f.pier
Posts: 1405
Joined: Fri Aug 25, 2000 2:50 am

Possibility To Remote Fly An Airliner

Fri Feb 10, 2006 3:53 am

All the most recent series of airliners are fly by wire. This means that there isn't a physical connection between commands and hydraulic and mechanical parts. Pilots give a digital command and computers "translate" it into a language comprehensible by automatic systems.
But this means that there is no need to have pilots physically in the plane.
I perfectly know that the presence of pilots is important and I wouldn't like to have "self controlled" airplane. But in extreme cases, for example the Helios 737 which crashed besause of the pilot missing, a remote system could be useful and probabily save a lot of lives.
A simulator connected with the true airplane should give the pilots all information he/she need to safely land the airplane.

Of course it should be "supercontrolled", for example this system should be activated only after authorization of the airplane manufacturer by special secret codes.

But this would exclude the possibility to hijack airplanes. What do you think?
 
RichardPrice
Posts: 4474
Joined: Sat Apr 23, 2005 5:12 am

RE: Possibility To Remote Fly An Airliner

Fri Feb 10, 2006 4:06 am

It would simply open up the possibility to hijack from the ground. Any computer security can be broken (yes, quote me on that, security is never meant to stop completely, just deter long enough that it doesnt matter) so having a total override system would mean the hijackers no longer have to risk themselves.

Plus you have other issues such as lag, latency, pilots being unable to 'feel' the aircraft etc.
 
FLALEFTY
Posts: 372
Joined: Tue Jan 31, 2006 11:33 am

RE: Possibility To Remote Fly An Airliner

Fri Feb 10, 2006 4:16 am

The short answer is "Yes" - especially the more modern, computer-controlled planes.

To do this would require installation of high-speed telemetry on the aircraft so all of the flight control parameters could be monitored and controlled in by a ground station. There would also be the need for a network of ground-based telemetry receivers/repeaters to allow continued control of the plane once it clears the horizon.
 
cbphoto
Posts: 1120
Joined: Sat Dec 20, 2003 6:23 am

RE: Possibility To Remote Fly An Airliner

Fri Feb 10, 2006 4:52 am

I have heard FedEx is looking at a freight system in the future that is completly pilotless! How far in the future is unknown!!
ETOPS: Engines Turning or Passengers Swimming
 
satx
Posts: 2771
Joined: Mon Apr 11, 2005 7:26 am

RE: Possibility To Remote Fly An Airliner

Fri Feb 10, 2006 5:13 am

As I understand it, the US military already has such capability with 'drone' aircraft and my guess is that we will eventually see remote controlled commercial aircraft as well, perhaps only a few decades away. I would expect the future costs of running an airline to play a big role in promoting remote control, and that freight aircraft will serve as a crucial link between military and passenger applications. I could see remote piloting technology first being used as backup for rare emergency situations that could result from both the captain and first officer being incapacitated. Perhaps the first such application will have built-in safety systems that require the cockpit to first hand-over control before the plane can be controlled from ground operators. This would obviously be a double-edged sword, but the fear people have of remote control might prevent full adoption in the early stages. Eventually the potential uses could be broadened to replace one of the pilots. Looking far off into the future of passenger aircraft, both pilots may eventually be removed from the cockpit after decades of acceptable performance is eventually achieved

If conventional hijacking continues to be a major problem in the near future, this may also play a role in speeding up the testing and eventual adoption of remote piloting technology.
Open Season on Consumer Protections is Just Around the Corner...
 
dartland
Posts: 514
Joined: Wed Apr 20, 2005 5:09 am

RE: Possibility To Remote Fly An Airliner

Fri Feb 10, 2006 5:22 am

This is a LOOOONG way off.

People are still complaining that they don't like it when a subway train moves without a human conductor (like Line 1 in the Paris Metro or the L train in NYC's subway) -- airplanes?

Although you can't refute that what seems crazy and impossible one day, often becomes reality down the road...
 
antiuser
Posts: 646
Joined: Thu May 27, 2004 9:43 am

RE: Possibility To Remote Fly An Airliner

Fri Feb 10, 2006 5:52 am

Current technology already allows it to be done. Pretty much all you'd need for a basic/rudimentary setup would be a series of servos and relays to control the yoke/stick, rudder, throttles and trim. However, as long as computers are fallible, there'll always be a human element in the cockpit.
The pilot's functions are likely to change, though. Basically the "pilot" would be someone who can operate the flight computer and know how to hand-fly the plane in case of computer failure - which, in some aspects, isn't so disparate from the way it currently is.
Azzurri Campioni del Mondo!
 
Bobster2
Posts: 1523
Joined: Mon Jan 17, 2005 3:04 am

RE: Possibility To Remote Fly An Airliner

Fri Feb 10, 2006 7:33 am

The NY Times had an article on this subject after 9/11. It was proposed that the pilots would initiate an automatic landing mode if they felt threatened by hijackers. All controls in the cockpit would then be immediately disabled and there would be no way to enable them again. Even if hijackers got into the cockpit, they could do nothing to prevent the plane from landing at the nearest airport.
"I tell you this, no eternal reward will forgive us now for wasting the dawn." Jim Morrison
 
N908AW
Posts: 864
Joined: Thu Aug 11, 2005 1:05 pm

RE: Possibility To Remote Fly An Airliner

Fri Feb 10, 2006 8:14 am

Quoting Bobster2 (Reply 7):
Even if hijackers got into the cockpit, they could do nothing to prevent the plane from landing at the nearest airport.

How safe would that be!?! The military is still working on making unmanned planes that make things go boom. Transporting 300+ people and baggage plus loads and loads of cargo unmanned? I don't think so. Insurance comanies right now are having a cow about the Eclipse 500's intent to design the plane for one pilot. Imagine zero.
'Cause you're on ATA again, and on ATA, you're on vacation!
 
bond007
Posts: 4423
Joined: Mon Mar 14, 2005 2:07 am

RE: Possibility To Remote Fly An Airliner

Fri Feb 10, 2006 8:18 am

Remember the 707 (I think) that they crashed in the desert to test some new fuel gel. It was all remotely controlled.

Of course, technically it's possible now...acceptable to pax....not for many decades.


Jimbo
I'd rather be on the ground wishing I was in the air, than in the air wishing I was on the ground!
 
satx
Posts: 2771
Joined: Mon Apr 11, 2005 7:26 am

RE: Possibility To Remote Fly An Airliner

Fri Feb 10, 2006 8:21 am

Quoting N908AW (Reply 8):
How safe would that be!?! The military is still working on making unmanned planes that make things go boom. Transporting 300+ people and baggage plus loads and loads of cargo unmanned?

It's not an all-or-nothing gambit. You don't have to get rid of the pilots just to try it out. This will be a very slow change that will start small and won't affect most commercial passenger aircraft for a very long time yet. However, I'm convinced that it will eventually happen. The potential cost savings are just too high to leave it on the drawing board forever.
Open Season on Consumer Protections is Just Around the Corner...
 
Bobster2
Posts: 1523
Joined: Mon Jan 17, 2005 3:04 am

RE: Possibility To Remote Fly An Airliner

Fri Feb 10, 2006 8:50 am

Quoting Bond007 (Reply 9):
Remember the 707 (I think) that they crashed in the desert to test some new fuel gel. It was all remotely controlled.

The remote pilots actually screwed up in that test or there was some hardware failure. They did manage to crash, anybody can do that, but they crashed in the wrong place. They were supposed to hit something on the ground that would tear open the fuel tanks, but they wobbled all over the place and lost control, and missed the target. If you watch the video, you can see that the plane was out of control at the end.

