Aviation regulators push for more automation so flights can be run by a single pilot
 

https://www.theregister.com/2022/11/21/pilot_single/

Well that won't happen any time soon. Less pilots can never offer a greater safety margin.

That would be fewer pilots - in several accidents I'm not sure that having two pilots was of benefit. These crashes seemed to depend heavily on the assumption by one pilot that the other pilot was doing some action and the other pilot doing the same, with neither doing the correct thing. PIA8303, for example, where the PIC had essentially committed to slamming it down, and the FO thought they were going around when he retracted the gear. Without the FO action it would have been significantly different.

What I would expect is a slight increase in the number of flights where the lone pilot fell asleep.

Overall it doesn't seem to me like a large positive action to reduce the number of pilots in a plane, regardless of special cases where it wasn't. It's very difficult to determine when things went right only because there were two and I'm not keen to find that number out by having those accidents with one.

I agree that there are examples of more than one not being helpful or not working together as intended. But the entire aviation safety system is based on redundancy and in the cockpit where it matters most they want to go single pilot? There are many incidents where even three persons could barely handle what was going on. Whenever things go crazy more than one is needed. A cabin attendant to help won't be enough.

I fully agree with you but caution the usage of English in such matters.

"less pilots" means with none. ie 'without pilots'. I'm sure you didn't intend that.

"Fewer pilots" would mean less than x - when usually x=2....

What's wrong with no (0) pilots? We are already crashing aeroplanes semi regularly with 2-3 pilots on board, so long as the accident rate (level of safety) remains constant what difference does it make?

Unless you are advocating for a zero accident rate, which is an impossible dream for crewed operations,.....why are we setting the safety bar higher for uncrewed ops? (or less crewed ops)

Alpha

There may well be cases where having differing opinions between flight crew members resulted in a crash/hull loss/etc, but that doesn't mean that we should never have two pilots. People are always quick to point out where pilots have made errors, but what we don't hear about is how many times having two pilots *avoided* an accident - ie where one made an error that was caught by the other. (The same argument applies for wanting pilotless aircraft - where the aircraft couldn't handle a situation/suffered failures/etc and one or more of the pilots saved the day). A bit like the old 'where do we need armour on our warplanes?' story. People are quick to say 'where all the bullet holes are!" or "pilots cause (X) number of crashes" but the reality may well be 'where the bullet holes *aren't*" - 'pilots prevent (Y) times MORE crashes'

My point wasn't about pilots making errors. They do and they will continue to do so.

The point was, if one pilot, or no pilots, can be shown to have the same error rate than multi pilot, then what difference does it make?

Or is the safety bar higher for less crewed/uncrewed ops than it is for multi crew operations? If it is, why?

Alpha

Very wrong reasoning to think only about accidents that were created by pilots while omitting all the accidents that didn't happen thanks to pilots.
The latter are far more frequent.
Automation bugs happen really frequently.

Automation is the biggest business in town at the moment. Whether it be picking, packing and delivering your latest Amazon purchase, non-driver Uber cabs or drone technology.
I think it's inevitable as the power of computers and AI increases that humans are taken out of the loop. Why not aircraft? We already have autonomous trains.
I think I read somewhere that USAAF drones fly with an almost perfect safety record.
The Mars probes are effectively autonomous due to the delay between here and Earth and then there is the latest Moon lander tests.

Would I be happy to fly on a pilotless aircraft? No. But then people thought they would die of asphyxiation on a train.
Eventually attitudes change.

Skynet anyone?

This needs to be called for what it is. It's about money & distraction.

Money. While the regulators may be calling for more automation, it'll be the airlines calling for less pilots. The airlines know single pilots ops "will never fly", the regulators and public opinion will see to that. What they do want is less pilots per flight. Instead of augmented crews on long haul flights, the airlines want a 2 pilot crew with both pilots on the deck for take-off & landing - the cruise will be single pilot while the other is in the CRC. There's a potential 33%-50% saving in crew costs on all long haul flights. Job done, back slaps & bonuses all round.
Distraction. While the media will give this oxygen, there'll be push-back from the punters and the airlines will suddenly "listen to the people" and announce they are ditching the single pilot idea and that there will always be 2 pilots in the cockpit, what they won't say is that there will only ever be 2 pilots on the aircraft, irrespective of sector length.

