Pilotless Large Aircraft

Dreampilot said:
You will see a pilotless cargo plane in the next 5-10 years.
Click to expand...

Not trying to criticize you at all, but I'm wondering what credibility you have for making this statement. The majority of your posts indicate that you stopped flight training prior to obtaining a private pilot's certificate, which certainly doesn't give you any grounds to make an educated prediction concerning unmanned aircraft. Do you know what a charted visual approach is versus a precision approach, and how much fuel/time it saves to execute one versus the other? Are you aware that airline pilots make "grey matter" decisions - that a computer could never have a "yes/no" answer for - on a weekly basis? What is your experience with weather radar, and how many times have you been told by center that there's an area of moderate to extreme precipitation on your route of flight, only to look out the window and see blue skies ahead and the aforementioned weather below you? Have you ever had a flight attendant call you and inform you that there's a heroin addict in the lav with a needle stuck in his stomach? How about the cargo pilots who discovered stowaways on their aircraft? How is a computer going to ascertain if they're seeking asylum or trying to use the aircraft for an act of terrorism?

ALL of the situations above NEED human intervention and require extreme attention to detail. MOST of the situations above don't even involve the act of flying an aircraft, which I'd argue has ALWAYS been a SMALL portion of a professional pilot's responsibilities. When people ask this question, they fail to realize that stick and rudder skills, although critical to a well rounded aviator, is just one facet of the aviation industry. The litigious, policy-driven environment we operate in requires an aviator's skill, a legal mind, and a customer service attitude.
 
The question posed to one of the human factors gurus, Earl Weiner, was about automation and airplanes. Weiner said, "The question is not the role of man in the automated airplane because there is none. The question should be the role of automation in the manned aircraft." Just because you can automate does not mean you should automate.

Increased capabilities introduce often unexpected consequences. For example, when engineers introduced retractable landing gear, they also introduced the ability to land gear up. So with each new advance comes a new potential for error and the more complex the system, the more arcane and probably unexpected the event.

Automation is ddd.. dumb, dutiful and dirty meaning it does what it is programmed to do, does it repeatedly and does it in environments where we don't want people. Armed UAVs are a good example. We can send them into high threat areas without the consequence of losing a pilot. But automated systems can not yet think outside the box and that is the conundrum. We all do fairly well with the routine and we can even cope with the blinking red light (well.. most of us) but it is the surprise factor that creates the problem. Something not foreseen or planned for. SURPISE!!!

Already we are seeing very highly automated machines that are single pilot. But as Weiner noted, the combination of the dry computer (the software, hardware, etc) and the 'wet computer' (human brain) still offers the greatest redundancy for the lowest cost. For this reason many airplanes that can be flown single pilot (Citations and the new VLJs) are not flown single pilot because 1) insurance 2) the cheap redundancy 3) and then there is workload.

It has been noted that automation in the cockpit reduces workload where it is already low (cruise) and often increases workload where it it already high (programming FMS for approach changes). And automation often obscures the evolution of a problem until the problem is no longer manageable. This was demonstrated in the Air China 747-SP that had an engine roll back, slowed.. autopilot cut loose and the crew went for Mr. Toad's wild ride (see http://tinyurl.com/2mlu6y) for full report.

Fully automated pilotless airplanes in the near future? The question is more WHY than anything else.
 
Orange Anchor said:
The question posed to one of the human factors gurus, Earl Weiner, was about automation and airplanes. Weiner said, "The question is not the role of man in the automated airplane because there is none. The question should be the role of automation in the manned aircraft." Just because you can automate does not mean you should automate.

Increased capabilities introduce often unexpected consequences. For example, when engineers introduced retractable landing gear, they also introduced the ability to land gear up. So with each new advance comes a new potential for error and the more complex the system, the more arcane and probably unexpected the event.

Automation is ddd.. dumb, dutiful and dirty meaning it does what it is programmed to do, does it repeatedly and does it in environments where we don't want people. Armed UAVs are a good example. We can send them into high threat areas without the consequence of losing a pilot. But automated systems can not yet think outside the box and that is the conundrum. We all do fairly well with the routine and we can even cope with the blinking red light (well.. most of us) but it is the surprise factor that creates the problem. Something not foreseen or planned for. SURPISE!!!

