Uplinker

All concerns are justified but examples quoted are wrong on many counts. First, number in front doesn't help. AF447 was crashed by two in front. SFO crashed with three in front. Qantas 747 400 was stalled three times in a hold near HongKong by three in front who not only didn't know the correct holding speed they erased the speed that was in FMS to insert wrong speed and whose combined experience was 40000hrs. So let's put more the merrier theory to bed.
Second, I am very sure Airbus doesn't want to repeat 737MAX saga by shelling out billions to squash the case.
Third, the bogey of EMER DES is not applicable to A350 because if the system cannot prevent the cabin pressure loss and estimates cabin altitude to rise beyond 10000ft in 10seconds it arms the auto EMER DES and if not cancelled or activated, at 15 seconds performs EMER DES without any initiation from pilot that includes cabin announcement, commencing descent to MORA/10000ft, offsetting track by 2.5nm, turning radar tilt down and squacking7700. And if there TCAS RA on the way down it will perform the escape automatically. Unreliable speed is not going to happen either because it has another set of pitot static in the engine that gives alternate speed if that also fails then from GW, CG and lift equation it gives you back up speed. All changes are automatic, the pilot notwithstanding one or two is just informed about the change. All he does may be say thank you and order coffee. AP can be engaged even after dual engine flame out.
Four, Airbus has already done 550hrs on ATTOL Automated Taxi Takeoff and Landing program where the pilot only started in the engines(I hope so) rest of the flight was done automatically. So there may be more automatics on the way in these trials. Offcourse no matter how safe it will be a conflict of interest for the community and no one is going to like it. Boeing may be the biggest opponent of this because as of now they don't have an equivalent. But should it come through they won't remain far behind.

Pilots will not gain more handling skills, practical pilotage experience and stay current if they are denoted to be cruise observers until something serious breaks. This is neither one thing nor the other. Could it be meant to prove that it doesn't work to come to the conclusion better leave pilots out altogether and go Tesla?

vilas, #82; a thoughtful post, reminding us that there are alternative, justifiable points of view

For those who disagree, ok, but please provide a reasoned and argued debate.
Starting with this quote ?

“If you want to fly as [traditional pilots] say they do, then go fly gliders, become test pilots, for all I care go to the moon. But flying for the airlines is not supposed to be an adventure. From takeoff to landing, the autopilots handle the controls. This is routine. In a Boeing as much as an Airbus. And they make better work of it than any pilot can. You’re not supposed to be the blue-eyed hero here. Your job is to make decisions, to stay awake, and to know which buttons to push and when. Your job is to manage the systems.”

Quoted text from 2009 book Fly by Wire: The Geese, the Glide, the Miracle on the Hudson. - William Langewiesche

"Boeing may be the biggest opponent of this because as of now they don't have an equivalent. But should it come through they won't remain far behind".

Boeing is on this now, and has been working on this concept for some time. The 777X autoflight sytem & FBW was designed with this concept in mind, and will continue to serve as the initial platform as time passes. Their plate is pretty full at the moment, so it remains to be see whether it actually comes to fruition in this decade.

alf5071h

Flying attracts those who are thrill seekers, who thrive on excitement. You know what, passengers don't want thrills, the management don't want thrills. The professional pilot who understands this uses his skill to make the flight routine, uninteresting, boring even. That is difficult for those of us who have a high threshold for stimulation.

vilas, yes but I never argued for "the more the merrier", I argued for a minimum of two pilots - not the same thing. The examples you give would not have been solved if only one (duff) pilot was in the cockpit. Like I say, improve pilot training and testing.

The A350, clever as it might be, will not be quite so clever if it suffers a major electrical, or other major system failure.

alf5071h, I have spent hundreds of hours crossing the Atlantic and the Indian oceans, staring forwards into blackness for hours on end, with nothing going wrong, while the aircraft is on A/P. I don't have a problem with doing that at all.

