UAV Airliners.

I was really thinking about designing new aircraft which completley remove any trace of the pilot. For example the area at the front of the plane would be dead space which could be used to either provide more seat or on longer flights to accomodate luxuries for those lucky enough to be in first!!!

I have no doubt that the Russians were rather anxious when they sent up the Buran shuttle to orbit on its maiden voyage with no one on board. I believe that if a bankrupt collapsing socialist country could manage it then, 20+ years later it would be easy to create.

People seem to forget that computers already control their safety when they in hospitals. Earlier this week I was on the Docklands Light Railway (a small network in London). It is almost ocmpletely automated apart from a single staff member on each train who's role is security.

Jensy

I really don't think that automation is too far away, although I think manned aircraft still have a ways to go. Both military and commercial. I do, however, think it would be great if the first-classers got to look through a giant glass dome FORWARD instead of a side view. Nothing like seeing a runway on final to end the trip.

Originally posted by pepsi78
I belive nasa's space shutle lands by it's self with out interferece from the pilots, they cant land it by them self, it's a flying brick, and it lands just fine with out any one pressing any buttons .

You sure?

I thought the Shuttle wasn't capable of that, but the Buran was.



Anyway, for those worried about pilotless aircraft, fear not, the FAA/JAA will not allow passenger aircraft with no on-board pilots for 30/40 years at least.

Im pretty sure I read a statistic that said more plane crashes or near misses are caused by pilots making mistakes with the autopilot and thinking they are going in a different direction to what they are, than by autopilot malfunction.

I think that this is particularly interesting as it shows that more accidents are caused by the babysitters of the computers than the computers themselves!!

People are capable of changing their minds and changing the way they go about things. They might usually crusie at 33,000 feet, but they might change their mind and go to 35,000 or go down to 30,000 if they feel like it. These are all irreelvant examples, but they show the idea. Computers can't change their minds. They don't have heuristic processors and can't learn or adapt their programming. If it's int he programming, the computer will do it. If not, then the computer is going to have a problem.

All this mumbo-jumbo basically means that a computer can't adapt like a human, but since this adaption can sometimes be a hindrance (example- making a stupid decision because at the time it "seemed" like the right thing to do), computers occasionally have the advantage over us. Particularly at times of great stress (although there usually aren't many of these in airlines. Usually.) when a normal human is...well... stressed, computers can handle themselves according to their programming. It's a double-edged sword. Sorry if this makes little sense, I'm just trying to say that computers can't make stupid decisions that aren't in their programming.

Originally posted by gfad
Im pretty sure I read a statistic that said more plane crashes or near misses are caused by pilots making mistakes with the autopilot and thinking they are going in a different direction to what they are, than by autopilot malfunction.

Just want to interject that I have been told that the main cause of an aircraft misshap is pilot error. I heard that many years ago so I dont know if it still holds true.

I think most of you proponents are thinking of standard case scenarios. I'm thinking in the lines of something bigger than just a flight from New York to LA. Computers are far better than people at fine-control - but when it comes to handling rapidly changing and 'chaotic' events - people are more likely to either regain control or survive - because their goal isn't necessarily to regain control. That will always be a computer's objective - to regain control of the aircraft - but if you've ever flown an aircraft in an accurate enough sim or in real life - you know that sometimes you have to give up your control in order to regain any shred of it. A human can give up that quest for control and begin a quest for survival over 'doing it by the book'. A person can also look outside and improvise with what is there (to crash into).

Then there are other circumstances... what happens if the entire control grid goes down due to terrorism, a pissed off employee wanting to demonstrate vulnerable programming shortcuts, war, etc? The planes are never capable of recieving confirmation about what is going on at an airport - whereas a person can use a cellphone to contact someone on the ground - or look out the durn window for himself. Then work out who is landing when with other pilots over the radio (not impossible for computers - if they could ever get past 'there is no way of knowing what is down there').

There are a million extraneous events that are unplanned that a person can adapt to in about a few seconds but a computer is unable to adapt to - and will lead the entire plane to destruction.

Im afraid I disagree again Aim.

In an emergency situation a human responds much WORSE than a computer. It is well known that a person who isnt stressed works much better than someone who isnt and the stress levels inside the cockpit of a plummeting plane are pretty high!

Whereas a computer never gets stressed or reacts rashly as it has no concept of danger.

A computer can evaluate probabilities of events in milliseconds and decide on the best course of action before a human pilot may have even noticed there was an event to react to.

I recently saw a demonstration of a computer system that could control the flight surfaces and engine thrust of a plane on a tiny and fast responding scale. This allowed for a plane which say has lost control of all its tail control surfaces could still land safely. An experienced pilot was put in the simulator cockpit and was put in a situation identical to that which led to the Sioux City crash. He crashed within minutes as the planes was basically uncontrollable, but when the new system was turned on he could fly the plane safely to the nearest runway almost as if nothing had happened.

Its the human minds flaw that they have a basic survival instinct. A computer doesnt have this which means that although survival may be its objective it doesnt attempt ridiculous or dangerous alternative courses of action unless it has evaluated other options and found that to be the optimum.

