how long

[schnitzel2k3]

Step 1: Convince paying passengers to hop on with no one up front.

Step 2: Developing security (I'm not talking steel doors) to keep people from hijacking the tech enroute.

To get this certified and implemented at the airline level will be decades, particularly after the Max software debacle. It will be interesting to see what the cost savings is, as I can only imagine this tech will be mondo expensive upfront until the Chinese figure a way to simplify or copy it.

[tsgarp]

Single pilot on a modern heavy jet sure, it's actually less workload than flying a Navajo SPIFR.

I've never flown a Navajo SPIFR but I have flown fast jets including really old ones with scattershot instrument panels, black ball attitude indicators, antiquated navigation radios and of course no autopilot SPIFR, and can say with some authority that your statement is sometimes true but usually false. It all has to do with task saturation and workload which, if you've flown transport category jets you know can be very, very busy. Improper workload management is one of the primary causes of errors.

From a practical standpoint of single pilot transport category aircraft...who is going to train that pilot and how?

Who monitors that pilot for errors?

If there's a problem what comes first, flying the plane or dealing with the problem?

I’ve flown high performance single turbines, transport category and light piston twins SPIFR. Transport category was by far the easiest. Automation, another body in the cockpit and plenty of power make life easy; Engine quits? Let the automation keep the climb-out while the other guy reads you the checklist of what switches to throw. Light twin SPIFR, no automation vacuum gyros, no auto-feather no wx radar and a half engine on each wing; Engine quits? Your day is about to become a bag of suck. Single engine high performance turbine? Engine quits? Pull the handle and send Martin-Baker a thank you note.

[leftoftrack]

Step 1: Convince paying passengers to hop on with no one up front.

Step 2: Developing security (I'm not talking steel doors) to keep people from hijacking the tech enroute.

To get this certified and implemented at the airline level will be decades, particularly after the Max software debacle. It will be interesting to see what the cost savings is, as I can only imagine this tech will be mondo expensive upfront until the Chinese figure a way to simplify or copy it.

step 1: offer pilots as an option for an extra $2

step 2: they can steal the software but can't put it all together with hardware.....whats the problem?

step 3: the Max demonstrated how easily the FAA can be bought, not sure why it would be an issue.

[Rockie]

Transport category was by far the easiest

It can be easy if you’re just sitting there like a passenger but then the other pilot has to do your job too.

[complexintentions]

Single pilot on a modern heavy jet sure, it's actually less workload than flying a Navajo SPIFR.

I've never flown a Navajo SPIFR but I have flown fast jets including really old ones with scattershot instrument panels, black ball attitude indicators, antiquated navigation radios and of course no autopilot SPIFR, and can say with some authority that your statement is sometimes true but usually false. It all has to do with task saturation and workload which, if you've flown transport category jets you know can be very, very busy. Improper workload management is one of the primary causes of errors.

From a practical standpoint of single pilot transport category aircraft...who is going to train that pilot and how?

Who monitors that pilot for errors?

If there's a problem what comes first, flying the plane or dealing with the problem?

I’ve flown high performance single turbines, transport category and light piston twins SPIFR. Transport category was by far the easiest. Automation, another body in the cockpit and plenty of power make life easy; Engine quits? Let the automation keep the climb-out while the other guy reads you the checklist of what switches to throw. Light twin SPIFR, no automation vacuum gyros, no auto-feather no wx radar and a half engine on each wing; Engine quits? Your day is about to become a bag of suck. Single engine high performance turbine? Engine quits? Pull the handle and send Martin-Baker a thank you note.

Sure, sure. Everything is easy until it isn't. You're too cool for school with the "Martin-Baker". Nuthin' to it.

I don't even bother putting up rebuttals to this type of thread of which there are already dozens of variations. It's pointless and I grew tired of debating against false equivalencies and straw men. I only check back now and then to see if it ever changes. Nup.

I will say, that even discussing automation replacing pilots when there are tens of thousands unemployed worldwide is hilarious.

[photofly]

Automation replacing pilots will fix the unemployed pilot problem very very quickly: they'll all go and be unemployed something else.

[trey kule]

I just read today in Flying that for $85k USD, you can have an autoland system installed in a TBM. New ones come with it standard.
But yep. There is no evolution in automation. Pilots will never be replaced. These systems on a small GA are there to enhance the pilot experience, not takeover when a pilot cant fly.

[Rockie]

Autoland is a simple, dumb system that follows an ILS signal, performs a flare maneuvre based on radalt then tracks the localizor while on the ground. Hardly intelligent, subject to strict environmental limits and needs to be watched like a hawk with the pilot spring loaded to do a go-around in the event of failures or inadequate performance. Not to mention it needs to be programmed to begin with after the pilot first decides conditions are suitable for its use.

It escapes me why people always point to autoland as the harbinger of pilotless passenger jets. It's been around for over 50 years and hasn't really gotten any better than when it was first introduced.