I have also twice had to emergency descend and dump fuel before making an unscheduled landing. If I wanted thrills, I would fly aerobatics in Pitts or Zlins. I agree totally with the last two sentences of your quote. :ok:

Some of the most fatiguing flying I've ever done has been two crew at the limit of FDLs. If I could have left the perfectly competent FO in charge while I sent back for a kip in a comfortable crew rest, and then likewise for him, it would have enhanced safety,

Just to refocus this thread. A350, single pilot, on cruise.

Not - Remove all aircrew, let the computer do all the flight controls. Can we have a little bit of perspective? Yes, it may be the start of a slippery slope, but it's a proposal.

Also, can someone tell me the number of hull losses from events happening whilst at cruise versus hull losses at departure or final approach?

Witness Boeing and the Max fiasco where they tried to save money - a fiasco that killed 300 people.

the difference is that it seems airbus has not cut corners with boat anchor systems like Boeing. Boeing added a control system with only 2 AOA. Airbus added more redundant multi sensor systems to improve safety. Cost a shedload no doubt.

apples for apples the 2 manufacturers do things very differently. No problems with the fear of pulling circuit breakers in a Boeing. The FCOM doesn’t really allow it unless your in a dire circumstance.

Give you a free lunch if you can find a cb on the 350 overhead. There are authorised computer resets in the fcom, different philosophy to Boeing.

I don't think the business case is clear already, as we do not know the applicable regulations that ICAO and EASA will issue in terms of certification/operations/flight time limitations...

Let us have an example:

You can have singe pilot cruise ops (EMCO) certified by proving, among a load of other things, that the probability of an emergency descent multiplied by the probability of an auto emer des sytem failure stays below a certain safety threshold. What if one pack or one bleed is MEL? the probability of an emergency descent goes through the roof, as you have no redundancy anymore (been there...) .Still under that treshold? If not, no extended minimal crew operation today. Crew is augmented, or flight cancelled if no standby crew. And what if the failure happens in flight? Of course, you expect the second pilot to come back from the bunk. But then, what is the maximum flight duty period? The FDP originally planned with EMCO, or basic max FDP? Depending on the regulation, you may end up diverting for some very minor failures.

They can do it, ok. But will that be really profitable in the long run? Not sure. Time will tell...

procede

How do you know the captain would have not pulled up as well? And if other first/second officers were assigned to that flight would all of them pull up using your logic?

No, but this particular second officer would not have been there if they would have been allowed to fly with a two man crew. More humans in this case simply meant more chance of human error.

Second officers generally aren't that experienced and especially inexperienced pilots are know to have really bad panic responses to unexpected situations.

I was just looking at this video on fatal accidents in GA due to students locking the controls in a panic response:

So all this PM/PF confirm/confirmed FPM is complete bunk then? It's good to see engineers and managers have it all sorted.

What about the Germanwings incident where the co-pilot deliberately crashed the aircraft after locking the Captain out?
Since then, many airlines insist on minimum 2 people on flight deck at all times. So, when one needs the lav-a cabin crew must remain on the FD until the other crew returns.

No! They did it with a single AoA. They didn't bother to connect the other one which is installed. Only after the crashes they connected them.

my bad. Thanks for correcting me. That makes it even more ridiculous. This is the benefit of modern systems in Airbus. You don’t bolt on something to fix something. You can use the various redundant systems and change the software function to suit. Sure some things are hard wired, however the data is shared on one of the many redundant data bus’s.

Just ask the regulatory agencies if they accept all the ETOPS operations to go Single Engine? Why not? the engines are so reliable our days!!!!
Single Pilot Operations are the big push from the Industry (Airbus and Boeing) because they've realized it is impossible to crew the number of aircraft they are planning to sell. This was a study done pre pandemic but it is still valid.

Also final reserve fuel. What's the point? Haven't heard of anyone for years using any part thereof.

