Single Pilot Airliners?

[AirFrame]

They decided when it would launch, and under what conditions. But once it left, they had *no* control save the self destruct.

The point is that we have the technology to fly aircraft autonomously now. I recall reading that an airliner had done an entire flight autonomously, from pushback to shutdown, somewhere in Europe, a number of months ago. Taxiing included. I don't recall the exact details, it just showed up in my news feed.

The technology isn't advanced enough to be sure of 100% success, but we don't have that with the meat popsicles up front either. The accuracy and confidence in them will only increase over time, and eventually they'll replace us. It's no longer a question of if, but when. What is more likely to slow the advance, is politics/marketing... will the public be comfortable getting into a plane with no human pilot on board...

[altiplano]

Don't forget, GPS autolands are already here...

Not yet they aren't, but GLS CAT II/III autolands are on the way. However while the aircraft 3-dimensional position in space will be GPS LAAS derived the actual FLARE, RETARD and ROLLOUT engagement will still be triggered from pre-programmed Radalt values just like the current ILS system. The rollout itself will be guided by GPS position instead of localizer.

VERY cool...I can't wait.

They are here as in aircraft are capable is what I mean.

Implementation is inevitable.

[Rockie]

Implementation is inevitable.

Of course it is, work is underway this minute to do just that. But I fail to see the relevance to single or no pilot airplanes. While revolutionary GPS is a position determining device with amazing accuracy and that's it. It isn't a brain. It's a means of determining the position of the aircraft that existing means of guidance can use, but in terms of procedures not much different than current ILS's or other navaids. It does nothing to provide autonomy.

The point is that we have the technology to fly aircraft autonomously now. I recall reading that an airliner had done an entire flight autonomously, from pushback to shutdown, somewhere in Europe, a number of months ago. Taxiing included. I don't recall the exact details, it just showed up in my news feed.

Do you mean this?

https://www.theverge.com/2018/2/1/16961 ... est-flight

Do you fly airplanes AirFrame? If you do I recommend you do the same as Altiplano and take note of how often you have to not just intervene, but remove the autopilot completely from the picture because of dynamically changing circumstances. Out here in the real world away from sterile launch and recovery facilities it's rather messy and extremely busy. Automation will never cut it in the real world all by itself. It requires artificial intelligence equal to what we mere humans are capable of, and as I said when real AI is finally developed it will be infinitely more significant and consequential than just flying airplanes back and forth.

[B208]

It requires artificial intelligence equal to what we mere humans are capable of

Not really.

[Rockie]

It requires artificial intelligence equal to what we mere humans are capable of

Not really.

A well thought out and presented argument B208. Bravo.

[SuperchargedRS]

So having a 250hr FO, how's that any diffent than just being single pilot?

Frankly might be safer, just fly the plane vs baby sit Jr. in the right seat plus fly the plane.

[Rockie]

You properly train the 250 hour pilot. Seem obvious to me, and I’ll take him/her over an autopilot that knows and thinks nothing every time.

You have a very poor opinion of new pilots don’t you?

[AuxBatOn]

Rockie, what, in a modern airliner cockpit, cannot be managed by a single pilot?

[Rockie]

Rockie, what, in a modern airliner cockpit, cannot be managed by a single pilot?

Under ideal, benign circumstances...nothing.

At all times though we use the concept of “mutual support” where both pilots ensure the other pilot is not making a mistake. Most of our procedures are designed to ensure critical actions are correctly understood and carried out. Contrary to what some people think CRM is not about getting along with others, it’s using well established threat mitigation strategies to ensure the safest possible operation.

When things get very busy as they are wont to do out here in the real world one pilot flies and the other manages the flight while still backing each other up to prevent errors. Some places you go and some situations require both pilots operating at 100% until the engines are shut down, because most major airports are operating to capacity and beyond. Throw weather into the mix and other operation challenges at an airport and everybody’s workload increases.

Then there’s the times things go wrong with the airplane caused by hard equipment or computer failures. One pilot flies sometimes under difficult circumstances without the aid of the all powerful autopilot while the other works the problem and prepares the aircraft and crew for a safe landing.

An autopilot keeps the blue side up when it’s working properly, but it doesn’t recognize or self correct it’s own errors. It doesn’t think or do anything by itself. It doesn’t adapt when things start to go wrong...it just quits. It is not a source of information to aid in decision making. It cannot manage a flight or a crew. It can’t communicate with anyone. And most importantly it is not a back up to ensure the other pilot isn’t making an error.

