Posted Tuesday 27th November 2007 15:10 GMT
'Human passengers will probably always insist on having a human pilot up front.'
What a mindless prediction to make. Always? Perhaps for the next few decades, if that. When robot pilots become safer then humans, why would you insist on having a human?
Posted Tuesday 27th November 2007 15:36 GMT
If you remove the soldiers and pilots from a military system you achieve several things but two stick out for me:
1) The conduct of a war, the tactics and strategies, used by the mechanized country will have essentially no restrictions. If there are no Americans on the battlefield, then there is no way for the American public to know if the war is being fought according to Hoyle, or at least Geneva.
2) Politicians in control of these systems have carte blanche to use violence in international relations. Indeed, from having robot weapons to having a point and click attack interface in the White House would be a minor development.
A democracy like the USA has (in theory) check on the power of the executive: Congress has the power to declare war, the exec does not. In the past when the exec fought undeclared wars - say Cambodia, Laos, Central American insurgencies - it got caught out because people talk. But if the exec has this kind of weapon system, then congressional approval of war becomes irrelevant: the exec simply does not tell the congress that there is a war. The executive branch becomes, if it so wishes, autonomous in it's ability to attack and destroy anyone it perceives as enemy - congressional oversight intended to keep the exec 'sane' falls away.
These weapon systems undermine a principle check on executive power in the USA. They are not really intended to 'save American lives', so much as they are intended to dramatically extend the power of the executive. They are fundamentally a tool in a fight between the exec and congress over which body controls the USA federal government.
I ask: when, in history, having acquired unrestricted military power, has a government executive/king/emperor failed to attempt imperium?
Posted Tuesday 27th November 2007 15:41 GMT
"Anyway, an airliner wouldn't become a lot more efficient for having its pilots removed because it already has to carry hundreds of people."
How about the efficiency of not having to carry two £50,000 - £100,000 salary meat bags, who hit the vodka bar directly before the flight and health and safety requirements about hours in the air and time spent sleeping/in the vodka bar.
Maybe removing the pilots isn't such a commercial no no afterall.
Posted Tuesday 27th November 2007 15:41 GMT
"When robot pilots become safer then humans, why would you insist on having a human?"
Because most people are not rational, particularly when it comes to flying and fear, or indeed relinquishing any kind of control to a robot! I would put money on it that even if a robot was doing the flying, the general public would insist on there being a trained pilot sitting in the cabin just to keep an eye on things!
It doesn't matter how statistically safe a robot piloted plane, train or car could be people are going to be a bit wary of it because if they cannot drive themselves, they would rather delegate the responsibility to another member of Homo sapiens rather than a collection of wires.
That's just one of the irrational animal based tribal responses which makes us human! Maybe it will not be always and until the end of the universe/human race - but it'll be a bloody long time till people get over it!
Posted Tuesday 27th November 2007 16:21 GMT
Why use robot warplanes? What's the point? As I see it their only function is a re-usable weapons platform.
Why not develop better missiles? If you can exceed range (no return journey), cheaper, cheaper/stealthier ships (no need for a runway). A 'single use' missile (jet or rocket powered) can replicate all the functionality in a robot plane, you can carry and support more in the air.
I don't get it.
Posted Tuesday 27th November 2007 16:21 GMT
...and I'd like to welcome you to this completely automatic flight. We're cruising at 30,000 feet and I'd just like to reassure you that due to the careful design of this robot plane, nothing can go, nothing can go, nothing can go...
Posted Tuesday 27th November 2007 16:21 GMT
What O/S do they use again?
Posted Tuesday 27th November 2007 16:21 GMT
"I ask: when, in history, having acquired unrestricted military power, has a government executive/king/emperor failed to attempt imperium?"
Pompey the Great when he landed at Brundisium on his return from his Eastern campaigns. The Senate expected him to march on Rome, but instead he dismissed his troops and bade them return home peaceably.
The Emperor Diocletian voluntarily abdicated his power and went into an unmolested retirement.
Count Belisarius consistently refused the imperium despite his personal popularity and military prowess.
that's just three that come to mind.
that's just two that come to mind.
