52 comments

[ 4.3 ms ] story [ 126 ms ] thread
(comment deleted)
Wow we are really making some stupid decisions because of the unrealistic impression people have about technology.
How does a single pilot flight allow the only pilot to visit the lav?

Putting a cabin crewmember alone in the cockpit seems like a terrible bet that no emergency (or even abnormal) will occur over the thousands of such occurrences.

That is not the idea. It's having full autonomy but still having one pilot because going from two to zero would be considered too much of a jump right now. The reason Airbus is pushing for one pilot is because they can do the full autonomy.

Next will be Traffic Controllers - "Airbus Unmanned Traffic Management" - https://www.airbus.com/en/innovation/autonomous-connected/ai...

> How does a single pilot flight allow the only pilot to visit the lav?

Autopilot becomes the first officer. When the pilot goes to the lavatory, they hand control to the autopilot. (Same in case of incapacitation.)

In practice, I assume the first flights to test this will be on short routes. The pilot needing to go to the bathroom would constitute an emergency.

> Same in case of incapacitation.

Though how tf would they land?

> how tf would they land?

Most planes—including many GA planes—can land themselves.

Yes the Cirrus vision even has a land at nearest airport mode that can be activated by a passenger in the back.

It's not exactly graceful though. It'll just announce its intentions on the emergency channel and picks the nearest runway it wants. You'd better have a really good reason for it. It's still better than activating the CAPS parachute though which will usually total the airframe and can cause injuries.

Most? Like more than half? I don’t think that’s correct, even with a flight crew that's not incapacitated, but vanishingly few can auto-land with an incapacitated crew.
Aircraft have been reliably landing themselves for decades now. Tom Scott did a video where he was coached through a simulated landing of an airliner with zero flying experience: https://www.youtube.com/watch?v=AbTDzPUDxqY

The technology is even to the point where drones can reliably land on aircraft carriers. US Navy fighter pilots now have software assistance for landing on carriers (Magic Carpet/Precision Landing Mode). What was once one of the most dangerous feats in aviation is now a routine and reliably safe operation. https://www.navair.navy.mil/news/Navys-latest-flight-control...

Until the 1970s, flight crews had four critcial roles: pilot, first officer, engineer and navigator [1]. Technology improved, namely on-board electronics and satellite-based communication and navigation, and the navigator became redundant.

Until the 1980s, flight crews had three critical roles: pilot, first officer and flight engineer [2]. Technology improved, namely sensors and onboard computing, and the flighe engineer became redundant.

I don't have a strong view on making redudant the FO. But we've eliminated roles before without compromising safety. For ETOPS-equivalent flights in good weather, it strikes me as fine--if the pilot is incapacitated the plane should be able to land itself at the nearest airport.

[1] https://en.wikipedia.org/wiki/Air_navigation

[2] https://en.wikipedia.org/wiki/Flight_engineer

"Single-Pilot Operations Are Under Increased Scrutiny" - https://aviationweek.com/air-transport/safety-ops-regulation...

Under some circumstances its already happening, in a way...."EASA is among many regulators that, under very specific guidelines, allow one pilot to take a nap in their seat while the other stays at the controls."

> Under some circumstances its already happening, in a way...."EASA is among many regulators that, under very specific guidelines, allow one pilot to take a nap in their seat while the other stays at the controls.'

For what it's worth, it's also the norm in general aviation.

A much higher accident rate also appears to be the norm in general aviation.
There is a massive difference between a pilot napping in the cockpit (and thus being available with a few seconds lag) and having no second pilot at all.

There is also a huge difference between going from 4->3 or 3->2 and 2->1. You're loosing any redundancy!

So many of the near disaster scenarios I watch/read about (I'm an avid watcher of both Mayday and Admiral Cloudberg) are only near disasters because of having two pilots (or sometimes more) in the cockpit at the time of the incident. I recall one the other day where the plane was only saved because they happened to have four qualified pilots in the cockpit that day. The captain, the FO, someone doing a checkride on the captain and someone doing an evaluation on the pilot doing the checkride. Unusual situation but it happens.

The problem, as I see it, is that there are people here optimizing for the average workload and the average workload deviates by orders of magnitude from the workload during emergencies.

In normal situations the pilots are fairly idle most of the flight (talking about long haul here, not regionals with 10 movements per day) and probably use 10% of their training. When it matters, that other 90% of their training is absolutely critical to saving lives.

Similar situation with flight attendants. 99% of flights they are (exceptional) waiters and waitresses, but, 1% of the time they have to draw upon little-to-never-used experience which is the difference between life and death.

Controlled rest is completely different. It is not already happening.
> it strikes me as fine--if the pilot is incapacitated the plane should be able to land itself at the nearest airport.

The proposal on the table is not to have only one pilot on the flight. It's to have only one on duty during cruise phase only while the other one rests. This way they can skip the full double crew requirement during long haul.

So during takeoff and landing the cockpit will still be fully staffed. And in an emergency the resting pilot can be called. Toilet breaks would be covered by the other pilot too. So the cockpit is never empty. Doesn't sound like such a bad idea imo. During cruise the plane is normally flying itself from waypoint to waypoint anyway. Though I wonder how much quality of rest you can get in such a scenario (having to be woken up for every toilet break etc)

There will also be a remote hands support center, I'm less confident about that idea but as long as it's extra and not counted on it might be a good option.

