Aviation regulators push for more automation so flights can be run by a single pilot
Join Date: Jun 2002
Location: Dubai
Age: 55
Posts: 313
Likes: 0
Received 0 Likes
on
0 Posts
There is no doubt automation/AI etc can achieve just about anything these days, but I am not convinced it can handle multiple complex failures and 'think out of the square'. A fully automated aeroplane might have saved AF447 but I doubt it would have saved QF72 and QF32.
Current state of implemented AI is pretty good but far from reliable enough for situations where there is risk of large scale loss of life. In medicine a failure usually inolves a single fatality iany seriopus event, so risk of using it for diagnosis is very different and failure rate vs human judgement is probably better and more acceptable? Current status perhaps illustrated by this story .... Astronomer in Twitter limbo over 'intimate' meteor - BBC News
Join Date: Nov 2022
Location: Usa
Posts: 1
Likes: 0
Received 0 Likes
on
0 Posts
100% commercial flights will be pilotless in the future. When will that be, no one knows. Ai will easily fly a to b. Desperate for a back up pilot? Robot on a seperate operating system.
Quite a timely interview of Richard Champion de Crespigny who was flying QF32 when one of the engines went "bang". By an Australia journalist Peter FitzSimons. As there were 2 check pilots on board, they had 5 pilots.
One-pilot cockpits? Here’s what QF32 hero and ‘Sully’ Sullenberger think
Richard Champion de Crespigny flew with the RAAF for 11 years, before flying Qantas jets for three and a half decades. In 2010, he famously helped save the lives of 469 passengers and crew when an engine exploded on the Airbus A380 he was in command of. I spoke to him on Friday.
Fitz: Richard, the news this week is that – as a cost-saving exercise – regulators and a number of airlines around the globe are pushing towards just one pilot in the cockpit. We’ll get to that. But first, tell me the story of flight QF32 from Singapore to Sydney on November 4, 2010, while we reflect on how you would have gone if you were alone. Please start in the traditional manner for all disasters, with “First, I heard a loud bang...”
RDC: First I heard two loud bangs. The first one was an engine surge, or backfire of Engine No.2. The second was the sound of the turbine exploding into three pieces that exited the engine at over two and a half times the speed of sound.
Fitz: Captain, my captain, I’m with you in the cockpit, and we’re in trouble. What’s the damage?
RDC: It’s like a cluster-bomb. Shrapnel has hit the fuselage in over 400 locations, with 200 impacts on the fuselage, 200 on the wing, even 20 to the top of the eight-story high tail fin. 650 wires are cut, and half the networks fail. Twenty one of the aircraft’s 22 systems are degraded.
Fitz: Captain, I wish to report that, to use the technical term, we are in deep ****! What now?
continues...
One-pilot cockpits? Here’s what QF32 hero and ‘Sully’ Sullenberger think
Richard Champion de Crespigny flew with the RAAF for 11 years, before flying Qantas jets for three and a half decades. In 2010, he famously helped save the lives of 469 passengers and crew when an engine exploded on the Airbus A380 he was in command of. I spoke to him on Friday.
Fitz: Richard, the news this week is that – as a cost-saving exercise – regulators and a number of airlines around the globe are pushing towards just one pilot in the cockpit. We’ll get to that. But first, tell me the story of flight QF32 from Singapore to Sydney on November 4, 2010, while we reflect on how you would have gone if you were alone. Please start in the traditional manner for all disasters, with “First, I heard a loud bang...”
RDC: First I heard two loud bangs. The first one was an engine surge, or backfire of Engine No.2. The second was the sound of the turbine exploding into three pieces that exited the engine at over two and a half times the speed of sound.
Fitz: Captain, my captain, I’m with you in the cockpit, and we’re in trouble. What’s the damage?
RDC: It’s like a cluster-bomb. Shrapnel has hit the fuselage in over 400 locations, with 200 impacts on the fuselage, 200 on the wing, even 20 to the top of the eight-story high tail fin. 650 wires are cut, and half the networks fail. Twenty one of the aircraft’s 22 systems are degraded.