Posted Tuesday 27th November 2007 16:31 GMT
A quote, I forget who said it first…
“It is said, that one day airliners will be crewed by only a single human pilot and a dog. The pilot is there to feed the dog. The dog is there to bite the pilot should he try to touch anything.”
Posted Tuesday 27th November 2007 16:43 GMT
>When robot pilots become safer then humans, why would you insist on having a human?
We don't have robot trains, despite them actually being in use elsewhere.
Posted Tuesday 27th November 2007 16:49 GMT
General George Washington.
Upon Britain's surrender, he was offered kingship; he refused, allegedly stating (and I have to paraphrase, due to severe CRS) "You would replace one King George with another?"
@Matthew:
"Why use robot warplanes? What's the point? As I see it their only function is a re-usable weapons platform."
Exactly...what's got you confused here?
"Why not develop better missiles? If you can exceed range (no return journey), cheaper, cheaper/stealthier ships (no need for a runway). A 'single use' missile (jet or rocket powered) can replicate all the functionality in a robot plane, you can carry and support more in the air."
With no return journey, you cannot properly perform the duties of patrol and rendezvous; identify the target visually to ensure you're not blasting some civilian aircraft out of the skies.
And a single-use missile, if launched, must end its flight somewhere; bottom of the ocean, lost in the desert, through the roof of your house...
And that's before we begin discussing the ramifications of monetary expenditures on said "single-use" missiles...
'nuff said...
Posted Tuesday 27th November 2007 16:49 GMT
You need planes because missles in themselves are a limited weapon. You CAN put the missles a considerable distance out, like on a ship, or even another continent - we have those already. But distance-fired missles are usually limited in the accuracy of their delivery, unless they are hitting a static target that an hour's old GPS fix is good enough to hit.
But if you want to hit something that is mobile, or moving, it takes localized target designation, and and the ability to respond to target countermeasures. And that is something that planes are good at, even if they may use short-range missles to actually deliver the payload. They can aim a laser target designator and use remote sensors to keep it locked, and guide the missle in. They can also relay ground images and intelligence back to command, for target assessment and post-strike assessment, something no missle can do.
And planes can loiter for hours over the battlefield, able to rain down death in a matter of seconds, rather than a 5 - 20 minute guided missle journey from an offshore ship. If you are infantry that calls in airsupport on the battlefield, that 5 - 20 minutes will seem like hours - and the situation on the ground is likely to have changed.
Lastly, there is a matter of cost of munitions. A plane can drop unguided and semi-guided munitions - a missle is by nature an expensive precision-targetted munition, and very expensive, especially when not taking out high-value targets (ie, other planes, ships, bunkers, assassinations, etc.). If you need to napalm the entire side of a hill to cover your infantry's flank, that is horrendously expensive very quickly. Better to have a few overflights with re-useable planes dropping unguided napalm bombs on that hilliside.
The answer is planes are not going to go away on the battlefield. But having live pilots seems less and less cost effective (at least until enemies figure out how to EMP them to destroy their CPUs, or jam the comms links, or they have to fight over a tactical nuclear battlefield that has no comms available).
Robert
Posted Tuesday 27th November 2007 17:22 GMT
Ok, it is not an absolute rule. But it is a good rule of thumb.
Posted Tuesday 27th November 2007 17:22 GMT
The power of executive is held in check by the power of the purse strings, which is controlled by Congress. Precision-guided high-tech weapons are VERY expensive, and only Congress (in the US) can appropriate enough monies to allow for the use and replacement of stocks in any meaningful amounts. Does that mean the executive couldn't fire off a few dozen on his own? Nope. But it means that the executive couldn't order any large-scale efforts and hope to have the stocks replentished without the approval of Congress.
Having said that, given the power of the defense establishment's lobbyists and campaign contributions, it is not clear that even Congress is much of a control mechanism...
Robert
Posted Tuesday 27th November 2007 17:25 GMT
quite right, I say...
if they do need a UCAV, why not design a controlling system that can fly an existing platform? reusability, anybody?