[Edited 2006-02-10 00:53:23]
"I tell you this, no eternal reward will forgive us now for wasting the dawn." Jim Morrison
 
HikesWithEyes
Posts: 623
Joined: Sat Jan 24, 2004 10:54 am

RE: Possibility To Remote Fly An Airliner

Fri Feb 10, 2006 12:17 pm

It will happen eventually. Once the systems prove to reliable, the
airline industry will be all over it.
You can say goodbye to ALPA and the hassles of CBA.
People have previously posted that the flying public won't accept it,
but a hundred years ago people were saying that they wouldn't go
up in an airplane, period.
As far as hijacking on the ground goes, it is always a possibility, but
the airline operations centers would get increased security and I believe
that most would have backup centers in case of power outages,disasters,
etc.
First, benzene in my Perrier, and now this!
 
masseybrown
Posts: 4444
Joined: Wed Dec 11, 2002 2:40 pm

RE: Possibility To Remote Fly An Airliner

Fri Feb 10, 2006 2:45 pm

I don't see pilotless passenger aircraft within 25 years; but it's easy to imagine a remotely controlled cargo flight with only one pilot carried "for emergencies" or union rules as the case may be. The freighter wouldn't be pilotless, of course; the pilot would merely stay on the ground and fly by a longer "wire."

The autopilot, data link, and telemetry equipment all exist. About two years ago, the USAF operated a Global Hawk from California to Australia and back, including fueling stops, without a pilot.
 
N766UA
Posts: 7843
Joined: Thu Jul 29, 1999 3:50 am

RE: Possibility To Remote Fly An Airliner

Fri Feb 10, 2006 2:47 pm

Quoting CBPhoto (Reply 3):
I have heard FedEx is looking at a freight system in the future that is completly pilotless! How far in the future is unknown!!

Screw FedEx then.
This Website Censors Me
 
bond007
Posts: 4423
Joined: Mon Mar 14, 2005 2:07 am

RE: Possibility To Remote Fly An Airliner

Fri Feb 10, 2006 9:34 pm

Quoting Bobster2 (Reply 11):
The remote pilots actually screwed up in that test or there was some hardware failure. They did manage to crash, anybody can do that, but they crashed in the wrong place

Well, this was hardly a good example...my fault. They didn't use much more sophistication than used on a remote control model aircraft. I guess it did prove that taking off and cruising is easy with little or no complex equipment. Also with autoland systems, which of course they didn't use, it would have been a different story - they were trying to crash the plane after all!


Jimbo
I'd rather be on the ground wishing I was in the air, than in the air wishing I was on the ground!
 
MD11Engineer
Posts: 13916
Joined: Sun Oct 26, 2003 5:25 am

RE: Possibility To Remote Fly An Airliner

Fri Feb 10, 2006 11:40 pm

How do you think the ex jet fighters converted to target drones are being flown? Even the 1950s Snark missile (an early nuclear cruise missile) could, if the mission was called off, be landed on skids using remote control. There is a huge skid runway iat Cape Canaveral built for this purpose when the USAF tested this missile back in the 1950s.

After 9/11 it was considered to install a "hijack" button on aircraft with FBW, which would disable the aircraft controls, only reversable by MX on the ground, and allow a ground station to take over the aircraft and land it, but it was considered to risky, that a terrorist organisation might hack into the system and hijack an aircraft to crash it without ever getting near it.

Jan
Je Suis Charlie et je suis Ahmet aussi
 
User avatar
yowza
Posts: 4281
Joined: Fri Jul 29, 2005 7:01 am

RE: Possibility To Remote Fly An Airliner

Fri Feb 10, 2006 11:45 pm

Remote pilots are an excellent idea. No point having pilots on board. If something tricky happens or control is lost all you need to do is find a passenger to fly the plane with instructions from the tower. From what I've seen in movies this is incredibly easy.  Yeah sure

YOWza
 
viv
Posts: 2953
Joined: Tue May 31, 2005 5:17 pm

RE: Possibility To Remote Fly An Airliner

Sat Feb 11, 2006 12:11 am

In the future, an airliner crew will consist of one man (or woman) and one dog.

The man's job will be to feed the dog. The dog's job will be to bite the man if he tries to touch the aircraft controls.
Nikon D700, Nikkor 80-400, Fuji X Pro 1, Fujinon 35 f/1.4, Fujinon 18 f/2
 
star_world
Posts: 943
Joined: Mon Jun 25, 2001 7:52 am

RE: Possibility To Remote Fly An Airliner

Sat Feb 11, 2006 12:21 am

Quoting N766UA (Reply 14):
Screw FedEx then.

 Confused Nice childish solution to that! Will you apply the same logic to all aircraft manufacturers that consider this approach? Do you refuse to ride on the unmanned monorails that you find in lots of airports?
 
N766UA
Posts: 7843
Joined: Thu Jul 29, 1999 3:50 am

RE: Possibility To Remote Fly An Airliner

Sat Feb 11, 2006 6:57 am

Quoting Star_world (Reply 19):
Nice childish solution to that! Will you apply the same logic to all aircraft manufacturers that consider this approach? Do you refuse to ride on the unmanned monorails that you find in lots of airports?

It's not a solution. I'm speaking as a pilot, not as a passenger. If I were a monorail operator I'd be pissed too.
This Website Censors Me
 
avgroupie
Posts: 13
Joined: Tue Sep 06, 2005 7:37 am

RE: Possibility To Remote Fly An Airliner

Sat Feb 11, 2006 7:41 am

Quoting Dartland (Reply 5):
People are still complaining that they don't like it when a subway train moves without a human conductor (like Line 1 in the Paris Metro or the L train in NYC's subway) -- airplanes?

In fact wasn't it the BART system in the San Francisco Bay area that had a fully automated train ( during start up testing - no passengers ) plow through the barriers at the final stop and end up in the parking lot ?

This level of commercial flight automation will probably occur, but not likely in my lifetime. If I am still around, I'd want at least a geek on board that could access the flight deck and execute "control/alt/delete" when the system no longer responds.  eyepopping 
 
fanofjets
Posts: 1979
Joined: Thu Apr 13, 2000 2:26 am

RE: Possibility To Remote Fly An Airliner

Sat Feb 11, 2006 1:14 pm

Quoting Bond007 (Reply 9):
Remember the 707 (I think) that they crashed in the desert to test some new fuel gel. It was all remotely controlled.

This was a one-time test of a fuel retardant. The aircraft referred to was a Boeing 720-061 (sometimes cited as a 720-027, as the airframe was purchased from the Braniff block of aircraft). The use of remote-control technology in this case is obvious.

Before:

View Large View Medium
Click here for bigger photo!

Photo © Marcus Anderholm



After:

View Large View Medium
Click here for bigger photo!

Photo © AirNikon

The aeroplane has unveiled for us the true face of the earth. -Antoine de Saint-Exupery
 
DeC
Posts: 535
Joined: Mon Aug 15, 2005 5:12 am

RE: Possibility To Remote Fly An Airliner

Sat Feb 11, 2006 1:25 pm

I would never fly with a remote-controlled airplane. Computers can do a lot of things but only guided, not by themselves; they keep making all those errors and all...imagine a plane run by Microsoft and getting an 'illegal operation - shutting down' error at FL360! Big grin
DEC
 
RDUDDJI
Posts: 1708
Joined: Fri Jun 04, 2004 4:42 am

RE: Possibility To Remote Fly An Airliner

Sat Feb 11, 2006 2:58 pm

I actually met some hippie who believed a conspiracy theory 4-5 years ago, that the CIA took remote control of the four planes on 9/11 and crashed them on purpose to further the West's agenda. It's sad what this world is coming to...
Sometimes we don't realize the good times when we're in them
 
ILCFII
Posts: 39
Joined: Sat Feb 11, 2006 12:31 am

RE: Possibility To Remote Fly An Airliner

Sat Feb 11, 2006 3:09 pm

This whole idea may not be all to far off. My school just got a grant to study UAV's.

www.aero.und.edu/f5_News/f1_Current%...News/filereader.php?id=012706a.xml

Rumor is that FedEx wants to be the first to fly a plane completely by remote control.
 
satx
Posts: 2771
Joined: Mon Apr 11, 2005 7:26 am

RE: Possibility To Remote Fly An Airliner

Sat Feb 11, 2006 3:15 pm

Quoting DeC (Reply 23):
I would never fly with a remote-controlled airplane. Computers can do a lot of things but only guided, not by themselves; they keep making all those errors and all...imagine a plane run by Microsoft and getting an 'illegal operation - shutting down' error at FL360!