Fitz: Captain, I wish to report that, to use the technical term, we are in deep ****! What now?
continues...
Well, if they want to remove a seat from the flight deck, they may as well remove 150 from the cabin.
Join Date: Feb 2008
Location: victoria bc
Age: 82
Posts: 77
Likes: 0
Received 0 Likes
on
0 Posts
For as long as pilots continue to fly airplanes there will be two of them. When the time comes for airplanes to be flown by other means, there may very well be a designated crew person aboard but his duties and responsibilities will not be what we now consider “ piloting” the airplane. Pilots ( plural ) fly the airplanes or something else does. That something else will not be a solo pilot.
Join Date: Nov 2003
Location: Germany
Posts: 137
Likes: 0
Received 0 Likes
on
0 Posts
HI, I think that the accident record of large drones -and airplanes without pilots are drones, mindless drones- is not good. The only crash database for drones i know is here: https://dronewars.net/drone-crash-database/
One has to subtract those that were shot down, but add those crashes that didn´t make it into this unofficial record. Given the relatively low number of large drones (hundrets, not tens of thousands) , I think that we are looking at many more crashes than piloted aircraft would have had.
One has to subtract those that were shot down, but add those crashes that didn´t make it into this unofficial record. Given the relatively low number of large drones (hundrets, not tens of thousands) , I think that we are looking at many more crashes than piloted aircraft would have had.
So is the industry prepared to move the weakest link in the chain from the flight deck to guy who wrote the code?? In reality this really doesn't do anything it would just maybe fix one problem and create a new one that doesn't exist at the moment.
A330 upset, note the use of the word "probably" in the report, not a comforting assessment. Captain retired with PTSD, one tough day at work. The report has a good discussion on computerisation.
One of the aircraft's three air data inertial reference units (ADIRUs) started outputting intermittent, incorrect values (spikes) on all flight parameters to other aircraft systems. Two minutes later, in response to spikes in angle of attack (AOA) data, the aircraft's flight control primary computers (FCPCs) commanded the aircraft to pitch down. At least 110 of the 303 passengers and nine of the 12 crew members were injured; 12 of the occupants were seriously injured and another 39 received hospital medical treatment.
Although the FCPC algorithm for processing AOA data was generally very effective, it could not manage a scenario where there were multiple spikes in AOA from one ADIRU that were 1.2 seconds apart. The occurrence was the only known example where this design limitation led to a pitch-down command in over 28 million flight hours on A330/A340 aircraft, and the aircraft manufacturer subsequently redesigned the AOA algorithm to prevent the same type of accident from occurring again.
Each of the intermittent data spikes was probably generated when the LTN-101 ADIRU's central processor unit (CPU) module combined the data value from one parameter with the label for another parameter. The failure mode was probably initiated by a single, rare type of internal or external trigger event combined with a marginal susceptibility to that type of event within a hardware component. There were only three known occasions of the failure mode in over 128 million hours of unit operation. At the aircraft manufacturer's request, the ADIRU manufacturer has modified the LTN-101 ADIRU to improve its ability to detect data transmission failures.
It is generally accepted that, for all but the simplest systems, it is impossible to guarantee the correctness of all the system requirements and associated assumptions.
The current aviation system has an enviable safety record; however, advances in technology are placing an increasing strain on our ability to assure the integrity of new and anticipated systems.
https://www.atsb.gov.au/sites/defaul.../ao2008070.pdf
Although the FCPC algorithm for processing AOA data was generally very effective, it could not manage a scenario where there were multiple spikes in AOA from one ADIRU that were 1.2 seconds apart. The occurrence was the only known example where this design limitation led to a pitch-down command in over 28 million flight hours on A330/A340 aircraft, and the aircraft manufacturer subsequently redesigned the AOA algorithm to prevent the same type of accident from occurring again.
Each of the intermittent data spikes was probably generated when the LTN-101 ADIRU's central processor unit (CPU) module combined the data value from one parameter with the label for another parameter. The failure mode was probably initiated by a single, rare type of internal or external trigger event combined with a marginal susceptibility to that type of event within a hardware component. There were only three known occasions of the failure mode in over 128 million hours of unit operation. At the aircraft manufacturer's request, the ADIRU manufacturer has modified the LTN-101 ADIRU to improve its ability to detect data transmission failures.