Sounds like shit to me.

16hr long haul and everytime the other guy needs to piss you get woken up. Then you both have to fly the approach completely cooked from the shit rest. And thats without wx or diversions. Clearly not thought out.

True, it's not ideal.

I wonder though if there were similar complaints when the flight engineers were phased out, which obviously worked out fine (things are actually a lot safer back then).

For example they could do things like pee tubes (With M/F suitable private receptacles) which some private jets have in the cockpit. So it's not necessary to wake up the other pilot all the time. But yeah a long bumpy flight would be exhausting for both.

How do you prevent the pilot from crashing it like that GermanWings flight?
This is the setup that would prevent that scenario. Out of flight envelope, the second pilot, ( the automation ) would take over.
and if the human pilot needs to do emergency maneuvers that the computer doesn’t agree with?
If the pilot disagrees with the computer, surely the pilot must be wrong, there is no way the engineers made fatal mistakes writing software. Software does not have bugs. Look how safe MCAS was.
Note that more than a few airbus accidents happened because the pilots didn’t trust the computer. The reports usually end in; ‘if they had switched to automatic…’
> If the pilot disagrees with the computer, surely the pilot must be wrong

Empirically speaking, this is the case. Many crashes would have been avoided had autopilot flown. This is as true in air transport as general aviation.

… in normal conditions.

In other-than-normal conditions the computer is almost always wrong.

It can't, like its is already the case with the Airbus, where the computer wont allow maneuvers, that would compromise the integrity of the air frame due to excessive G's.
Only in Normal Law though.

These protections are increasingly lost as you move into alternate laws 1 and 2. And no protections whatsoever in direct law.

In most emergencies issues pile up so quickly that these modes tend to activate more often than not (and confusion about which mode is active has led to several fatal crashes)

Mentour pilot on YouTube has these cases and explains them well, including those different law modes.
Your comments prove you aren't in the industry.
You know what? I just had a chat with Spock here on the bridge....He suggests logic may be our best ally in this debate, not the sector of our Starship crew's employment...
How much money can this actually save airlines? Maybe a few thousand dollars per flight? Is it really worth it to the possibly hundreds of passengers on board? Who would want to save $10 on their ticket in exchange for removing all human redundancy from flight operations?
Upgrade to an extra pilot for $50 now!
Those ticket buying screens are going to need a lot of extra space: Click here to check-in your luggage, one extra selection box to be away from the emergency exit, extra $50 for not flying on a Boeing, $100 for a two pilot flight...
If only they put all the options on a single screen. That would risk the customer having any idea what they're buying. I'm just waiting for the day I click to the next screen only to learn I am purchasing a standing room ticket with the option of a seat for an extra fee.
Suddenly you have two windows open with booking.com, they open randomly during the flight selection. Then you train Google captcha to recognize American hydrants and get a ban for "acting like a bot". The airline compensates the losses by creating onboard lottery.
They would open more routes if not pilot restricted. Instead of training more pilots, they want to spread them thinner to maximize profits. It’s a bad idea.
These MBAs deserve eating toasts with strawberry lubricant instead of jam, spread thinnly of course.
> Maybe a few thousand dollars per flight? Is it really worth it to the possibly hundreds of passengers on board?

Empirically, yes, a lot of people will choose a €10 cheaper flight, particularly in Europe, where that might be 50% off.

The responsibility for the lives of hundreds on board will be further diluted. It's about profit extraction.
I would actually argue two things here.

#1, I think it's easier to go from 2 pilots -> 0 pilots. If your automation and remote control are good enough and robust enough to survive without a pilot, they are good enough to fly without a pilot.

#2, If you actually want to reduce crewing in a way which makes sense with less compromise on safety, allow long-haul flights to drop to a single pilot in the cabin during cruise phase. Long haul flights might fly with 4 or 6 pilots today, enabling to always have 2 pilots in the cockpit. It seems reasonable to me to reduce this number and allow only one pilot during cruise, where a second (or more likely more) pilot can join in the cabin within a minute or two in the event of any issue.

In the event of some issues. Physically moving to the seat could be impossible in time.

The a380 is certified to 853 people aboard, just pay the fucking skills. Skilled people are why an economy exists, paying people is the point.

I'm not sure that we really should ever do it, but, I think if you're going to consider doing it you should consider doing it in the least risky way possible (long haul flights during cruise phase).

Let's even start with cargo only flights.

I do think the learning would be valuable because I know eventually we will develop pilotless planes, just not sure it will happen in my lifetime.

Rather than announcing "is there a doctor onboard" they might be announcing "is there a pilot onboard"
i feel like quite a few folks in this thread greatly underestimate the things that may go wrong with automating flight controls. Just a recent example: in the past months, mainly due to Israeli military ops afaik, GPS spoofing (!not jamming!) experienced a significant uptick. If you spoof GPS, you can have all the redundancies in the world, they’ll all agree over the spoofed data. There are videos of resulting egpws misfires [1] already. Also, (enhanced) GPS is also used to recalibrate inertial navigation to keep the system as accurate as possible during the progression of the flight, just saying…

[1] https://youtu.be/-aExG6YUwFg?si=y6R26YdIXgtdyGdi