They’re after reserve crew. Making the 2nd pilot “reserve crew” cuts a lot of personnel costs up front, not only salary, and Cathay is a logical early adopter because they do a lot of 2-crew flights. All other LH operators are gonna watch this one very carefully... and of course, once this is done they will plan the next step, which is getting rid of the 2nd pilot, because the lifetime costs of a pilot (salary, retirement benefits, hotel, ground transport, etc) add up to a pretty penny and pilots can’t be traded or put up as collateral (yet) (Not valid in the sandbox)

Freight will go uncrewed first, until the first accident caused by an unexpected/unannounced GPS jam, which is something that’s already a problem, but unknown to the bean counters who get lost in their Lincoln Navigators. After that one, we’ll rest easy until at least 2100, by which time there’ll be demand for pilots on Mars, and all of us here will be either winged or horned (or reincarnated)...

Well one already has-Lufthansa and they have already said no because it doesn't meet the safety requirements:

I think LH might know a thing or two about long haul flying and crew costs.

LH will roster a senior first officer as third crew member for long range cruise not take one less.

Which is all fantastic until they screw up. I have personally saved myself from an autopilot in an airliner trying to kill me twice in five years. Sure they're great most of the time but they are not infallible and need to be monitored. The other issue that really irks me is the way regulators treat automation screw ups. If a pilot tries to spear an airliner into the ground he gets called in for counseling and revaluation. If the automation does it then they just shrug their shoulders. I know of one particular known software issue which should have grounded the fleet until it was fixed but despite being reported the regulator didn't have the courage to do something and take on a manufacturer. They just sat there and watched the reports pile up until 6 months later a software update came out. This in my opinion is why automation is treated with almost God like reverence. If we saw some real data on how many times pilots save the day from automation killing them I think the attitude in the industry might change. However it looks like they are hell bent on proving the hard way why two pilots and good automation is the only real safe option.

A great description of why full automation is still years away. The drone industry is going to pioneer this for us and I fear is not going to be as smooth and accident-free as they are all claiming . The attitude of the regulator(s) is going to be interesting to watch ,

Enthusiasts for fully automated flight might like to review the multiple Bulletins issued by Boeing over the last decade or so regarding flight path “anomalies” in both VNAV and LNAV in RNP approaches in FMC updates U11.0 / U12.0 / U13.0 etc.
Fixed now , but a reminder ; there is no such thing as fault free software.

A drone crash database.
https://dronewars.net/drone-crash-database/

TukwillaFlyboy

The problem with this statement is it assumes that humans are fault free. The software assistance wouldn't be in the picture at all if pilot error was not the biggest cause for incidents/accidents. Human performance is highly variable during day and night. With the same individual on day to day basis, when not rested even if he was off any duty, country to country, region to region. Proper Selection and training is not possible even in the US as Atlas, Colgan accident prove. So it's question of finding which one is more dangerous and expensive. Are CAT III operations, RVSM possible without automation?

I personally have conducted RNP approaches with the software glitches I referred to above and watched the autopilot quietly drift off profile with no warning.
I am a big fan of RNP , GLS and GNSS etc.
But don’t tell me you are going to trust an autopilot to fly the 400 SLF, including your own family, throughout a whole flight sequence.
I’m certainly not.

Sitting in front as the PF Pierre-Cédric Bonin who crashed AF447 had his wife on board.
what difference did it make?

vilas

Consider that pilots, potentially, while bot error free, commit different kinds of errors and have the capability for creative problem solving. They are not preprogrammed computers limited to actions and reactions thought of ahead of time. Think of Sioux City, Hudson River, Gimli, etc.

For all the money presumably saved, consider the cost of a single catastrophic accident. When things are working normally, the automation can perform more efficiently and in some cases with higher performance. But not so win non normal circumstances.

I am trying to be the devil's advocate so the community understands how technology fares against them. Soux city was a great feat, Hudson, gimli is nothing compared with it. However for Sioux city problem a software to use thrust/weight and asymmetry is possible and was being developed but due low probability of repeat it wasn't pursued. In Hudson, gimli case providing data base of gliding range without power and an additional electric hydraulics for AP can be even better managed. All credit to Sullenberger for doing what he did but if you critically analyse except the decision to ditch rest of the things human error under pressure was evident. He wasn't maintaining proper speed and when his speed dropped 19kts below Vapp it triggered alpha protection which prevented stall but also hampered proper flare and aircraft impacted heavily. So automated Airbus protection did help. Computers can handle data base of similar or possible accidents and quickly provide the lone pilot with correct solutions, and with better automation execute itself. As against this Mangalore accident in 2010 (not the only one) despite good aircraft and perfect weather due to poor Piloting158 people lost their lives. These things will be taken in consideration.