Planes have more and more software in them with every new design, and yet they crash less than they used to. Just because the IFE goes on the fritz doesn't mean the rest of the aircraft is at risk. Personally, I don't mind the idea of an automated flight or a flight controlled by remote operators. As long as the technology is proven and the government doesn't shield the manufacturers and airlines from legitimate litigation, I'll fly it.
Open Season on Consumer Protections is Just Around the Corner...
 
Mir
Posts: 19107
Joined: Mon Jan 05, 2004 3:55 am

RE: Possibility To Remote Fly An Airliner

Sat Feb 11, 2006 9:21 pm

Quoting Star_world (Reply 19):
Will you apply the same logic to all aircraft manufacturers that consider this approach?

I certainly would.

Quoting Star_world (Reply 19):
Do you refuse to ride on the unmanned monorails that you find in lots of airports?

Big difference between those and a plane. You can have systems in a monorail that will bring it to a stop should a computer problem occur. That's not possible in an airplane.

-Mir
7 billion, one nation, imagination...it's a beautiful day
 
ImperialEagle
Posts: 2238
Joined: Sat Jan 14, 2006 10:53 am

RE: Possibility To Remote Fly An Airliner

Sun Feb 12, 2006 6:41 am

Quoting HikesWithEyes (Reply 12):
It will happen eventually. Once the systems prove to reliable

AND I can think of a couple of U.S. carriers, whose management would like to convert right now! Take the ones based in MSP or ATL for instance!

Personally, I'm looking for humans up front----no hair or gray hair is preferred. Wink
"If everything seems under control, you're just not going fast enough!"
 
star_world
Posts: 943
Joined: Mon Jun 25, 2001 7:52 am

RE: Possibility To Remote Fly An Airliner

Sun Feb 12, 2006 6:57 am

Quoting Mir (Reply 27):
Big difference between those and a plane. You can have systems in a monorail that will bring it to a stop should a computer problem occur. That's not possible in an airplane.

As other people have pointed out already in this thread, the onboard systems on aircraft are already advanced enough that they can operate to minima below what a pilot flying could. By no means am I saying that the technology is suitable today for fully automated flights, but given that most modern aircraft already have a computer layer sitting between the pilot inputs and the flight controls themselves it's only a matter of time.

I would actually argue that delivering automation on an aircraft would be less of a challenge than other transport types - such as cars for example. The margins for error there are much smaller when the concrete barrier at the edge of the road is less than a second away. In addition, the environment is already highly controlled - one automated aircraft could co-exist with many other 'traditional' ones; this would not work as well with cars where the other vehicles' movement cannot be anticipated as easily.

I'm not completely insensitive to pilots' views on the subject - if my job was threatened by advances in automation I'd hardly be very enthusiastic. But technology is always going to continue to progress, whether you like it or not. I'm sure horse carriage drivers weren't too pleased with the invention of the car.
 
planemaker
Posts: 5411
Joined: Mon Aug 25, 2003 12:53 pm

RE: Possibility To Remote Fly An Airliner

Sun Feb 12, 2006 10:16 am

Quoting N908AW (Reply 8):
I don't think so. Insurance comanies right now are having a cow about the Eclipse 500's intent to design the plane for one pilot.

Eclipse has had insurance for owners lined up for quite some time. It was one of the first things that they tackled... obviously.

Quoting Star_world (Reply 29):
By no means am I saying that the technology is suitable today for fully automated flights

The technology is already there... the Global Hawk flew from California to Australia completely autonomously with absolutely no human intervention from taxi out to landing. BTW, UAV flight is permitted in US commercial airspace (under enhanced seperation but it is permitted nevertheless already.)

Quoting Star_world (Reply 29):
but given that most modern aircraft already have a computer layer sitting between the pilot inputs and the flight controls themselves it's only a matter of time.

People who are really worried about computers crashing should NEVER fly Airbus or E-jets (or the 787 in the future) since there is no control link between pilot and any aerodynamic control surface.

Quoting Star_world (Reply 29):
would actually argue that delivering automation on an aircraft would be less of a challenge than other transport types - such as cars for example.

Honda has already been testing (for delivery in 2007) autonomous steering.

Quoting Star_world (Reply 29):
I'm not completely insensitive to pilots' views on the subject - if my job was threatened by advances in automation I'd hardly be very enthusiastic. But technology is always going to continue to progress, whether you like it or not. I'm sure horse carriage drivers weren't too pleased with the invention of the car.

Or elevator operators, as another example. The change to a crewless cockpit will be very gradual... just like the change from the very early 7 crew cockpit to the nnow standard 2-man crew. First there will be a change to just one pilot (computer operator)... and then there will be none.
Nationalism is an infantile disease. It is the measles of mankind. - A. Einstein
 
superhub
Posts: 398
Joined: Sat Jan 21, 2006 4:17 pm

RE: Possibility To Remote Fly An Airliner

Sun Feb 12, 2006 10:30 am

Quoting Dartland (Reply 5):
People are still complaining that they don't like it when a subway train moves without a human conductor (like Line 1 in the Paris Metro or the L train in NYC's subway) -- airplanes?



Quoting Avgroupie (Reply 21):
In fact wasn't it the BART system in the San Francisco Bay area that had a fully automated train ( during start up testing - no passengers ) plow through the barriers at the final stop and end up in the parking lot ?

The MTR and KCR in Hong Kong are all automated. However, a driver is still put in there just to ensure everything works properly.

I believe the Dockland Light Railway in London are driverless. There is a conductor whose job is to open/close doors and to switch the key to get the train moving.

Yes. The BART crash happened during testing. But now, all BART trains are automated, again with a driver to monitor if everything is ok. Also, drivers do drive trains when there are speed limits (especially if the central control system detects a ghost train)

Quoting F.pier (Thread starter):
I perfectly know that the presence of pilots is important and I wouldn't like to have "self controlled" airplane. But in extreme cases, for example the Helios 737 which crashed besause of the pilot missing, a remote system could be useful and probabily save a lot of lives.
A simulator connected with the true airplane should give the pilots all information he/she need to safely land the airplane.

Yes. It would probably have saved lives in the Helios crash. But I am not that comfortable about the controlled-on-the-ground system. Without pilots, what happens if the simulator on the ground gets overtaken by terrorists? The terrorists will just fly the plane, and those on the plane cannot do anything about it. Hence, there should be pilots on board to override the ground command.

But then if pilots on board can override the ground command, terrorists who hijack on board can do the same too.

You can argue that the military have pilot-less planes and there is no problem with terrorism. But the military has a totally-secure area for people on the ground to fly them. Even if terrorists try to take over, I am sure the army will neutralise them long before they can take over (after all, the army have plenty of guns). But if the ground control simulators are placed in, say, an airline headquarters, it will cost a lot to make the building as secure as an army base.
 
NAV20
Posts: 8453
Joined: Thu Nov 27, 2003 3:25 pm

RE: Possibility To Remote Fly An Airliner

Sun Feb 12, 2006 10:51 am

Quoting Star_world (Reply 29):
would actually argue that delivering automation on an aircraft would be less of a challenge than other transport types - such as cars for example. The margins for error there are much smaller when the concrete barrier at the edge of the road is less than a second away.

Have to disagree, Star_world. A car is 'positively-located' by virtue of the fact that it is running on a road - so the designer of any control systems has a fixed datum to start from. An aeroplane operates in a fluid environment in which it can (and does) move constantly up, down, and sideways as well as forward - so the designer would have to 'guess' where it might be at any one time. Aeroplanes also, of course, travel at speeds of the order of ten times those of cars. And I suppose it's worth mentioning that, unlike cars, if they 'stop', they crash..... .

Besides, surely it's a matter of priorities? If safety is the goal, why not start with automating cars, which are arguably the most dangerous form of travel on a seat-mile basis; as opposed to airliners, which are currently the safest?

Automating cars would be a lot simpler than automating airliners; largely a matter of installing a GPS-based guidance system, sensors to follow lane-guides installed in the roads, and further sensors to maintain safety distances and prevent collisions. And it would save literally hundreds, if not thousands, of lives daily, on a worldwide basis - starting tomorrow. The same can't be said for automating airliners.