At the risk of sounding like a cheerleader CRM doesn't just compensate for human weaknesses, it utilizes our strengths to achieve the safest outcome. Autopilots do not contribute at all to CRM except for relieving the PF of the need to be actually hands on in steady state situations, and as I've said many times the busier it is the busier the PF is in manipulating the autopilot to the point where it's easier to get rid of it altogether. I've posted a link below to a Flight Safety Foundation document on CRM and approach and landing accidents. It's only a few pages long and an easy read.

https://flightsafety.org/files/alar_bn2-2-crm.pdf

[Brantford Beech Boy]

Single Pilot Airliner.

Been around for decades....

BBB

[B208]

It requires artificial intelligence equal to what we mere humans are capable of

Not really.

A well thought out and presented argument B208. Bravo.

Rockie, well thought out and presented arguments have, in the past, been ineffective with you, (pearls before swine sort of thing). However, against my better judgement I will favour you once again with my wisdom. 99% of the time controlling a modern aircraft is simply implementing a response in reaction to pattern recognition. That is something computers are much better at than humans (there are already computer programs that provide highly accurate medical diagnoses). Granted, there are the 1% of cases, (like Sully and the Gimili glider), where a computer would not have saved the day. However, a computer most likely would not have allowed the Collgan or Air France to happen.

From a technological stance we are very close to having pilotless aircraft. The technology just needs a bit more time to prove its maturity and reliability. Same argument applies to single pilot airliners.

[AuxBatOn]

Rockie, what, in a modern airliner cockpit, cannot be managed by a single pilot?

Under ideal, benign circumstances...nothing.

At all times though we use the concept of “mutual support” where both pilots ensure the other pilot is not making a mistake. Most of our procedures are designed to ensure critical actions are correctly understood and carried out. Contrary to what some people think CRM is not about getting along with others, it’s using well established threat mitigation strategies to ensure the safest possible operation.

When things get very busy as they are wont to do out here in the real world one pilot flies and the other manages the flight while still backing each other up to prevent errors. Some places you go and some situations require both pilots operating at 100% until the engines are shut down, because most major airports are operating to capacity and beyond. Throw weather into the mix and other operation challenges at an airport and everybody’s workload increases.

Then there’s the times things go wrong with the airplane caused by hard equipment or computer failures. One pilot flies sometimes under difficult circumstances without the aid of the all powerful autopilot while the other works the problem and prepares the aircraft and crew for a safe landing.

An autopilot keeps the blue side up when it’s working properly, but it doesn’t recognize or self correct it’s own errors. It doesn’t think or do anything by itself. It doesn’t adapt when things start to go wrong...it just quits. It is not a source of information to aid in decision making. It cannot manage a flight or a crew. It can’t communicate with anyone. And most importantly it is not a back up to ensure the other pilot isn’t making an error.

At the risk of sounding like a cheerleader CRM doesn't just compensate for human weaknesses, it utilizes our strengths to achieve the safest outcome. Autopilots do not contribute at all to CRM except for relieving the PF of the need to be actually hands on in steady state situations, and as I've said many times the busier it is the busier the PF is in manipulating the autopilot to the point where it's easier to get rid of it altogether. I've posted a link below to a Flight Safety Foundation document on CRM and approach and landing accidents. It's only a few pages long and an easy read.

https://flightsafety.org/files/alar_bn2-2-crm.pdf

In percentage, how often do you feel you NEED a second pilot in the cockpit?

[Rockie]

In percentage, how often do you feel you NEED a second pilot in the cockpit?

Depends on the phase of flight.

Preflight - 100%
Start/Taxi/Takeoff - 100%
Climb - 50%
Cruise - 10%
Descent - 80%
Approach and landing - 100%
Taxi in - 100%

That's in a normal routine flight with nothing unusual going on.

Remember that the primary goal is to capture, prevent and if necessary correct errors which occur prevalently during takeoff, approach and landing. Those are also the most demanding workload and situational awareness times especially if weather is a factor. When something is wrong with the airplane or there is some sort of passenger or crew issue underway the 2nd pilot is essential at all times.

Advocates of single or pilotless aircraft assume the airplane will fly itself...it won't...ever. They also assume nothing goes wrong with the airplane and that if it does a single pilot can handle it because the plane flies itself. See previous sentence.

To put it in terms you might better understand, do your odds of survival in a hostile zone increase, decrease or stay the same in a two-ship versus by yourself?

In my opinion pilots who think an autopilot can do their job need to find another line of work because they do not understand this one.

[lostaviator]

The plane being able to fly itself is just one piece of a very large puzzle. Airports will need to be upgraded to help these planes navigate the ground, air traffic control networks will need to be updated to allow them to essentially control aircraft movements, the public will have to be convinced. And then there's the whole part of terrorism.... If a plane has a direct link to someone on the ground, there is always a chance of unlawful interference. No one tries to hack USAF drones because, why would you? Add 100,200,300 lives to the mix, and ISIS militants everywhere will be having wet dreams.