[leftoftrack]

Autoland is a simple, dumb system that follows an ILS signal, performs a flare maneuvre based on radalt then tracks the localizor while on the ground. Hardly intelligent, subject to strict environmental limits and needs to be watched like a hawk with the pilot spring loaded to do a go-around in the event of failures or inadequate performance. Not to mention it needs to be programmed to begin with after the pilot first decides conditions are suitable for its use.

It escapes me why people always point to autoland as the harbinger of pilotless passenger jets. It's been around for over 50 years and hasn't really gotten any better than when it was first introduced.

Or.......you haven't been paying attention, Pilotless autoland already exists as an emergency backup for an incapacitated pilot. though you do need someone to touch the big red button.

https://youtu.be/IcVuubU4BTU

[tsgarp]

Transport category was by far the easiest

It can be easy if you’re just sitting there like a passenger but then the other pilot has to do your job too.

Nah. I’ve carried weak FOs into high density IFR with some systems broken. Still easier than doing hand flown single pilot IFR with all system working.

[photofly]

It escapes me why people always point to autoland as the harbinger of pilotless passenger jets. It's been around for over 50 years and hasn't really gotten any better than when it was first introduced.

50 years ago, autoland couldn't download the weather, select the nearest airport with a suitable runway, choose the correct approach for an into-wind runway of suitable length, announce its intentions to ATC on the correct frequency, fly the approach and land on the centreline, in a crosswind.

At least I don't think it could. It can now.

[Rockie]

50 years ago, autoland couldn't download the weather, select the nearest airport with a suitable runway, choose the correct approach for an into-wind runway of suitable length, announce its intentions to ATC on the correct frequency,

Autoland doesn't do that. Some other unthinking programs may do that based on very limited logarithms, and the list of things it doesn't consider but need to be considered is very, very long. I'm curious, how does it announce its intentions on the correct frequency? What if ATC can't accommodate this inhuman voices' demands?

fly the approach and land on the centreline, in a crosswind.

Autoland does this up to certain wind limits, just like it did 50 years ago. It also does it exactly the same way - track the localizor and flare using the radalt. Every airplane I've flown does it exactly the same way with the exception of the military planes I've flown that didn't have it. If technology were developed reliably allowing pilots to see through precipitation and fog, that would provide far more capablility and safety and render autoland a poor second choice in poor weather and basically unnecessary.

[photofly]

Autoland doesn't do that.

"Autoland" does do that. If "Autoland" doesn't, but something else does, then the quibble is about the definition of "autoland" vs the definition of "something else", which is dull and I'm not going to go there. The point is, press a button, the plane flies to a suitable runway and lands safely, with no further interaction with the occupants.

I believe the system that Piper (in the Meridian M600) and others have now certified for installation is all GPS based, probably with a radio altimeter. It doesn't require a localizer.

It's an emergency system, for use when the pilot has been incapacitated. It uses a synthesized voice to announce a mayday and declare intentions to ATC, it doesn't request or wait for a clearance. For now. And for now, there's no way for ATC to transmit a clearance.

But that was true of weather information too, 50 years ago, and a computer that was capable of holding the information required to choose an airport and a runway 50 years ago was too big to fit in an airplane. Now, obviously, it fits in any GPS navigator.

The argument about whether autoland is better or no better than what was available 50 years ago isn't an argument at all. The point is that an autoland system with awesome performance, without a pilot, can be added to small GA airplanes at non-military prices, and that ground/space systems to support it already exist and are available to all. Things only more sophisticated, from here.

[valleyboy]

SPIFR - is not difficult or a high work load even if by chance the auto pilot quits working working -where is all this shit coming from - Damn

[trey kule]

Rockie.

Instead of dazzling us with your rather uninformed perspective, why don’t you take a few moments and google and read the article which describes what the TBM autoland will actually do. A bit more than what you are purporting in your self described wisdom.

And this is an $85K fix to a small GA plane. What’s the next step?
Take a step back (or, in your case maybe many) From your ego.RPVs and UAVs are already here. And time and effort will improve them , I would think.

There are some very smart people developing cars that are driverless. And I think , one day, they will succeed. Why would that not evolve into aircraft?

[mixturerich]

Since this is a public forum I’m gonna just say pilotless airplanes are the devil.

I don’t want to be a passenger drone “pilot” anytime soon.

[Rockie]

Rockie.

Instead of dazzling us with your rather uninformed perspective, why don’t you take a few moments and google and read the article which describes what the TBM autoland will actually do. A bit more than what you are purporting in your self described wisdom.

And this is an $85K fix to a small GA plane. What’s the next step?
Take a step back (or, in your case maybe many) From your ego.RPVs and UAVs are already here. And time and effort will improve them , I would think.