[Edited 2006-02-12 02:56:28]
"Once you have flown, you will walk the earth with your eyes turned skywards.." - Leonardo da Vinci
 
planemaker
Posts: 5411
Joined: Mon Aug 25, 2003 12:53 pm

RE: Possibility To Remote Fly An Airliner

Sun Feb 12, 2006 6:22 pm

Quoting NAV20 (Reply 32):
Besides, surely it's a matter of priorities? If safety is the goal, why not start with automating cars, which are arguably the most dangerous form of travel on a seat-mile basis; as opposed to airliners, which are currently the safest?

"If safety is the goal..." - ah, life is never so simple. If vehicle safety was really a goal, then there are many things that could be legislated to reduce, overnight, the very high number of accidents that happen. Of course many drivers and most of the manufacturers wouldn't like it... but as you said, "it is a matter of priorities"... and car safety obviously just isn't one of them... really.

BTW, as I mentioned earlier, Honda has been testing autonomous steering for introduction in 2007. And Mercedes has already introduced auto braking in their new S Class model (via two radars - short and long) and auto cruise control.

Autonomous flight is already a reality with the Global Hawk. And much of today's commercial flying is already automated. So it is only a matter of time before it becomes 100% - at first with a pilot for passenger reassurance, and then without.
Nationalism is an infantile disease. It is the measles of mankind. - A. Einstein
 
Mir
Posts: 19107
Joined: Mon Jan 05, 2004 3:55 am

RE: Possibility To Remote Fly An Airliner

Sun Feb 12, 2006 6:47 pm

Quoting Planemaker (Reply 30):
People who are really worried about computers crashing should NEVER fly Airbus or E-jets (or the 787 in the future) since there is no control link between pilot and any aerodynamic control surface.

I believe the rudder is not FBW in the Airbus, and the ailerons are not in the E-Jet (or at least I think that's what a Shuttle America pilot told me).

-Mir
7 billion, one nation, imagination...it's a beautiful day
 
africanflyer
Posts: 13
Joined: Sat Feb 11, 2006 12:34 am

RE: Possibility To Remote Fly An Airliner

Sun Feb 12, 2006 6:58 pm

Quoting RichardPrice (Reply 1):
It would simply open up the possibility to hijack from the ground. Any computer security can be broken

Indeed. Reliance ONLY on computers is as bad if not worse than reliance ONLY on people.

Quoting SATX (Reply 4):
This would obviously be a double-edged sword

Therein lies the perennial problem.

Quoting Antiuser (Reply 6):
Current technology already allows it to be done

And as others pointed out, has since a long time.

Quoting Bobster2 (Reply 11):
If you watch the video, you can see that the plane was out of control at the end

Any link to this video please?

Quoting HikesWithEyes (Reply 12):
It will happen eventually

I agree, it's the trend of all things, efficiency, profit, safety.

Quoting MD11Engineer (Reply 16):
that a terrorist organisation might hack into the system and hijack an aircraft to crash it without ever getting near it

Already this has happened with more primitive navigation systems. In Southern Africa a simple false radio beacon planted by South African intelligence led to the crash of Mozambique's presidential plane.

Quoting Mir (Reply 27):
You can have systems in a monorail that will bring it to a stop should a computer problem occur. That's not possible in an airplane

I'm not so sure the equation is accurate. If a car simply stops on a highway that is likely to result in death. If a plane simply continues on a straight course, if nothing is noticed ahead or heading to that place, does seem the equivalent of "stopping in a safe place".

Quoting Star_world (Reply 29):
I would actually argue that delivering automation on an aircraft would be less of a challenge than other transport types - such as cars for example. The margins for error there are much smaller when the concrete barrier at the edge of the road is less than a second away. In addition, the environment is already highly controlled - one automated aircraft could co-exist with many other 'traditional' ones; this would not work as well with cars where the other vehicles' movement cannot be anticipated as easily.

Agreed, the margins are usually less with cars, once an aircraft is high up in the sky.

Quoting Superhub (Reply 31):
I believe the Dockland Light Railway in London are driverless. There is a conductor whose job is to open/close doors and to switch the key to get the train moving.

Yes, and also London's busiest and most reliable high speed Tube Line, the Victoria Line. The "driver" just sits there and does little else than open and shut the doors. The tube trains movement is remotely controlled from a central control station. The DLR goes very slowly, presumably as it is in a less controlled environment (outdoors) and also because people can see and know there is no driver, not even a conductor.

Quoting NAV20 (Reply 32):
An aeroplane operates in a fluid environment in which it can (and does) move constantly up, down, and sideways as well as forward - so the designer would have to 'guess' where it might be at any one time. Aeroplanes also, of course, travel at speeds of the order of ten times those of cars. And I suppose it's worth mentioning that, unlike cars, if they 'stop', they crash.....

Not sure that is really a good comparison. Depending on an aircrafts position, this can be determined with accurace and be even minutes away from any potential crash, so the speed here does not matter when comparing to the seconds it takes for a car.
 
Mir
Posts: 19107
Joined: Mon Jan 05, 2004 3:55 am

RE: Possibility To Remote Fly An Airliner

Sun Feb 12, 2006 7:27 pm

Quoting Africanflyer (Reply 35):
I'm not so sure the equation is accurate. If a car simply stops on a highway that is likely to result in death. If a plane simply continues on a straight course, if nothing is noticed ahead or heading to that place, does seem the equivalent of "stopping in a safe place".

A plane can't just pull over to the side of the road when things go awry. It has to be brought down to earth eventually, and in a controlled manner. Otherwise death is certainly going to result.

Something else to consider: if the planes are being remotely controlled from the ground, how is the delay going to impact the ability to hand fly? Not every approach can be flown down to the ground by the autopilot. Not every approach is an ILS. It would appear that visual approaches wound be impossible, and that would require spacing on approaches to be wider, which would not mix well with an increase in the number of planes in the air (i.e. delays).

-Mir
7 billion, one nation, imagination...it's a beautiful day
 
NAV20
Posts: 8453
Joined: Thu Nov 27, 2003 3:25 pm

RE: Possibility To Remote Fly An Airliner

Sun Feb 12, 2006 10:00 pm

Quoting Planemaker (Reply 33):
Autonomous flight is already a reality with the Global Hawk. And much of today's commercial flying is already automated. So it is only a matter of time before it becomes 100%

Glad we agree about cars being 'a suitable case for treatment' in the automation field, Planemaker (but they'd better leave MY car alone!  Smile).

The significance of innovations like Predator and Global Hawk are greatly exaggerated. Predator flies at only about 90 knots and Global Hawk at about 350. Both crash fairly regularly. The military don't worry much about crashes; their priority is to avoid human casualties from enemy action, and they tend not to worry too much about what happens to the odd bit of unmanned hardware, however expensive it may be. But you would expect airlines operating aircraft full of paying passengers to feel the need to be a lot more careful.

Besides, a neighbour of mine (radar specialist) was in the RAAF at the time of the Global Hawk trip here, and he said that over a hundred Australian servicemen were mobilised for the event - and God knows how many Americans came over as well. The thing had to be escorted by accompanying aircraft every mile of the way, and minutely controlled by a series of ground-stations while it was over land; and of course it could only land at a designated airfield, which had been specially beefed up to the equivalent of better than Category 3 capability for the occasion; and checked, double-checked, and triple-checked.