This will happen. Maybe not soon, but it is the next logical step. The step after will be pilotless aircraft. Again not soon, as in our lifetimes, but it will happen. It won’t take more than a generation or 2 (up to 50 years) for public perception to switch to trusting automation over manned crew. Once they are fully used to everything in their lives being automated. There are already over 1400 self driving cars in the US. Over the next 20 years I see that being the majority of all cars.

There will still be accidents, but I think it will be the same or slightly lower than crewed aircraft by the time it happens. It will start with cargo, then the military, then commercial.

Envoy in the USA just had a flight out of ORD where the captain had a heart attack just on rotation. The line check pilot flew a wide pattern and landed back in ORD. Sadly, the captain did not make it. My condolences to the family. This is why you have 2 pilots at the controls.

Instead of single pilot cockpits it would make more sense to pair a two man cockpit with unmanned "slaves", like an unmanned freighter, that follows in automated formation distance (maybe in goose formation style for drag avoidance?) and that can be remotely controlled by the manned master aircraft and handed over to some drone landing team close to destination. Sort of the best of both worlds but not the worst like one pilot only for everybody.
Convincing the flying public will be a major hurdle.

It’s already being done. Airbus delivery flights from Blagnac for more then one carrier are trialing the Airbus single pilot concept.

And here's another good reason why we should never have single pilot cockpits. I doubt automation would help in this situation.

Captain suffers Heart Attack on takeoff | Emergency Return to O'Hare

Would you care to elaborate what you mean please?

My questions are these:

*if there is one pilot and the rest of the automated systems are designed by humans, do we really free ourselves from human error? Have a look at the QF72 for example.

*further to that, if the PIC is responsible under the legislation can they rid themselves of the automation if deemed necessary? If not do we charge the software designers, engineers, risk assessors and ultimately boards of the companies that will inevitably fail us?

Full delivery crew available but only one pilot and the airbus observer present in the cockpit at any one time. For the cruise only so far. And not over the PRC because of terrain etc and sh@t ATC. Aircraft controlled via datalink essentially from Toulouse. Early days yet but it is certainly being trialed on delivery flights. And doesn’t take a genius to work out the airlines involved.

The reasoning is balanced when you consider all the accidents that didnt happen thanks to technology and systems (TCAS/TAWS,etc). Its not just pilots that prevent accidents. And for further consideration, what about those accidents where to avoid an accident the pilot was simply following a direction from a technology/system...the same system that could have avoided the accident if it were left to do so.

Not to mention, the numerous accidents where the pilots made a concious decision to ignore what the technology/system was advising them to do....

​​​​​​​Alpha

When all is normal, a single pilot might be able to manage the flight without incident. But the level of safety, and especially the ability to cope with non-normal conditions is greatly reduced. One need only look to the safety level of general aviation. This level of safety must not be accepted for pt 135/121 operations.

Who in their right mind would want to do this job by themselves??

The social aspect (and salary…still, hehe) is to a great deal the reason why I am still in it. The Covid pandemic opened a whole lot of new doors and flying isn’t a given for me anymore. Can’t imagine I am the only one…..

The majority of the flying public only care about price - if its cheap enough, then whether 3,2,1 or 0 up front sadly won't matter.

The majority of the flying public seems to have a fear of flying and will finally pick the option it considers the safest. Imagine electrical fires or similar and then the remote control or AI left having to take over.

If we would want to abandon redundancy as a concept we can go single engine, single generator, single hydraulics and so on. Making aircraft less complex, more lightweight and tickets cheaper? And who would need that costly airport fire brigade?

Whatever it takes to keep labour costs low and falling. Makes you wonder who the government is representing, the people or the corporations? Okay, there's no wondering... it's not the people.

With single pilot ops the officials would be saying something a lot worse.

There is no credible single pilot operation of any large airplane. Without the redundancy of a second pilot, full automation is the only alternative. The first fully automated flight will be the first with a single pilot.

