilot, System Reliability Assurance Engineer.
The F/E's told us to be back.
Present day technology, not installed on present day aircraft, is already capable to overcome the challenges.
All the cases listed above included, the HF kills more people these days than it saves. Problem with humans is we cannot be elevated to a higher level of sentient beings. Less prone to error, more emotionally stable, easier to replicate (cross-generation training), repairable.
The future aircraft case to consider:
- autonomously capable
- remotely piloted
- locally supervised on-board by 1 ex-pilot, System Reliability Assurance Engineer.
The F/E's told us to be back.
Join Date: Jun 2007
Location: Wanderlust
Posts: 3,406
Likes: 0
Received 0 Likes
on
0 Posts
Within a year Air bus is introducing a robust AP that won't trip outside the envelope i.e. alpha prot but take action to come out with intimation on FMA. AP will be available with unreliable speed, ADR 1+2+3 Fail. And an alternate AP will be available which will maintain VS0, Hdg and ATHR will maintain a fixed speed. So coming decade will give a clear idea where aviation is heading.
Join Date: Oct 2003
Location: If this is Tuesday, it must be?
Posts: 651
Likes: 0
Received 0 Likes
on
0 Posts
the HF kills more people these days than it saves
I have no doubt that one day we will see single pilot and autonomous airliners, but we have a long way to go to that point. People, especially the techno-geeks, get blinded by the capabilities of their new toy while forgetting about 1) reliability and 2) the fact that not all the world has the necessary infrastructure. We've had autoland technology since the 60s, but what proportion of airports have Cat3 on every runway? I'd love to see some of this autonomous technology cope with African ATC...
Only half a speed-brake
BizJetJock
Also not what I said
Thought-provoking statement, hopefully. Crew working well don't enter the equation, the crew today working beyond what the Future Gen setup is able to deliver is what fits the comparison. What you call saves are HF solutions to situations rooted in the presence of the necessary but unsupervised and uncontrolled HF element onboard to begin with.
BTW don't omit the remote piloted part, please. Three standalone layers as opposed to a single layer of 2 humans now.
Even Uplinker's case is admitted to be a one-man save, is it not?
+ autonomous
+ remotely piloted from C&C
+ on-board supervised with live (single) backup.
UPS 6
Air Canada snowplough @ SFO
Were those saves or kills? Only through huge luck is how those did not become the largest disasters of all aviation time.
On-board HF will lose the reliability battle through a long-term continuous improvement programme of the HW and SW solutions, still controlled by human operator teams, remotely.
Also not what I said
Thought-provoking statement, hopefully. Crew working well don't enter the equation, the crew today working beyond what the Future Gen setup is able to deliver is what fits the comparison. What you call saves are HF solutions to situations rooted in the presence of the necessary but unsupervised and uncontrolled HF element onboard to begin with.
BTW don't omit the remote piloted part, please. Three standalone layers as opposed to a single layer of 2 humans now.
Even Uplinker's case is admitted to be a one-man save, is it not?
+ autonomous
+ remotely piloted from C&C
+ on-board supervised with live (single) backup.
UPS 6
Air Canada snowplough @ SFO
Were those saves or kills? Only through huge luck is how those did not become the largest disasters of all aviation time.
On-board HF will lose the reliability battle through a long-term continuous improvement programme of the HW and SW solutions, still controlled by human operator teams, remotely.
Last edited by FlightDetent; 4th Sep 2021 at 17:21.
Only half a speed-brake
vilas
Yes.
And the Future Gen (cargo first) airliner with enough system integration to take advantage of that will be built around 2035 with design-useable life of 2040-2080. My suggested setup will be probably viewed as overly conservative then.
Yes.
And the Future Gen (cargo first) airliner with enough system integration to take advantage of that will be built around 2035 with design-useable life of 2040-2080. My suggested setup will be probably viewed as overly conservative then.
Only half a speed-brake
calypso
you start. JACDEC is a good source for inspiration.
Send one example where HF did a save that today's tech onboard future installation would not be able to. I'll equal that and then we take turns.
For handicapping, be invited to choose how to restric my scope. Either the "west" accidents of the past (historical HF) or the "asian" accidents of present (HF of the next three decades).
you start. JACDEC is a good source for inspiration.
Send one example where HF did a save that today's tech onboard future installation would not be able to. I'll equal that and then we take turns.
For handicapping, be invited to choose how to restric my scope. Either the "west" accidents of the past (historical HF) or the "asian" accidents of present (HF of the next three decades).
Join Date: Oct 2015
Location: Tranquility Base
Posts: 77
Likes: 0
Received 0 Likes
on
0 Posts
FlightDetent, not wanting to push down your credibility, merely trying to understand where your reasoning is coming from: are you or have you been a professional pilot with an airline?
For me as a professional pilot in an airline it is hard to follow what you are saying. Every day I go to work we as a crew make dozens of decision to get the job done in a non-ideal, real-world, complex aviation environment. I don't see software to be able to manage that, also not with so called "future" technology, because it would still have to fit into the non-ideal, real-world, complex aviation environment.
For me as a professional pilot in an airline it is hard to follow what you are saying. Every day I go to work we as a crew make dozens of decision to get the job done in a non-ideal, real-world, complex aviation environment. I don't see software to be able to manage that, also not with so called "future" technology, because it would still have to fit into the non-ideal, real-world, complex aviation environment.
Join Date: Jul 2008
Location: USA
Posts: 22
Likes: 0
Received 0 Likes
on
0 Posts
People hold out AF447 as a pilot failure but seem to forget they were responding to a complex systems failure. The BAE report lists 13 other occurrences of the same system failure with different outcomes. If the report captured 13 occurrences, how many have/had occurred fleet wide, that were mitigated successfully by those up front?