As far as I can tell, so far, you could only replace the present 'tariff' of two pilots travelling with the aeroplane with literally dozens of ground-based and air-based controllers per aeroplane. And they would mostly be working on information which, due to the need for satellite relay, would be anything up to two minutes out of date. The thought of such a system being in operation at Heathrow or Kennedy or even Sydney, with up to 20 aircraft at any one time stacked for miles around flying holding patterns and clamouring for landing clearance, beggars the imagination - even if the weather was good.  Smile

In addition, on the basis of my friend's recollection of the interior of Global Hawk, you could maybe design an automated airliner that would 'work' after a fashion. But the fuselage, like those of Global Hawk and Predator, would be so bung-full of satellite dishes, servos, transmitters, aerial arrays, relays, and the like that there wouldn't be any room left for passengers.......
"Once you have flown, you will walk the earth with your eyes turned skywards.." - Leonardo da Vinci
 
wukka
Posts: 884
Joined: Wed Jan 26, 2005 1:08 pm

RE: Possibility To Remote Fly An Airliner

Sun Feb 12, 2006 10:33 pm

Quoting Antiuser (Reply 6):
Pretty much all you'd need for a basic/rudimentary setup would be a series of servos and relays to control the yoke/stick, rudder, throttles and trim.

Most of that is already there in a FBW ship. As was mentioned above, I would think that the most difficult part of the whole thing would be the high-speed telemetry involved.

A one second delay in "pilotless" command response in a crosswind at a 50 ft. RA reading could potentially put your bogies down in the dirt somewhere between 18L and 18R... not something that I'd be comfortable flying on.

...but this is coming from a guy that watches computers go ass-out for a living and has an inate distrust for CAT III and "full autoland" because of it. Again, as was mentioned above, I'd rather have the bald or greying guy up front making the ultimate decision at crunch-time.
We can agree to disagree.
 
planemaker
Posts: 5411
Joined: Mon Aug 25, 2003 12:53 pm

RE: Possibility To Remote Fly An Airliner

Mon Feb 13, 2006 4:17 am

Quoting NAV20 (Reply 37):
In addition, on the basis of my friend's recollection of the interior of Global Hawk, you could maybe design an automated airliner that would 'work' after a fashion. But the fuselage, like those of Global Hawk and Predator, would be so bung-full of satellite dishes, servos, transmitters, aerial arrays, relays, and the like that there wouldn't be any room left for passengers.......

Thank you for your thoughtful post. However, I have to disagree with the thrust of your neighbour's information and the conclusions that you have drawn from them.

Certainly the interior of the Global Hawk should be "bung-full" of equipment. It is, after all, a surveillance aircraft - and all that gear is fitted into a fuselage that is only 44-ft. long. But the navigation gear is only a very small fraction of the spy gear. The prime navigation and control system consists of only two KN-4072 INS/GPS systems.

Regarding realiability, you should recall that the Global Hawk has been certified by the FAA to use civilian air corridors in the United States with no advance notice.

And almost 2 years ago, Boeing's UCAV program had two X-45As taxi together and demonstrate coordinated flight. Also, another flight had the X-45 pick up several "SAM" radar signals and take completely autonomous evasive action.

More recently, in June 2005, NASA carried out a technology demonstration using six GA planes that landed in rapid succession under IFR without any ATC in the time it would normally take to land a single flight under current aviation rules.

Considering where the industry is already at now technologically, I really don't understand how people can doubt that there will be pilot-less commercial flight... even if its 20, 25 or 30 years from now. Of course it won't happen overnight. Current UAVs will be followed by UCAVs, then freighters and then commerical airliners. Of course, for a while during the transition there will be one pilot kept in the cockpit as "back-up" to calm down any nervous pax.
Nationalism is an infantile disease. It is the measles of mankind. - A. Einstein
 
NAV20
Posts: 8453
Joined: Thu Nov 27, 2003 3:25 pm

RE: Possibility To Remote Fly An Airliner

Mon Feb 13, 2006 7:27 pm

Thanks also for your constructive posts, Planemaker. Have to query this, though:-

Quoting Planemaker (Reply 39):
But the navigation gear is only a very small fraction of the spy gear. The prime navigation and control system consists of only two KN-4072 INS/GPS systems.

You'll know better than I that, in a conventional aeroplane, many systems are 'automated' but that, at present, the pilots set them. For example, the autopilot/autothrottle will fly a given course and profile at a given speed, but the pilots set those parameters by hand. The same goes for the ILS systems. In addition, the pilots also constantly monitor any number of instruments and displays.

In a pilotless airliner, the ground operators would need to be able to make all those settings remotely, and also 'see' all the instruments. That would require a huge amount of telemetry, and a large number of additional servos etc.; all of which would add cost, take up space, and add weight.

Therefore, if the object is to save money, any savings on the pilots' salaries and overheads would be negated as soon as the lost space and extra weight required even a single row of passengers to be left behind. That's without taking into account the much increased capital and revenue costs of installing and servicing all the extra equipment (and of course paying all the ground controllers).

In the area of safety, neither Global Hawk nor Predator yet approach even the safety standards of high-performance piloted military aircraft, leave alone those of civilian jetliners:-

"Hornburg put Predator’s crash rate last year at 32.8 per 100,000 flight hours, and this year’s at 49.6. “If you want to talk about Global Hawk, which we are measuring, the accident rate for the Global Hawk right now is 167.7. That is unsatisfactory.”

(It was interesting, to say the least, to learn that of the four Global Hawk prototypes that were lost (out of six), one crashed because it received a radio signal intended for another UAV hundreds of miles away, and another went down because of a 'typo' by a ground controller operating a computer keyboard!).

http://www.nationaldefensemagazine.o...sues/2003/May/Pentagon_Unhappy.htm

I can quite see why UAVs can be useful to the military (because they can venture into places where a manned aircraft would be at too great a risk of being shot down) but I don't think the systems are anywhere near safe enough to be used even for cargo aircraft at the present time.

The other thing that occurs to me is the obvious 'What if?' - what happens if something goes wrong? There were three emergencies just in the past year in which, had human pilots not been at the controls, serious accidents would probably have occurred; the A310 which lost its rudder en route from Cuba, the 747 which lost engine power and had to glide-land at Heathrow, and the A320 with the stuck nosewheel at LAX.

I don't see how ground operators could possibly have brought any of those three aeroplanes back safely.

[Edited 2006-02-13 11:35:12]
"Once you have flown, you will walk the earth with your eyes turned skywards.." - Leonardo da Vinci
 
planemaker
Posts: 5411
Joined: Mon Aug 25, 2003 12:53 pm

RE: Possibility To Remote Fly An Airliner

Tue Feb 14, 2006 3:24 am

NAV20, thank you again. I am enjoying our discussion.

Quoting NAV20 (Reply 40):
...but I don't think the systems are anywhere near safe enough to be used even for cargo aircraft at the present time.

I agree with you... at the present time. And "present time" is the key word. I have always said in 20, 25 or 30 years from now. When one considers the technology that we have now that is not yet operational, coupled with what is already coming down the "pipe"... and what is still to be developed, I believe that there can be no doubt about a pilot-less future.

Quoting NAV20 (Reply 40):
For example, the autopilot/autothrottle will fly a given course and profile at a given speed, but the pilots set those parameters by hand. The same goes for the ILS systems.

Yes, he sets them by hand... but most of the time through a keypad (FMS) that could easily be done remotely.

Quoting NAV20 (Reply 40):
In addition, the pilots also constantly monitor any number of instruments and displays.

"Constant" is no longer (some could argue that on longer sectors it hasn't been that way for quite some time - I've flown several times across the Atlantic in the cockpit jumpseat and the crew chatted away more than monitored.) The trend has increasingly be going to "set & forget" and the "black cockpit" where the pilots are given an aural warning and a text message on the EICAS only when there is an incident/anomaly. Otherwise the screen reamins "black" (unless the crew actually want to view the "system" or "instrument" on the EICAS and pull it up.)

Quoting NAV20 (Reply 40):
In a pilotless airliner, the ground operators would need to be able to make all those settings remotely, and also 'see' all the instruments.

Please remember that "all the instruments" are not really instruments but software code. And as mentioned earlier, the flight will be programmed into the FMS before take-off. Thus, a ground-based pilot would only have to monitor or send a command only if there is an incident/anomoly that the onboard systems haven't resolved.

Quoting NAV20 (Reply 40):
That would require a huge amount of telemetry, and a large number of additional servos etc.; all of which would add cost, take up space, and add weight.

Abolutely no additonal servos would be required. And telemetry is not a problem... it could quite easily be handled even today - we don't have to wait 20 years.

Quoting NAV20 (Reply 40):
Therefore, if the object is to save money...