There are some very smart people developing cars that are driverless. And I think , one day, they will succeed. Why would that not evolve into aircraft?

It’s a more sophisticated parachute.

Pilotless autonomous drones exist and are great at flying missions where nobody is at risk on-board in sterile airspace cleansed of other traffic except for the kind that wants to shoot it down. Passenger aircraft on the other hand operate in very messy airspace chock full of countless things constantly getting in their way. And they break.

Look at the abnormal section of any large aircraft operating manual. Look at the MEL. Listen to a busy arrival or departure frequency, or enroute frequency during weather. Sit in a flight deck and watch a crew enter a busy terminal in poor weather, deal with runway changes and maybe a diversion and you’ll have some small idea why we are a very long way from pilotless passenger aircraft.

Not saying it won’t happen, I am saying it won’t happen until true artificial intelligence has been created. Even then I wouldn’t dream of getting on it unless it also possesses a strong survival instinct and is capable of feeling fear. That one thing alone keeps us out of more trouble than you realize.

Driverless cars? They are two dimensional, slow moving land machines following a hard path on the ground and when it breaks or in an emergency it will (you hope) stop. And that’s the key isn’t it? A vastly simpler machine doing a vastly simpler task in a vastly simpler and benign operating environment where everything in it can slow down and stop. I had this conversation 40 years ago with a truck driver and nothing has changed.

You can also shove your ego comment up your ass.

[lownslow]

All fragile egos and fear mongering aside, I believe getting to the point of operating pilotless airliners has a weird and maybe impassable middle step. Everyone talks about removing the FO next and just having one pilot to watch systems and override as required but I’m not sure I would trust every single pilot in the world to not fall asleep at the switch or to not have a shitty month and just Germanwings the thing into the side of a mountain. So now you have to go from a two pilot flight deck to zero pilots in one step and that would be an awfully hard sell. Bet you could charge quite the fare to sit in the first row with the windshield though.

Technologically, anyone who mentions the success of US military drones is of course right but as stated they fly in pretty sterile airspace. IIRC in China they’re starting to fly small autonomous cargo planes, modified PAC 750s (picture a Caravan but whoever built it lost the instructions). To save complexity they don’t seem to have the capacity to taxi, instead being towed to the departure runway then after landing they just shut down on the runway and get towed in. Of course the airspace is tightly controlled there so I doubt they even have provision for solving conflicts in flight. If you have low enough standards, the whole aftair ought to be robust enough to hop on as a passenger.

I wonder if there’s much provision for MEL on UAVs or if they need to be perfect for every launch. In the end it may be cheaper to just pay two pilots than to ground every drone-liner for every little snag.

[tailgunner]

I agree with Rockie.
To the posters who are using military UAV’s as examples, do not look at their respective attrition rates, it is fearfully high. It’s just that they are secretive, unmanned, and operating over politically sensitive areas, so they are not widely reported on. They have operational challenges, but perform their assignments well enough to usurp manned aircraft in certain areas and roles.
In civilian operating roles, a loss rate currently seen by military operations would be unacceptable, and severely unaffordable.
As Rockie points out, there are a thousand little decisions that occur in a pilot’s/ crew’s mind that positivity affect each flight. Something as simple as having the ground crew push slowly to avoid contamination, delaying engine starts, preventive deice, taxiway choice, taxi speeds, stopping to allow the FA’s to secure the cabin, returning to the gate when a passenger feels unwell, working through an MEL, proper use of the brakes to give a smooth taxi, requesting an alternative departure path, requesting a different runway for departure, flap selection to enhance safety, choosing to FLEX, taking an intersection TO, wake turbulence decisions, making appropriate PA’s to ease passenger worries and to explain delays etc. These are decisions that we make but probably don’t even realize we do.
Cheers

[Rockie]

If pilot training were a simple matter of pointing the airplane in the right direction with nothing standing in the way we could do it in a fraction of the time. Real life isn't anything like that though and a great deal of initial simulator time is spent training the crew to fly the aircraft safely while dealing with specific aircraft technical issues and faults. Towards the end LOFT scenarios are introduced with not only aircraft but weather, passenger and security issues to name a few. Also included are command type scenarios where a person's ability to solve problems requiring them to jump from A to Z without going through the rest of the alphabet are developed. Perhaps even violate an SOP, company policy or air regulation along the way to get the airplane on the ground safely.

That's before they get to the airplane where the situations are frequently even weirder than a training department can conjure up and lots of real live people are involved, not just an instructor or examiner sitting in the back of the sim taking score. We develop judgement and encourage pilots to trust it while making their decisions.

When a machine can pass not only a transition course but a command upgrade program, interact effectively with all the humans it'll have to (CRM in case anyone forgot), then demonstrate judgement and a healthy respect for all the things that can kill you in flight then I'll consider getting on an airplane flown by it unsupervised. Let me know when there's a machine that can do that.