I think a computer would stand better in an unpredicatable emergency than you give it credit for.

Again, ignoring the other aspects which are more relevant than mechanical failure.

A machine's goal is control. That is its fatal flaw.

A human's reaction to extraneous events are far greater than those of a machine's. Pure and simple. We can handle events and values greater than our ability to comprehend. A machine hits a critical error and behaves eratically.

Originally posted by Aim64C
A machine's goal is control.

Incorrect. Quite incorrect, actually. A machine's goal is whatever it is programmed to do. It evaluates and decides according to it's PROGRAMMING. If the machine is programmed to regain control asap after losing it, it will do that (or try to). If it's programmed to do something else, say, sliding backwards after losing control in a stall, pointing at the enemy's lead and launch a missile... It will do that. The secret is in the programming. A computer can contain much more information and process that information much faster than a human can, that's very simple. If you had a supercompuer calculate all (or a very great deal of) possibilities and scenarios, loaded that all into a program, you may just have a computer with better control than a human would.

A human can adapt, that's our advantage.
A computer works a whooooole lot faster than we do and is unaffected by fear and stress. That's their advantage.

A machine's goal is ALWAYS control because it is always programmed to control something - or there would be no need for it at all.

Originally posted by Aim64C
A machine's goal is ALWAYS control because it is always programmed to control something - or there would be no need for it at all.

If there was a total lack of control and the machine couldn't do anything, then the computer would need to regain control, and would proceed to do so, assuming that the parameters were met in its programming. If someone wanted to, when an airplane entered a flatspin one could program it to regain control for the pilot. But they could also program it to eject the pilot. Or drop every munition. Or even self-destruct. It ALWAYS depends on the programming within the computer regardless of anything that happens to the aircraft. If a computer is not programmed to do something, it won't. If it is programmed to do something, then it will. Very simple.

A computer will only take action that is dictated by its programming. Unless it is capable of changing its own programming, which means it would be AI which it isn't.

Which is exactly why a pilot will always be better in extreme cases rather than a computer. Extraneous solutions or problems are not addressed by a computer.

Say what you want to - I'll shoot your robot planes out of the air faster than you can build them if your objective is to remove people from responsibility.

Originally posted by Aim64C
A machine's goal is ALWAYS control because it is always programmed to control something - or there would be no need for it at all.

That doesnt make much sense! Following your line of thought surely a pilots objective is also control since all they can do is control the plane. Your logic doesnt make sense though. You are using two different definitions of control as the same thing. The first is control - an entitiy influencin something. The second is control - having a stable and flyable aircraft.

All a computer (or a pilot for that matter) can do is control (1) the plane. But a computer would be better at regaining control (2) of an aircraft in an emergency.

Also when you are talking about "extraneous situations" pilots are not trained in these situations either! Pilots are trained in basic emergencies but not obscure and highly unlikely ones which you are discussing as a barrier to UAV airliners. In these situations a human pilot would probably also crash and burn.

"Say what you want to - I'll shoot your robot planes out of the air faster than you can build them if your objective is to remove people from responsibility."

Resisting inevitable change again ...

Computers will soon be controlling all aircraft, whether you choose to resort to terrorism or not.

I can see the benefits of using computers to control the majority of operations. But like others, I still want somebody who can manually take control should an emergancy arise. My reasoning? I don't want the blue screen of death to ever become literal.

Originally posted by gfad
You are using two different definitions of control as the same thing. The first is control - an entitiy influencin something. The second is control - having a stable and flyable aircraft.

Glad to know I wasn't the only person struggling with that . You don't even need a stable aircraft to do things. Flyable? Dispensable too (although not permanently). Take the Raptor. A rocket with wings (Wile E. Coyote, anyone?), if you will. I'm gonna take flak for this comment, but it's not so stable-which is why it needs a powerful flight computer to manage a lot of the functions so that the aircraft can fly. However-need to dodge a missile? Let the plane flip out. I couldn't predict where it's gonna go, can the missile? Maybe, maybe not.

Also, with TVC it allows you to CONTROL the plane without it necessarily being FLYABLE. The idea being that were you to stall the aircraft you could still use the engines to force the aircraft to fall tail-first. pointing straight up. And were your opponent coming straight down at you, you might even hit them with a well-timed missile launch. So you needn't be flying stably to have control, or even to have a tactical advantage.

Now the thing is, can a human say "this will be my maximum amount of maneuvering before the aircraft becomes irrecoverably out of control"? No, unless we can calculate these things in our heads (and it's a long equation), so we have to guess. A computer can, though. And it can do it while it is still relevant. By the time you finished the equation on your notepad the figures would be grossly inaccurate because all the while, those figures have been changing. So we can't know our best way to maneuver while still maintaining the ability to recover the aircraft.

Also, humans are affected by FEAR. We don't want to die... Most of us. So we probably won't go anywhere near the limit of maneuvering while maintaing the ability to recover because we don't want to make a boo-boo and get smacked by reality for it. Or the ground. But a computer can't fear, it can only calculate, then do. So, placed alongside a human pilot and having to complete the same maneuvers, the computer would most likely do them the fastest and with the most effect. Meaning that in combat, it would have the advantage.