Already we are seeing very highly automated machines that are single pilot. But as Weiner noted, the combination of the dry computer (the software, hardware, etc) and the 'wet computer' (human brain) still offers the greatest redundancy for the lowest cost. For this reason many airplanes that can be flown single pilot (Citations and the new VLJs) are not flown single pilot because 1) insurance 2) the cheap redundancy 3) and then there is workload.

It has been noted that automation in the cockpit reduces workload where it is already low (cruise) and often increases workload where it it already high (programming FMS for approach changes). And automation often obscures the evolution of a problem until the problem is no longer manageable. This was demonstrated in the Air China 747-SP that had an engine roll back, slowed.. autopilot cut loose and the crew went for Mr. Toad's wild ride (see http://tinyurl.com/2mlu6y) for full report.

Fully automated pilotless airplanes in the near future? The question is more WHY than anything else.
Click to expand...


"The National Transportation Safety Board determines that the probable cause of
this accident was the captain's preoccupation with an inflight malfunction and his failure
to monitor properly the airplane's flight instruments which resulted in his losing control of
the airplane.

Contributing to the accident was the captain's over-reliance on the autopilot
after the loss of thrust on the No. 4 engine."

Another case of the pilot being wrongfully blamed. The only way this would have been prevented is if he was holding the yoke himself for the entire flight. They have some balls don't they? What if there WAS no pilot on board, and the automated system was ALL that was available? Can a computer land a structurally damaged plane? It wouldn't have had to because it would have run straight into the ground anyway...
 
Dan64456 said:
"The National Transportation Safety Board determines that the probable cause of
this accident was the captain's preoccupation with an inflight malfunction and his failure
to monitor properly the airplane's flight instruments which resulted in his losing control of
the airplane.

Contributing to the accident was the captain's over-reliance on the autopilot
after the loss of thrust on the No. 4 engine."

Another case of the pilot being wrongfully blamed. The only way this would have been prevented is if he was holding the yoke himself for the entire flight. They have some balls don't they? What if there WAS no pilot on board, and the automated system was ALL that was available? Can a computer land a structurally damaged plane? It wouldn't have had to because it would have run straight into the ground anyway...
Click to expand...
I would advocate that the pilot should always have his hands on the controls. I come from different training than the rest of you. When you are at 200' AGL and the alt hold quits, it's nice to be a nanosecond away from shutting the autopilot off. I also always have my hands on the controls as PNF in any critical phase of flight - different training different thought process. I know that many out there won't agree with this, it's not taught (to my limited knowledge) by flight schools.

I would submit (without the actual proof) that if a computer is required to fly a F-16/F-18 then it most likely could land a plane with a structural defect.

Having sat on my fair share of FNAB boards in the Navy (Fleet Naval Aviator Evaluation Board) that in multiplace aircraft, it's headwork that accounts for most pilots loosing their wings (not flying ability). And a computer can not do headwork yet, for that reason alone, they will keep the man in the cockpit.
 
with pilots being paid so low, why would they ever get rid of them? before they get rid of the pilots I think they are going to make them pilots/flight attendants first and fire all flight attendants :) that way pilot can set up the airplane for a flight and then start serving coffee :D so you only need a crew of two instead of 6 or 8 :D at least until computers get 100% reliable...then...well, you get the picture :D
anyway, jokes aside, I still think they would need someone on the ground controlling those pilotless airplanes.
maybe some airplaneless pilot like me? ;)
 
UAVs are controlled by humans, but they use the UAVs in places where they don't want to risk the pilot. If you have a bunch of pax in the back, then most likely even if fully automated, there is going to be a pilot up front.
 
Dan64456 said:
"The National Transportation Safety Board determines that the probable cause of this accident was the captain's preoccupation with an inflight malfunction and his failure
to monitor properly the airplane's flight instruments which resulted in his losing control of the airplane.
Click to expand...

First, the NTSB doesn't tell you why an accident happened. It tells you what happened as in probable cause. Cause is not found. It is constructed.

And there were cues but they were masked. The autopilot controlled the 747 until it reached its authority limits and BAM.. it let's go. The airplane begins rolling and lights and bells are going on. SURPRISE!!. Yes, I know they are supposed to monitor but studies have proven time and time again humans are POOR monitors. So... let's take a crew, put them in a dark, mostly quiet environment.. some low frequency vibration in cruise.. night.. been on duty for a while and let's let an event slowly evolve until BAM!