Should you think having the humans role in the cockpit replaced with automation is a good idea you only need to review incidents and accidents caused by automation, A320 written off because software wouldn't allow the crew to flare for landing for example. Replacing pilot error in the cockpit is merely replacing it with human error on the ground by way of software coding/design errors, hardware design, inability to test all electronic failure modes etc Failure mode of the 777 upset in Oz had previously been deemed as not possible.

Currently, when one of the pilots takes a break, they put a senior crew member on the flightdeck, so there are always 2 up front...
So it appears this is already happening...

Yes that was true, that is a problem with an technologically outdated FMS system running legacy software with 14 plus versions of software updates and patches.......

As a general rule I suspect that for every automation induced error someone can find an equivalent pilot error to balance it and vice versa. It will never be possible to prevent all fatal crashes but I think most would agree that the goal is to reduce such crashes to as low as possible bearing in mind the economics and practicality of such interventions.

What automation offers is consistency. Over the years there have been many great saves by pilots but equally there have also been many failures by pilots, be that through fatigue, inattention, misunderstandings, unprofessional behaviour or even downright recklessness. I have no stats but I strongly suspect the failures vastly outweigh the saves. OTOH I am sure there are many automation failures that might have resulted in fatal crashes if it were not for the human pilots intervening. This means that any aircraft with a reduced number of pilots will need rigorous testing but also that rigorous controls and testing are needed over software updates.

We can be certain that at some point a single pilot or autonomous airliner will crash through programmer error. The acid test will be if the average number of fatal crashes per year is considerably reduced through further automation.

turbidus

Really? People still do that? Not required with a working camera system.

I can think of many examples where pilots have saved the day, many times having to intervene over the automation.

Can't think of one example where the automation saved the day and did the opposite of the above.

Now, with dreadfully low experienced, low paid, and under motivated pilots coming into the ranks, that muddies the waters.

Perhaps the best way to increase safety is to intervene with better training and standards.

To a pilot that statement is immediately obvious, to the accountants and business MBA,s that run the airlines, that statement is like showing a cross to a vampire. Their recent iteration of reducing crew costs was the introduction of the MPL. Thats where you took a zero hour candidate,gave them ten hours in an aircraft then 100 hours in a sim and plonked them in the RHS to act as a low paid, debt ridden, seat warmer. There is no technical advantage in single pilot cruise ops ( or the Airbus acronym Single Human In Technology), only a supposed cost one.

There is also no need to have two pilots in the cockpit during cruise, especially not if they are poorly trained cruise only pilots. My only concern is that the remaining pilot falls asleep due to boredom or needs to go to the bathroom.

Having less crew on long haul flights also means that each pilot will get more opportunities to take-off and land, which are the critical parts of the flight.

You have just contradicted your own statement. Thats at least two reasons why you need two pilots in the flight deck during all stages of flight.

So instead of doing 10 take-off and landings a year they get to do 12? Not sure that is going to substantially enhance the flying skills of long haul pilots.

There is also no need to have two pilots in the cockpit during cruise, especially not if they are poorly trained cruise only pilots. My only concern is that the remaining pilot falls asleep due to boredom or needs to go to the bathroom.

Having less crew on long haul flights also means that each pilot will get more opportunities to take-off and land, which are the critical parts of the flight.

There are too many incidents/accidents due to human failure to brow beat technological intervention. Two Bonafide pilots in front with headsets off and speakers off. Not an isolated incident. In one case fighters were scrambled. One taking rest, other on ipad dials altitude instead of resetting heading leading to altitude bust. Capt sleeps for hour plus, starts ATC delayed descent, 9000ft out of slot never realised(sleep inertia) lands deep out of badly unstable approach and go around after reversing kills 158 people in good weather. No! This will not do. Now only question is whether technology can do better. May be not present machines but more redundancies and reliability can do it. Or can humans be changed?