Money is a factor, no doubt, but safety would be the first criteria. Pilot error is the number one cause of accidents/incidents. And accidents cost an airline a lot of money. As a side note, I am sure you are aware that the cost of maintaining a pilot roster is not just in their wage while flying.

Quoting NAV20 (Reply 40):
...any savings on the pilots' salaries and overheads would be negated as soon as the lost space and extra weight required even a single row of passengers to be left behind.



Quoting NAV20 (Reply 40):
That's without taking into account the much increased capital and revenue costs of installing and servicing all the extra equipment (and of course paying all the ground controllers).

As we have established, there will not be a requirement for any additional equipment or extra weight (in fact, the current gear just keeps getting smaller and lighter every year,) and an airline will gain space, if anything, because the cockpit will only require space for a single "compter operator." And on the ground, the airlines already have the army of "ground controllers" - the dispatchers and flight followers.

Quoting NAV20 (Reply 40):
had human pilots not been at the controls, serious accidents would probably have occurred;

Actuaries have to be very emotionless when analysing aircraft accidents for insurance companies, and unfortunately, we must be as well to have a factual discussion. And the facts are that pilot/human error is the cause of the majority of accidents/incidents.

Also, please remember that you are framing your points in the here and now... while we are actually talking about something that will happen in 20, 25 or 30 years from now.

And the switch from the current 2-man crew to no crew would not happen overnight. First there would be a switch to a single pilot and then none. And, of course, this will happen first in UCAVs, followed by cargo aircraft and then pax aircraft.
Nationalism is an infantile disease. It is the measles of mankind. - A. Einstein
 
NAV20
Posts: 8453
Joined: Thu Nov 27, 2003 3:25 pm

RE: Possibility To Remote Fly An Airliner

Tue Feb 14, 2006 7:52 pm

Enjoying it this end too, Planemaker!

Quoting Planemaker (Reply 41):
And "present time" is the key word. I have always said in 20, 25 or 30 years from now.

We’re obviously not far apart on the likely timescale. Must admit that I had a chuckle at that – remembering an economics lecturer who told me that we would cease to be dependent on fossil fuels within a similar timescale. That was a shade more than 30 years ago….  Smile

I’d accept your point that miniaturization of telemetry etc. has come a long way – but they’d have to find room for a satellite dish (maybe three, to provide ‘redundancy) and the motors to drive them. And if, as with Predator, each ground controller was set up in a full cockpit replica and provided with TV coverage of key views, all the instruments and panel displays would have to be available to them.

Generally the economics of any such exercise appear to be screwy to me. The most you could save per aeroplane would appear to be the salaries and overheads of two pilots, and say 500 pounds of weight. To achieve that, the aeroplane is going to require millions of dollars-worth of extra equipment and almost certainly carry fewer passengers; and on top of that you’d have to meet the capital cost of setting up all the ground-stations, and training and paying a large number of highly-skilled ground controllers.

Quoting Planemaker (Reply 41):
Pilot error is the number one cause of accidents/incidents.

With respect, that’s a bit of an over-simplification. Almost all accidents happen because of a combination of factors (called a ‘cascade’ by professional investigators). The primary cause of accidents is usually found to have been ‘pilot error’ in about 50% of cases. Other causes that frequently arise, and are sometimes found to be the primary causes, are weather, mechanical or structural failure, air traffic control mistakes, systems malfunctions, etc. I suspect that it rather suits all sides of the industry to play down causes other than 'pilot error’ in accidents, in their public pronouncements anyway. After all, the pilots are usually dead anyway – and to put undue emphasis on other factors which may have contributed would mean taking whole marques of aeroplane out of service while they were modified, closing whole airports until their air traffic control was brought up to standard, etc.

If you research individual accidents you’ll most often find that after the ‘primary’ cause was announced as ‘pilot error’, a whole raft of Service Bulletins, Airworthiness Directives, alterations to manuals etc. are brought out to put the other contributing causes right.

Besides, we hear a lot about ‘accidents’, and they are exhaustively investigated. We don’t hear nearly as much about ‘incidents’ – potential accidents which are avoided by alert and skilful piloting. Like the three incidents that I referred to above, all of which arose from mechanical or structural failures - no 'pilot error' involved. I remain convinced that any or all of those examples would probably have turned into fatal accidents if the aeroplanes concerned had been under ‘remote control’.

In fact (wild thought to finish off with  Smile) if we ever did ‘progress’ to fully-automated flight, we’d probably encounter a whole new class of accident. In the absence of any reports of developing problems from human pilots, without data from Cockpit Voice Recorders etc., and with the Flight Data Recorders just recording the 'systems' doing their various jobs, it would probably be next to impossible, in most cases, ever to find out what DID go wrong?
"Once you have flown, you will walk the earth with your eyes turned skywards.." - Leonardo da Vinci
 
planemaker
Posts: 5411
Joined: Mon Aug 25, 2003 12:53 pm

RE: Possibility To Remote Fly An Airliner

Wed Feb 15, 2006 7:08 am

Quoting NAV20 (Reply 42):
Must admit that I had a chuckle at that – remembering an economics lecturer who told me that we would cease to be dependent on fossil fuels within a similar timescale. That was a shade more than 30 years ago….

Yes, one can always get a chuckle from finite resource predictions. Some of my favourites are related with Thomas Malthus, Paul Ehrlich and the Club of Rome. According to that group, we shouldn't be here now.

Quoting NAV20 (Reply 42):
I’d accept your point that miniaturization of telemetry etc. has come a long way – but they’d have to find room for a satellite dish (maybe three, to provide ‘redundancy) and the motors to drive them.

I don't think that that will be a problem in 20+ years. Several airlines already offer their pax satellite TV or radio. And there is Connexion by Boeing, which offers passengers high-speed broadband internet connectivity via satellite.

Quoting NAV20 (Reply 42):
And if, as with Predator, each ground controller was set up in a full cockpit replica and provided with TV coverage of key views, all the instruments and panel displays would have to be available to them.

I guess that first we would have to discuss what a cockpit might look like in 20+ years. As I mentioned in my last post, I believe that if there even is a cockpit it will be very simple. No more than a couple of large monitors and perhaps a joystick and mouse. Who knows, "switches" and "controls" could very well be voice activated by then as they are now in some fighters. And, if not, certainly of a "touch screen" variety. Having internal and external TV coverage is no problem though it would not be required. Aircraft currently fly and land in zero-zero weather on autopilot.

Quoting NAV20 (Reply 42):
With respect, that’s a bit of an over-simplification. Almost all accidents happen because of a combination of factors (called a ‘cascade’ by professional investigators). The primary cause of accidents is usually found to have been ‘pilot error’ in about 50% of cases.

I agree with you that it is a bit of an over-simplification. However, to continue to put it simply, the reduction of that 50% is nevertheless a substantial improvement that would have a large economic impact.

Quoting NAV20 (Reply 42):
Generally the economics of any such exercise appear to be screwy to me. The most you could save per aeroplane would appear to be the salaries and overheads of two pilots, and say 500 pounds of weight.

Just imagine if just one airline didn't have their 6,000 pilots, and all the ancillary costs associated with them. I think that you could conservatively estimate the cost savings at over $2-billion per year. That would appear to be pretty compelling economics.
Nationalism is an infantile disease. It is the measles of mankind. - A. Einstein
 
User avatar
VirginFlyer
Posts: 3974
Joined: Sun Sep 10, 2000 12:27 pm

RE: Possibility To Remote Fly An Airliner

Wed Feb 15, 2006 8:35 am

This is a very interesting discussion. From reading the above posts, I think there are two different schools of thought being talked about. One is the remotely piloted aircraft, while the other is the completely autonomous aircraft (I believe this is the fundamental difference between the UAV and the UCAV).