So, what is the answer? Don't be distracted. That is like telling someone 'tell me when you are asleep." ??? If you are distracted you probably are NOT aware you are distracted.

The accident happened because he lost situational awareness.
How do you know.
He crashed.
Why did he crash?
He lost situational awareness.

That is a circular argument and goes nowhere to explain what happened and why.
What if there WAS no pilot on board, and the automated system was ALL that was available? Can a computer land a structurally damaged plane? It wouldn't have had to because it would have run straight into the ground anyway...
Click to expand...

It would still get classified as some form of human error. The system was designed by humans. The system was constructed by humans.

'Human error' is jargon used by lots of people who have little understanding or training in human factors. It explains nothing. It says nothing. It does nothing to prevent it from happening again.
 
The whole of the London's Docklands Light Railway is driverless, but still has a conductor to open and close the doors and check tickets - why? well simply because otherwise the train would set off with someone's arm hanging out a door. Why haven't they been able to adopt this system across the whole railway network? Simple - the DLR has a few drivers that "teach" the computer the route. Since it's on rails and receives signalling info via electric circuits in the track no driver is needed.

Now let's look at a plane - divert around weather, descend early, slow the approach, put in "S"-turns etc, etc - it's too unpredictable to possibly be computer programmed 100% successfully. Remember the old saying "A computer is only as good as the data you give it" in other words insufficient data means wrong decision. How many crashes have been avioded or made less severe by a pilot's experience and reactions to an unexpected situation?
MMMM...well just a few that spring to mind - that UAL flight that crash landed at Sioux City, with hydraulics lost, steering by varying right and left engine thrust; the BA777 that glided into LHR because the "computer" shut off the engines. The Transat Airbus that made it into some island in the Azores on TransAtlantic crossing when there was a freak fuel leak.
You could program the computer for these and some other freak accident will occur that will need a skilled pilot to resolve the situation (relatively) safely.

It ain't happening any time soon.
 
Pilotless aircraft wont happen anytime soon. But I do think the 2 man crew will be eventually replaced with a 1 man crew, and that it will happen in my life-time.
 
It seems like a couple of times a year we discuss this.

IMO you won't see it anytime soon, at least with passengers. One, cost is a big issue. Chances are you will not only need very advanced computers onboard, but also more advanced equipment on the ground to aid in approaches, redundancy etc. Liablility would be huge too.
Two, there are soooo many variables in the equation to consider. I believe the variables associated with 100% automated flight are as complex as the variables that create and drive weather systems. For all of our technology we still cannot completely predict exactly what the weather will do until it is about to happen. Sure we know it might rain or get windy, but the details are always hard to get perfect. Why? Because there are so many things that affect the outcome. (Think chaos theory. . . thank you Jurassic Park.) You would need computing power that could deal with tons of possibilities. Any pilot knows that no flight is the same and the need to adapt and improvise are daily issues. If flight always took place in a completely predictable vacuum, so to speak, I think it might be fairly easy to make "robot" pilots work. But, the computer processing power to safely manage all the potential variables of a safe flight exists only in the imaginary world of movies. Think something like Star Wars, I Robot or the Terminator. I don't doubt that one day we may get there, but we are talking a long way off.

Can you imagine writing all the code for a program that would deal with every possibility and contigency that a commercial flight faces and THEN making a CPU that can accurately run the program?
 
ppragman said:
I doubt it. Most freight companies are primarily concerned with the bottom line, do you think they have the capital to put into an airplane that doesn't need pilots?
Click to expand...

If executives could find a positive ROI, they'd put the capital into such a project.
 
ppragman said:
Further, I predict that freight companies would lose money because of all of the freight style that goes on. People bust mins, and push it to get in, people go the extra mile, and cut corners and rush so that they can get done on time, computers don't push it, they aren't flexible, and they won't try to hurry to get things done quicker. They won't do things like cancel IFR and proceed in VFR, or other things without incredible programming. I don't see it happening.
Click to expand...

That is one of the scariest things I've read in awhile. I really hope that the culture being described above is not very prevelant our community. Being a "hero" isn't necessarily directly correlated to making it though the career unscathed.