The reason why single pilot ops won’t be around for the foreseeable future is incapacitation. Whilst increased use of automation may make a single pilot airliner a reasonable prospect you then have to make it fully automated in case of that pilot being incapacitated. So you basically need a far greater level of automation nearing AI level. And the bigger problem, you’ll also need uninterrupted and perfectly reliable ground to air communication in case of that incapacitation. There’s enough problems with VHF comms interruptions to prevent that from happening without a massive overhaul of the worldwide communications network.

Basically a minimum of 30 years before the first scheduled single pilot flight happens. Every single aircraft manufacturers are planning is two crew, and they’ll be in operation for at least 25-30 years.

Pilots are their own worst enemies. You will have pilots losing their minds about how cool this plane is that can taxi, take off and fly in cruise by itself then wonder why their conditions are in the gutter. By the time this happens the people they pioneered this will be retired on their fat pensions.

The lawyers and insurers are going to make a lot of money.

"Less pilots can never offer a greater safety margin."

that's clearly not true historically - as the number of people on the flight deck has reduced safety rates have increased - of course it due to better higher tech and more automation enabling a reduction - not the other way round - but no-one would suggest INCREASING the numbers of n the flight deck would they?

And yes the tech occasionally goes wrong - but again probably not as often as humans

“We” are NOT crashing regularly! Safest couple of years in aviation history. But yes, sometimes but very very seldom accidents are caused by pilot error. Much more often accidents are prevented by pilot action! From my own experience as a pilot I wouldn’t putt my family in a pilotless aircraft. I have seen too many automation hick ups for that.

The aircraft in question are already highly autonomous with Auto TCAS and Auto emergency descent. Autonomous taxi trials are underway at Toulouse, and these flights have already been conducted, with the only real pilot input being altering the FCU, and a TCAS and visual lookout. Except as stated through Russian airspace ( when we used it ) due to lack of CPDLC and the PRC, due to terrain and paranoid standards of control that can’t help but interfere all the time. Albeit with a full crew available and rotated through as required. However that is more of a requirement for the entire crew to witness the operation as opposed to fatigue.
As to VHF outages a fair part of the world has CPDLC coverage, satellite enabled. I can’t remember when I last had a prolonged comms outage, even in the Dark Continent.
Incapacitation is a factor. That’ll be down to the regulators and our excellent (!) medicals. Likely to get more stringent and more invasive. A bigger factor is the chance of a “Eurowings” type event and indeed the effect on the mental health of the pilots involved due to isolation etc. That said, carriers in the less enlightened parts of the world have been quite happy to have their crews confined to their hotel rooms on layover with little or no interaction for almost 2 years now, consigned to endless loop rostering, with the only way out to the greater community and your families to spend a further 14, or 21 days locked up in a hotel at the end of the loop. And endless cycles of pointless testing.
So, it is being actively looked at. With the initial aim to reduce crews from 4 to 3, 3 to 2 etc etc. A beancounters wet dream.
Convince yourself that this is 30 years away if you wish.

Military drones seem to have an unusually high cruise phase crash rate. Why is that happening? We should find out before airliners go unmanned.
https://dronewars.net/2019/06/09/acc...drone-crashes/

They still fail to understand the FO is a CP in training...

As long as they don't use tesla software.....

Presumably because there is a huge amount of software and hardware invested in downing drones ?

It's more like restreamed mechanics being unaware of icing and stalls when steep turning in high mountains and such.

Back to the medical data?
I'm sure someone knows what the rate of incapacitation due to medical reasons is. .........

I think it’s not so much the reduced redundancy, although that plays a part, it’s the human interaction with two or more crew members bouncing ideas off each other, generating solutions and monitoring the results. You read about pilot error in accident reports but you don’t read about how good CRM stopped an accident chain before it had a chance to develop.

The automation will have to become better, as we are at the unfortunate time where it’s reliable up to the point where things get really bad, then it dumps the whole mess on the (single) pilot. Much in the way of self-driving vehicles in their current state of evolution. If the sole pilot who is awake is in the toilet when something bad happens (there was an opportunity for a joke there but I resisted it), then the aircraft must be able to cope.

If a non-pilot goes flying in a light aircraft with a qualified pilot, they have to assume the chances of survival are not good if that pilot becomes incapacitated, but do it anyway for fun. Will several hundred commercial travellers see it in the same light?