It is generally accepted that, for all but the simplest systems, it is impossible to guarantee the correctness of all the system requirements and associated assumptions.
The current aviation system has an enviable safety record; however, advances in technology are placing an increasing strain on our ability to assure the integrity of new and anticipated systems.
https://www.atsb.gov.au/sites/defaul.../ao2008070.pdf
The ADIRU OPS versions up to and including version -07 contained a latent software error in the algorithm to manage the sensor set used for computing flight control outputs which, after the unit went through a power cycle, did not recognise that accelerometer number-5 was unserviceable.
An anomaly existed in the component software hierarchy that allowed inputs from a known faulty accelerometer to be processed by the air data inertial reference unit (ADIRU) and used by the primary flight computer, autopilot and other aircraft systems.
When the hardware failure occurred, combined with the software anomaly, the crew were faced with an unexpected situation that had not been foreseen.
The software anomaly was not detected in the original testing and certification of the ADIRU.
https://www.atsb.gov.au/sites/defaul...503722_001.pdf
An anomaly existed in the component software hierarchy that allowed inputs from a known faulty accelerometer to be processed by the air data inertial reference unit (ADIRU) and used by the primary flight computer, autopilot and other aircraft systems.
When the hardware failure occurred, combined with the software anomaly, the crew were faced with an unexpected situation that had not been foreseen.
The software anomaly was not detected in the original testing and certification of the ADIRU.
https://www.atsb.gov.au/sites/defaul...503722_001.pdf
The “inevitable pilotless airliners” story happens every couple of years, someone finds the slightest comment to get media attention, technology aficionados make comments saying “it’s only a matter of time”, pilots say “not so fast”, manufacturers say “we’re exploring all possibilities”…… and eventually nothing eventuates, the stories are put to bed before the media cycle happens again.
Like the “Cathay A350 single pilot in cruise” stories that happened a few years ago. Once the media had run their stories it all went quiet.
Like the “Cathay A350 single pilot in cruise” stories that happened a few years ago. Once the media had run their stories it all went quiet.
Wouldn't remote controlled or AI guided airliners need some fail-safe guaranteed network connection at all times? How could this be guaranteed? What happens if Satellites get jammed or break down or reception gets affected by sun storms or similar? Then your shiny bot-plane suddenly flies on it's own into the next tropical depression without all the fall back safety promised?
Megan, thanks for the links - I wasn't aware of the A320 one. Of course what none of these reports can say is what would likely happened if there had been no crew intervention at all, and the automated systems just been allowed to continue doing their thing.
Less hair: or if in times of international tension the likes of Putin/Xi simply state that they have and will use the technology to disrupt satellite signals? I’ve said in other threads that a fundamental reason it’s unlikely that uncrewed passenger ops will take place in the foreseeable future is product liability – take the pilots out and you have no flying scapegoat, ALL legal responsibility is transferred to the manufacturer or operator. You have to completely rewrite things like ICAO Annex 2, 2.4: Authority of pilot-in-command of an aircraft. “The pilot-in-command of an aircraft shall have final authority as to the disposition of the aircraft while in command.”
The philosophical question is this. The industry can say “we have a 100% (not 99.9999999%) guarantee that we can take care of ALL CONCEIVABLE combinations of failures using neural network AI”. Are there INconceivable failures of mechanical parts, software coding and environmental aspects? How do you deal with that? They seem to have occurred in quite a few events.
And to deal with loss of comms for some reason – easily conceivable - all this stuff has to be on board each aircraft. The presence of (a) pilot or pilots provides the only totally independent intelligence able to make an attempt (not guaranteed successful) to deal with an inconceivable combination of events. When the detailed “terms and conditions” are written by the manufacturers lawyers, the operators lawyers are going to say “no way”!