Now, a remotely piloted aircraft, in my opinion, makes absolutely no sense for passenger aircraft. You still have to employ pilots, and they would still be bound by the same restrictions on duty time etc. The only financial savings would be in terms of accommodation at away stations. There would be definite risks in terms of what happens if the signal is lost (and this would be a problem, especially over the Pacific). It would be a very unnecessary over-complication of the system. The two primary reasons I have heard for having UAVs are that they allow the aircraft to conduct riskier missions without endangering the pilot's life, and they allow the aircraft's design to be greatly simplified (and reduced in weight) by removing the systems required to keep the pilot alive and accommodated. Quite obviously neither of these arguments applies to a passenger aircraft - they don't fly dangerous missions, and they need systems to keep the passengers alive. The same could be said for the vast majority of freight aircraft, which are pressurised, and which would not be risked on dangerous missions due to the value of the aircraft itself. Furthermore, because there is someone still flying the aircraft, there is no removal of the human factors element (other than issues such as hypoxia and the somatogravic illusion, neither of which should really come into play these days anyway if aircraft are maintained properly, and pilots are trained properly)

The other school of thought obviously is the completely autonomous aircraft. This obviously does remove the pilot completely, with the resultant cost savings. Given airline flying is basically point to point along predetermined routes, this wouldn't be that difficult on the face of it. However, the issue comes up of what happens when things have to change from this pre-determined set of rules, such as in an emergency situation

Quoting NAV20 (Reply 42):
potential accidents which are avoided by alert and skilful piloting.

I think NAV20 has struck the nail squarely on the head here. Yes, there are accidents caused by pilot error, and we hear a lot about it. However, there are also a lot of accidents prevented by pilot action. The fundamental problem with a computer is that it will quite happily go about doing its task as defined by a set of mathematical rules, but is not able to handle completely 'out there' situations.

An accident which leaps to mind straight away is United 232, the DC-10 which had a catastrophic failure of #2 (tail mounted) engine, and subsequently lost hydraulic power. The aircraft was only able to be brought back because of some lateral thinking by the pilots, who used the thrust settings on #1 and #3 engines to control the aircraft. The aircraft was kept in control almost to the runway, making a very heavy landing and breaking up on impact. There was loss of life, however nearly two thirds of those on board survived. Such a scenario was never considered by the engineers when they designed the aircraft. I think it would be safe to say that if the aircraft had been under completely automatic flight, it would not have been able to make this action by itself, and would have crashed with no survivors.

Now that we know about such a failure, the way to get around it could be programmed into computers (and indeed I believe it has, although it took some time to perfect). However, the real danger of 'unknown unknowns' still exists. In other applications, such as monorails, cars, etc, this can be acceptable. If things go completely out the window and the computer can't handle the situation, the vehicle can be brought to a stop pretty much as is. It would definitely be frustrating for the passengers, but it would not generally be life threatening. Unfortunately in an aircraft, you don't have the luxury of just stopping there and then because the computer doesn't have a way to handle the situation. Either the situation gets handled, or their aircraft crashes.

Computers as we know them today do not possess the capability to make lateral thinking decisions like this. At the end of the day, they are still simply electronic calculators, running along a pre-defined set of rules. They have become amazingly fast and powerful in recent years, but that does not change the fact that they are simply a mathematical apparatus. Perhaps one day we will have computers that can reason, and make logical jumps like humans. When that day comes however, we would then have to look at a much bigger picture than just 'pilotless aircraft' - it would be a world where humans are not needed (but let's leave that argument for The Matrix). I'd also be willing to bet that if computers were able to gain our human strengths - our abilities of lateral thinking and learning as things occur - that they would also pick up our weaknesses in terms of making incorrect decisions as a result of making the wrong lateral thought, or learning incorrectly. This leaves us right were we started, only it isn't Human Factors any more, it is Computer Factors. All that being said, I am quite sceptical that we will be able to produce computers who can think like us in my generation - it just seams a leap too far at this point in time.

Getting back on subject, where I believe the future for commercial aviation lies is in a greater integration of man and machine. We have two different resources available to us, both with their own strengths and weaknesses. Instead of casting one aside completely for the other, we should be trying to make the best use of both, in order to maximise the strengths, and negate the weaknesses. For this reason, I think taking the human out of the loop would be an absurd idea - every bit as absurd as suggesting we should remove all automation from aircraft, and fly them completely by hand, navigating completely by dead reckoning. What is needed is a sensible middle ground. We need to make use of computers to be able to handle the straightforward tasks much better than any human could. We need the human there to be able to handle the 'curve-ball' tasks much better than any computer could. Rather than trying to eliminate the pilot, the goal should be to improve the interface between pilot, computer and aircraft, so that they can work together even more effectively than what they already do.

V/F
"So powerful is the light of unity that it can illuminate the whole earth." - Bahá'u'lláh
 
bond007
Posts: 4423
Joined: Mon Mar 14, 2005 2:07 am

RE: Possibility To Remote Fly An Airliner

Wed Feb 15, 2006 8:56 am

Quoting VirginFlyer (Reply 44):
Now, a remotely piloted aircraft, in my opinion, makes absolutely no sense for passenger aircraft. You still have to employ pilots, and they would still be bound by the same restrictions on duty time etc.

Unless you had a team of 'pilots' on the ground that were monitoring the flights of perhaps many times the aircraft they could individually fly from the cockpit. One person could monitor many aircraft in cruise flight, and other available for individual control if a problem occurs.

Quoting VirginFlyer (Reply 44):
An accident which leaps to mind straight away is United 232, the DC-10 which had a catastrophic failure of #2 (tail mounted) engine, and subsequently lost hydraulic power. The aircraft was only able to be brought back because of some lateral thinking by the pilots, who used the thrust settings on #1 and #3 engines to control the aircraft.

Agreed, but since that accident, NASA and others have developed systems to control the aircraft with engine thrust only - this is probably best done by a computer and not a human ... not available for the DC10 mentioned though.

Quoting VirginFlyer (Reply 44):
Computers as we know them today do not possess the capability to make lateral thinking decisions like this

Yes, but given all possible data, I'd bet there are more accidents now because of incorrect lateral thinking by humans (not based on logic), than there would be if everything was handled in a mathematical logical way. Contrary to what some might think, 'seat of the pants' flying often doesn't work. That's why the first thing they teach you in instrument training is 'believe the instruments'.

Quoting NAV20 (Reply 42):
Other causes that frequently arise, and are sometimes found to be the primary causes, are weather, mechanical or structural failure, air traffic control mistakes, systems malfunctions,

I think you'll find that few accidents occur purely because of a structural failure, or system malfunction, that couldn't have been saved by pilot actions. In fact many incidents that are caused by system failures, end up being crashes because they are handled incorrectly by the pilot.

Remember what airline pilots do at EVERY step of the flight, including emergencies - they read a checklist and do what it says ... little more than a simple computer program  Smile

It probably wasn't that long ago that nobody believed we have trains that didn't have drivers!

Good discussion,

Jimbo
I'd rather be on the ground wishing I was in the air, than in the air wishing I was on the ground!
 
User avatar
VirginFlyer
Posts: 3974
Joined: Sun Sep 10, 2000 12:27 pm

RE: Possibility To Remote Fly An Airliner

Wed Feb 15, 2006 10:17 am

Quoting Bond007 (Reply 45):
One person could monitor many aircraft in cruise flight, and other available for individual control if a problem occurs.

Perhaps, but I don't see this being allowed - at the end of the day, regulators will demand that each aircraft has its own pilot - and for a very good reason. Can you imagine the outcry against the FAA (or any other regulator) if an aircraft were to crash because there were not enough 'operators' at the ground station to deal with simultaneous emergencies on multiple aircraft. It is an unlikely situation, but still entirely possible.

Quoting Bond007 (Reply 45):
NASA and others have developed systems to control the aircraft with engine thrust only

My point is, up until Sioux City happened, it was an 'unknown unknown' - no one had thought about the possibility of that happening, and how to handle it. It was a solution which was thought up on the spot, under pressure, based on the experience of the pilots. In my view, computers, as we know them today, are less able to handle an 'unknown unknown' than a human.

Quoting Bond007 (Reply 45):
Contrary to what some might think, 'seat of the pants' flying often doesn't work. That's why the first thing they teach you in instrument training is 'believe the instruments'.