Back on topic:

I've seen several articles on the robot freighter. They are looking at a large scale aircraft. Right now, there are very few companies that can afford large scale new freighters. A new 747-8 comes in around 290 mil a copy (if I remember right). How much would an equivelant size robo-machine cost? Sure most airports we fly to are on coasts, but they are in or near large cities. Without assurance that the asset wouldn't get lost in an accident, would you buy one and lease it to an operator? No one wants to own the first Unmanned Large Aircraft that wrecks into a city on takeoff and kills people on the ground.

As passengers, I doubt many would get in a tizzy if there is no wizard behind the curtain, as they want the cheapest direct non-stop from Binot to Orlando they can get. As people that have homes, the thought of a pilotless aircraft flying over their house would be terrifying.

From an ATC standpoint, aren't they still having restrictions around UAV activity?

Has anyone flown a new model of airplane? I've introduced two. The associated problems with teething a new type is enormous. I can't tell you the number of times the engineers at the aircraft factory scratched their heads, with a "Uh, it's not supposed to do that" look.

Of course the on-board expert argument has gone over ad nauseum, so I won't dive into that.

meritflyer is right. If they can make enough money off it, they will. The big stumbling block right now is how much money are they going to have to invest. As the economy is tumbling, spending gets cut. Especially discretionary spending. When times get better, the back burner stuff like this will be news again. However, our next generation of thinkers will find lots of stuff to "fix".

Will it ever happen? I'm sure it will. When? No idea. There's a lot of hurdles to pass.
 
Polar742 said:
That is one of the scariest things I've read in awhile. I really hope that the culture being described above is not very prevelant our community. Being a "hero" isn't necessarily directly correlated to making it though the career unscathed.

Back on topic:

I've seen several articles on the robot freighter. They are looking at a large scale aircraft. Right now, there are very few companies that can afford large scale new freighters. A new 747-8 comes in around 290 mil a copy (if I remember right). How much would an equivelant size robo-machine cost? Sure most airports we fly to are on coasts, but they are in or near large cities. Without assurance that the asset wouldn't get lost in an accident, would you buy one and lease it to an operator? No one wants to own the first Unmanned Large Aircraft that wrecks into a city on takeoff and kills people on the ground.

As passengers, I doubt many would get in a tizzy if there is no wizard behind the curtain, as they want the cheapest direct non-stop from Binot to Orlando they can get. As people that have homes, the thought of a pilotless aircraft flying over their house would be terrifying.

From an ATC standpoint, aren't they still having restrictions around UAV activity?

Has anyone flown a new model of airplane? I've introduced two. The associated problems with teething a new type is enormous. I can't tell you the number of times the engineers at the aircraft factory scratched their heads, with a "Uh, it's not supposed to do that" look.

Of course the on-board expert argument has gone over ad nauseum, so I won't dive into that.

meritflyer is right. If they can make enough money off it, they will. The big stumbling block right now is how much money are they going to have to invest. As the economy is tumbling, spending gets cut. Especially discretionary spending. When times get better, the back burner stuff like this will be news again. However, our next generation of thinkers will find lots of stuff to "fix".

Will it ever happen? I'm sure it will. When? No idea. There's a lot of hurdles to pass.
Click to expand...


You'd be surprised, there's a reason I would rather fly single engine piston than go back to the freight companies I know of up here. At my old job we did unspeakable aviation acts. The environment is not very conducive to pushing it either, and yet people do it.

Down in the states I don't know, but up here, its so prevalent its scary. The push from management at some outfits is obscene, and the percentage of pilot I know who won't hesitate to take a peak is pretty spooky.

Single engine piston we have to scud run often, but its better than busting mins.
 
Dan64456 said:
You'd probably be surprised, lots of people buy into that "Pilot Error" 99% bs with no question or research on the matter...
Click to expand...
The public may buy into the pilot error idea, but companies know that it's much easier to fire one crew after an accident rather than refitting the entire fleet after an accident involving a computer error. Just think back post 9/11 and how difficult they felt it would be to put "reinforced" doors on every flight deck. I think our jobs are safe.
 
It has already happened. The Flight Engineer was replaced with technology. The last airplane to be built requiring a flight engineer was the Boeing 747-300, which I don't think lasted very long. That was in the late '80's.

The day may come when technology will replace 1 pilot in the cockpit. But, for the foreseeable future, I don't think you will see pilotless airplanes.
 
You must log in or register to reply here.
Back
Top