Less hair: or if in times of international tension the likes of Putin/Xi simply state that they have and will use the technology to disrupt satellite signals? I’ve said in other threads that a fundamental reason it’s unlikely that uncrewed passenger ops will take place in the foreseeable future is product liability – take the pilots out and you have no flying scapegoat, ALL legal responsibility is transferred to the manufacturer or operator. You have to completely rewrite things like ICAO Annex 2, 2.4: Authority of pilot-in-command of an aircraft. “The pilot-in-command of an aircraft shall have final authority as to the disposition of the aircraft while in command.”
The philosophical question is this. The industry can say “we have a 100% (not 99.9999999%) guarantee that we can take care of ALL CONCEIVABLE combinations of failures using neural network AI”. Are there INconceivable failures of mechanical parts, software coding and environmental aspects? How do you deal with that? They seem to have occurred in quite a few events.
And to deal with loss of comms for some reason – easily conceivable - all this stuff has to be on board each aircraft. The presence of (a) pilot or pilots provides the only totally independent intelligence able to make an attempt (not guaranteed successful) to deal with an inconceivable combination of events. When the detailed “terms and conditions” are written by the manufacturers lawyers, the operators lawyers are going to say “no way”!
What about the probably 100 incidents a day worldwide that we never hear of where an airliner suffers an autopilot malfunction and the pilots simply disconnect, manually fly and then just write up the defect in the tech log? Or any other system defect automation can’t handle and again all that is written about it is a tech log entry?
Exactly.... the ATSB B777 report has some interesting material about the history of the ADIRU involved, as well as operating with disatch deviations.
With this 'RUSH' to work from home, they can let the pilots operate the flight deck from home, and the cabin crew can be 'upskilled' to deal with any issues that may occur, on the other hand the upskilling will have to include EFTO, Bomb Threats, diversions, multiple bird strikes, cabin fires, cargo fires, fuel freezing, tea pot failure, depressurisation, battery fires, and the millions of other items than can mess up the day. I think they do well to manage on two crew, prob better to get an extra on in to reduce the workload.
Am I the only one who fails to see the relevance of whether pilot(s) are single, married or in a committed long-term relationship?
On an unrelated topic, a counter argument to prospective solo operation of commercial air transport might be to consider further multi-skilling of first officers. All that unnecessary spare capacity (in between in-flight crises) could perhaps be employed in other on-board areas, such as children's entertainer, or singer-pianist at the A380's walk-up bar lounge?
On an unrelated topic, a counter argument to prospective solo operation of commercial air transport might be to consider further multi-skilling of first officers. All that unnecessary spare capacity (in between in-flight crises) could perhaps be employed in other on-board areas, such as children's entertainer, or singer-pianist at the A380's walk-up bar lounge?
Pegase Driver
Join Date: May 1997
Location: Europe
Age: 73
Posts: 3,669
Likes: 0
Received 0 Likes
on
0 Posts
It’s not just the incidents that are reported in the press and are known to the world.
What about the probably 100 incidents a day worldwide that we never hear of where an airliner suffers an autopilot malfunction and the pilots simply disconnect, manually fly and then just write up the defect in the tech log? Or any other system defect automation can’t handle and again all that is written about it is a tech log entry?
What about the probably 100 incidents a day worldwide that we never hear of where an airliner suffers an autopilot malfunction and the pilots simply disconnect, manually fly and then just write up the defect in the tech log? Or any other system defect automation can’t handle and again all that is written about it is a tech log entry?
QF32 is a good example and I talked with De Crepigny on this long ago , thinking outside the box saved them , no following the check lists or EICAS messages , which a computer system would have .
Full automnation carry pax ? not in our lifetime. but sincle pilot ? well Embraer is working very seriously on it , the only major diference between their Business jets and their airliners is the number of pax in the back. The cockpit automation is the same . When the number of flying hours of their single pilots business jets will be large enough to demonstrate ( or not) the safety case things might change and at least keep the discussion alive . But currently , not a good idea to implement.
Last edited by ATC Watcher; 5th Dec 2022 at 09:40. Reason: typos
The two-man cockpit B-21 is said to be optionally manned. I wonder if this technology is supposed to trickle down into the commercial world one day?