[Rockie]

However, a computer most likely would not have allowed the Collgan or Air France to happen.

Strange you would cite these cases B208 as both aircraft had autopilots and computers that failed. In Colgan's case the wings were iced up and it exceeded the autopilot's capabilities so it disconnected. In Air France the sensors were iced up which removed valid information for those computers to act on, so they quit as well.

In both cases the only thing that could have saved those airplanes was a properly trained pilot.

[B208]

However, a computer most likely would not have allowed the Collgan or Air France to happen.

Strange you would cite these cases B208 as both aircraft had autopilots and computers that failed. In Colgan's case the wings were iced up and it exceeded the autopilot's capabilities so it disconnected. In Air France the sensors were iced up which removed valid information for those computers to act on, so they quit as well.

In both cases the only thing that could have saved those airplanes was a properly trained pilot.

In both those cases the pilots on board failed to recognize the pattern. A properly written algorithm would not have. My position stands.

[AuxBatOn]

You really need a second guy to start and taxi an airplane? Your exageration takes away from your credibility in this discussion.

[trey kule]

There were engineers who stated that man could not go over 60 mph because it would suck the air out of their lungs.
That man could not fly. That supersonic flight was not possible. That a trip to the moon........impossible.

There are pilots now that cannot see that advances in technology will see single pilot heavy cargo aircraft in their lifetime.

[Rockie]

You really need a second guy to start and taxi an airplane? Your exageration takes away from your credibility in this discussion.

With all due respect you're a little out of your lane making that kind of statement. Pre-flight is when the flight profile is programmed into the computer and performance calculations are made and entered among many other things happening at once demanding your attention in the back, in the front and around the aircraft. Backup and confirmation is essential to capturing errors and omissions as history has proven. Once the aircraft is moving on the ground whether under its own power or not is one of the highest threat times in the entire flight as history has also proven. Both pilots must be heads up or mistakes can and will happen, particularly at large, very busy airports with contaminated surfaces, de-icing and low visibilities, runway changes, taxi route changes etc. I've made them myself and speak from experience, not assumption. The worst airline disaster in history happened because of mistakes on the ground...remember?

You will learn this if you ever step out of your current job into mine.

The goal is a 100% safety record that is achieved deliberately, not through luck

[AuxBatOn]

So, you need someone IN the cockpit to do this? Wouldn't it be acceptable to have someone do these tasks remotely, like cross-checking the data you program? When do you physically use the second pilot in the cockpit, doing piloting tasks (not related to programming computers).

[Rockie]

So, you need someone IN the cockpit to do this? Wouldn't it be acceptable to have someone do these tasks remotely, like cross-checking the data you program? When do you physically use the second pilot in the cockpit, doing piloting tasks (not related to programming computers).

The second pilot is physically there doing something all the time, all calculations, programming and pre-flight checks must be monitored and confirmed for accuracy and omissions, and task sharing is essential if anybody wants to leave the gate on time. Even then it's not enough at times. Really AuxBatOn, you're just going to have to take the word of someone who actually does this for a living...or not. The choice is yours, but remember that the next time a Navajo driver tries to tell you how to do your job.

[AuxBatOn]

Rockie,

I work with people that check my work from a control station on the ground with telemetered data. I think I know what I am talking about. Doing things remotely is possible.

[altiplano]

Don't forget, GPS autolands are already here...

Not yet they aren't, but GLS CAT II/III autolands are on the way. However while the aircraft 3-dimensional position in space will be GPS LAAS derived the actual FLARE, RETARD and ROLLOUT engagement will still be triggered from pre-programmed Radalt values just like the current ILS system. The rollout itself will be guided by GPS position instead of localizer.

CAT 1 GLS approaches are already certified and demonstrated through autoland and rollout.

They also are limited to runways equipped to do them on. A pilot on the other hand is limited by the aircraft limits and can land on thousands of runways the autoland cannot.

My point was:
All those runways without an expensive ILS beam will be able to be autolanded through a single area augmentation antenna and a GLS GPS approach.

The disruption will happen quicker than we can imagine once it's gets rolling.

[Rockie]

Rockie,

I work with people that check my work from a control station on the ground with telemetered data. I think I know what I am talking about. Doing things remotely is possible.

Apples and oranges so no, you don't know what you're talking about. How about you refrain from telling me what my job entails and I won't tell you what yours does?

[sanjet]

It is just a matter of time. The only issue is the cost to implement this method. Right now it’s still cheaper to train two guys/gals to fly an airliners and avoid logistics for a 1 pilot/pilotless airliner. I think within 2 decades the infrastructure will start being implemented for single pilot ops.