Returning to the airliner subject, we can see that a computer can maneuver the most effectively. This means that if an airliner were to lose control for whatever reason, or if something were to happen, it would be able to react near-instantaneously to the event, and perform with maximum efficency and effectiveness, because a computer can actually calculate the best way to do things, while a human can only guess.

Man, that was a long rant.

Computers will soon be controlling all aircraft, whether you choose to resort to terrorism or not.

Well played.

[edit on 8/16/2006 by Darkpr0]

[edit on 8/16/2006 by Darkpr0]

Originally posted by cyberdude78
I can see the benefits of using computers to control the majority of operations. But like others, I still want somebody who can manually take control should an emergancy arise. My reasoning? I don't want the blue screen of death to ever become literal.

"Systems" like the Global Hawk that are fully autonomous have dual seperate redundant systems... running totally DIFFERENT firmware, code, 'operating systems', etc. Should one fail the other should be unaffected.

When a global hawk engine fails, the computer can:

1. Find the closest airfield.
2. aim and decend appropriately.
3. Land on the strip.
4. Phone home.

I think computers can do just about everything better than humans, you just need redundancy.

As far as "the whole network failing"... you would not design it this way. Each flight would be "pre-programmed" and the computers would make adjustments as necessary should a problem on the ground occur.

I would check into the capabilities of the Global Hawk...

1. It files its own flight plans with the FAA.
2. Can taxi.
3. Take off.
4. Fly in the commercial lanes.
5. Land.
7. Taxi off.

It can also...

1. change destinations should a ground issue occur without intervention.
2. Land with a failed engine.
3. Recover from almost any flight issues.

All of the technology is there, it works and will be made safe enough for human passengers very soon.

Study the Global Hawk.

[edit on 16-8-2006 by Slap Nuts]

Originally posted by Slap Nuts

All of the technology is there, it works and will be made safe enough for human passengers very soon.

Study the Global Hawk.

Slap Nuts,
I do agree that we will see transport size UAV at some point. Cargo first of corse. But as good as the global hawk is, if I rember right, two of them did go down during the wars in the middle east. I dont really know when it will happen, but it will happen.

Originally posted by kilcoo316

Uhm, the newer aircraft can taxi around the major airports.

I continually wind up my pilot buddys with their entire work routine...

upon entering aircraft => autopilot on

just before leaving aircraft => autopilot off

What system can taxi a commerical aircraft, from the gate to the active? LOL

There much more to flying than just pushing a few buttons. But yeah I can see how that would "wind them up"; you know diminishing their value in a profession wherer they have worked very hard to get where they are.

Originally posted by Slap Nuts
"Systems" like the Global Hawk that are fully autonomous have dual seperate redundant systems... running totally DIFFERENT firmware, code, 'operating systems', etc. Should one fail the other should be unaffected.

When a global hawk engine fails, the computer can:

1. Find the closest airfield.
2. aim and decend appropriately.
3. Land on the strip.
4. Phone home.

I think computers can do just about everything better than humans, you just need redundancy.

As far as "the whole network failing"... you would not design it this way. Each flight would be "pre-programmed" and the computers would make adjustments as necessary should a problem on the ground occur.

I would check into the capabilities of the Global Hawk...

1. It files its own flight plans with the FAA.
2. Can taxi.
3. Take off.
4. Fly in the commercial lanes.
5. Land.
7. Taxi off.

It can also...

1. change destinations should a ground issue occur without intervention.
2. Land with a failed engine.
3. Recover from almost any flight issues.

All of the technology is there, it works and will be made safe enough for human passengers very soon.

Study the Global Hawk.

[edit on 16-8-2006 by Slap Nuts]

The Pentagon admits that the Global Hawk UAV has a crash rate more than 50 times that of F-16 piloted fighter jets, and has set a target to reduce this by 2009 (see graph)
NS

He and the manufacturer of the Predator, General Atomics Aeronautical Systems, blame the pilot of the drone for the accident. He worked for General Atomics, which flew the Predator under contract with the government. The pilot told investigators from the National Transportation Safety Board (NTSB) that his control panel froze up. When he switched to a second control station, he didn't notice that it was set to shut off the plane's fuel. The switch cut off the Predator's engine. The drone had been given permission to fly at 14,000 to 16,000 feet, an area that was closed to other planes. After the engine quit, it drifted until it struck the ground.
USA Today

Opps so much for no pilot error with UAV. I think you are confusing AUAV with UAV. Big difference there.

[edit on 16-8-2006 by Imperium Americana]

Imperum,
That link had a lot of good information on it, thanks. There is a lot of big names trying to get more acess to comershial air ports for uav's. Below is a quote.

The plan, which also involves giving unmanned aerial vehicles (UAVs) easier access to US civil airspace, is the result of a powerful partnership led by the Pentagon, NASA, Lockheed-Martin, Northrop Grumman and Boeing. Called Access 5, the group aims to loosen restrictions on where UAVs can operate within 5 years.