I'm aware of that. Like I said, pilots need to be properly trained so that they don't fall victim to the somatogravic illusion. For those who aren't familiar with the term, it is when, without any visual reference, your body misinterprets a change in pitch as a change in speed, or vice verse - an increase in speed feels the same to the 'seat of the pants' as a pitch up. A King Air crashed near here a number of years ago at night. After take off, the pilot had thought the aircraft was climbing steeply, so he pushed forward on the controls. The sensation of steep climb increased. Before he could realise his mistake, he hit the ground. This can be avoided very easily - trust your instruments.

The point I was getting at wasn't that pilots should be flying by the seat of their pants (having flown at night, I know that is a silly suggestion), but that in an unusual situation, where an 'unknown unknown' presents itself, the best solution could well require lateral thinking (such as United 232), and in this situation the human will most likely provide a better outcome than the computer.

Quoting Bond007 (Reply 45):
Remember what airline pilots do at EVERY step of the flight, including emergencies - they read a checklist and do what it says ... little more than a simple computer program

That's a very good point, and one which I'd like to expand on. If the computer goes through all its checklists, and still can't solve the problem, then it doesn't have anything else to fall back on, and the plane crashes. If the pilot goes through all the checklists, and still can't solve the problem, he can still fall back on his experience and knowledge, and not just the programmed guidelines, in an attempt to find solution - again, the United 232 scenario.

Like I said, I'm not advocating turfing computers out of the aircraft and having all flying done by the seat of the pants - indeed, I think that is every bit as bad as taking the pilot out and having all flying done by the computer. What I am saying is that an aircraft should use both computers and humans to achieve the best possible outcome. We have two resources with differing abilities (and differing strengths and weaknesses) available to us. At the very least, we'd be foolish not to make full use of both of them; I'd go as far as to say it would be negligent.

V/F
"So powerful is the light of unity that it can illuminate the whole earth." - Bahá'u'lláh
 
planemaker
Posts: 5411
Joined: Mon Aug 25, 2003 12:53 pm

RE: Possibility To Remote Fly An Airliner

Wed Feb 15, 2006 5:32 pm

Quoting VirginFlyer (Reply 44):
The other school of thought obviously is the completely autonomous aircraft. This obviously does remove the pilot completely, with the resultant cost savings.

I believe that we'll end up with pilot-less commercial aircraft. However, as I have posted, I don't believe it will happen for at least 20+ years, and both pilots won't be pulled out of the cockpit overnight either.

Quoting VirginFlyer (Reply 46):
Can you imagine the outcry against the FAA (or any other regulator) if an aircraft were to crash because there were not enough 'operators' at the ground station to deal with simultaneous emergencies on multiple aircraft. It is an unlikely situation, but still entirely possible.

How many emergencies per day do airlines have today? And in 20+ years when aircraft and systems have even higher reliability?

Quoting VirginFlyer (Reply 46):
If the pilot goes through all the checklists, and still can't solve the problem, he can still fall back on his experience and knowledge, and not just the programmed guidelines, in an attempt to find solution - again, the United 232 scenario.

But just how experienced or knowledgeable is the pilot? Using the United 232 scenario, the plane would have been totally destroyed if it weren't for UAL senior training pilot and DC-10 expert Dennis E. Fitch, who was fortunately travelling as a passenger. I read that in subsequent reconstructions of the flight in simulators, no pilot of any seniority had succeeded in reproducing Fitch's achievement of maneuvering the aircraft as far as the runway.

Already, computers fly and land aircraft more precisely and consistently than pilots. In 20+ years, commercial aircraft will obviously be even safer than they are now. So it will be up to the actuaries to work out at what point the risk of having pilots (whose errors cause approx. 50% of primary accidents) versus not having pilots is reached. With the continuous improvement in aircraft design, reliability and safety, and the continuous advancement in computer processing power and software, a pilot in 20+ years will be a redundant cost. At the end of the day, it will be a financial decision.
Nationalism is an infantile disease. It is the measles of mankind. - A. Einstein
 
bond007
Posts: 4423
Joined: Mon Mar 14, 2005 2:07 am

RE: Possibility To Remote Fly An Airliner

Wed Feb 15, 2006 9:54 pm

Quoting Planemaker (Reply 47):
But just how experienced or knowledgeable is the pilot? Using the United 232 scenario, the plane would have been totally destroyed if it weren't for UAL senior training pilot and DC-10 expert Dennis E. Fitch, who was fortunately travelling as a passenger. I read that in subsequent reconstructions of the flight in simulators, no pilot of any seniority had succeeded in reproducing Fitch's achievement of maneuvering the aircraft as far as the runway.

I agree it was an amazing feat, and they worked miracles. I guess my point is that suppose those engines could have been independently controlled by a computer, programmed to deal with such an incident (which was done by NASA later). Would the outcome had been the same .. better or worse?

Quoting Planemaker (Reply 47):
At the end of the day, it will be a financial decision.

I think it will also be a safety decision. Technology will enable us to 'remove' some human error from the equation.

Quoting VirginFlyer (Reply 46):
That's a very good point, and one which I'd like to expand on. If the computer goes through all its checklists, and still can't solve the problem, then it doesn't have anything else to fall back on, and the plane crashes. If the pilot goes through all the checklists, and still can't solve the problem, he can still fall back on his experience and knowledge

Well, the computer can always have a endless sequence of tasks to perform, even after the standard programmed checklist. But, in my scenario, this is the exact time that control would be handed over to a human pilot, or pilots, on the ground (more than available in a cockpit), to try and resolve the situation.

I agree ... 20 years!


Jimbo
I'd rather be on the ground wishing I was in the air, than in the air wishing I was on the ground!
 
NAV20
Posts: 8453
Joined: Thu Nov 27, 2003 3:25 pm

RE: Possibility To Remote Fly An Airliner

Fri Feb 17, 2006 2:01 pm

Quoting VirginFlyer (Reply 46):
What I am saying is that an aircraft should use both computers and humans to achieve the best possible outcome. We have two resources with differing abilities (and differing strengths and weaknesses) available to us.

I think that's exactly right, VirginFlyer. The three recent incidents I mentioned (they all happened within the past year) were less extreme that United 232, but all of them required a combination of 'lateral thinking', improvisation, and flying skill that only human pilots could have provided.

In the Air Transat A310 case, even the crew did not know (until they landed) that they had lost most of the rudder. But they DID know that (while cruising on autopilot) there had been a loud bang and a sharp change of attitude and 'flying feel' (plus a cabin crew member injured). They were able to turn the aircraft back and land safely by use of ailerons and elevators only. On the face of it, it's difficult to see how automatic systems could have done that (indeed, it's interesting to speculate on whether such systems would have 'realised' that anything was wrong; they might just have flown the aeroplane on to Canada  Smile).

The 747 near Heathrow suddenly lost engine power, and the captain decided to glide in to a landing. There is a radical difference in technique between landing a glider and landing a powered aircraft. Basically, you come in much higher, keeping height in hand until the last moment, because you don't have any reserve of power. Had 'systems' been in charge, they would only have been able to follow the long shallow approach dictated by the ILS beams - and the aircraft would inevitably have crashed a long way short of the runway.

The A320 landing at LAX, which we all saw on TV, went off so smoothly that it 'looked easy' - but it could have been very different. The pilot had to keep the jammed nosewheels off the tarmac for as long as he could; but still put them down before the speed decayed to the point where they would have come down anyway, probably so hard that the impact would have collapsed the gear. Choosing the best 'compromise' moment to ease the nose down (indeed, planning and carrying out the whole landing) would have called for a great deal of skill, judgment, and nerve.

It's hard to see how automated systems could be 'pre-programmed' even to recognise, leave alone deal with, situations like that. And if you're going to have 'ground controllers' good enough successfully to 'fly' aeroplanes that well from a seat on the ground, using closed-circuit TV or whatever, they're going to have to be so highly-trained - and so highly-paid - that you might just as well give them blue uniforms and gold rings and call them 'pilots' anyway.

Because that is what they would have to be.
"Once you have flown, you will walk the earth with your eyes turned skywards.." - Leonardo da Vinci