Posted Tuesday 27th November 2007 17:25 GMT
< "I ask: when, in history, having acquired unrestricted military power, has a government executive/king/emperor failed to attempt imperium?" >
< "that's just three that come to mind." >
You forgot Cincinatus. And George Washington, whom was offered the role of King of America, but followed the example of Cincinatus. Indeed, America has a history of such, in the formation of the Society of Cincinatus.
Still, E has a point... Such examples are notable by their very rarity. All the more reason to ensure that the Second Ammendment is protected... There may come a day, sooner rather than later, when the Executive Branch may need reminding for whom they work.
.
That all said, I approve, in general, of non-manned combat drones - They'll be neccessary to deal on an one-to-one basis with opposing drones, of which we can *guarantee* there will be some. Further, we needn't *assume* that they'll be tools of oppression; I'm quite sure that in several cases of the US Executive Branch's abuses of power, it wasn't *US* people who blew the whistle, but rather it was the Other Side. Robots or no robots, I can't see the Other Side in any future case of abuse of privilege and power just lying down and choosing passivity.
Posted Tuesday 27th November 2007 17:25 GMT
Call me irrational but no matter how good a robot is at piloting a plane I want a human at the helm.
Robots have no imagination or creativity and lack the ability to "think" illogically in a creative way. They can not be reasoned with, they can just run instructions and the way they perform is only as good as the instructions that they are given.
If the computer driving your docklands light rail train has a divide by zero error and you miss your next stop or come to a grinding halt you dont have to worry about the 30,000 ft of air between you and the ground and the crater that the 200 - 300 vehicle you are in is going to make when it reaches it.
A robot at the helm has no feelings either way about becoming one with the earth at high speed from a perpendicular direction however a human usually does and as a general rule of thumb will try to avoid it by all means possible.
The most successful and widely used fully autonomous aircraft used to date is a cruse missile and that is not the kind of cruise I wish to be boarding when I go on holiday.
Posted Tuesday 27th November 2007 17:40 GMT
Heck, they had them at least partially automated in the days of the F-8 Crusader.
The major difference with this is the lack of the backup meatsack.
Posted Tuesday 27th November 2007 22:29 GMT
Where's the ROTM warning in the article?
Posted Tuesday 27th November 2007 22:29 GMT
"...how to EMP them to destroy their CPUs, or jam the comms links..."
Good points, but I imagine a faraday cage around the sensitive bits would be good protection.
AFIAK current cruise missiles do not depend on a comm link to acquire their target. When the cruise was tested in western Canada in the '80s we were told that the tests were to develop or validate terrain recognition software, and that northern Alberta terrain was very good for that purpose.
Which begs the question: if an automated attack plane can recognize the terrain, and find it's target by itself, *and* it's comms get jammed, then is the owner of that plane going to be capable of recalling an attack (which seems a reasonable feature to have)?
Posted Tuesday 27th November 2007 22:29 GMT
is nearly the case. Autopilot and ILS systems, combined provide auto-landing systems and they are nearly as good as a pilot landing with good visibility. They have existed for quite a while now and are commonly used for low visibility landings. They apparently were originally not too good with strong winds, but I guess they can cope now.
Autopilot does the bulk of the work after journey decision by the pilot. I would not be to amazed if the journey was calculated by computer to minimize fuel usage using wind force (specially for east-west oriented flights). The pilot probably still has the chance of driving his enormous winged taxi in the traffic jam before take off.
Both planning and taxi could be automated. I am strongly for keeping a couple pilots on airliners, as shit happens and some crew is needed anyway. Manual pilot will probably be a sign of emergency in the coming years.
Posted Tuesday 27th November 2007 22:31 GMT
Besides the ability to carry a larger payload and exceed the Maximum G Factor these craft could also be used for spy missions or stealth attacks. Plus if they have no human onboard there are no qualms about self destruction if they get caught.
They could be used for radar and communications support in hostile territory; Kind of like some friendly eyes in the sky.
Posted Tuesday 27th November 2007 22:31 GMT
WRT Congressional control of the purse strings: certainly Congress has that. However Congress need only approve purchase of a few such weapons and the problems I see will exist. In the current state of politics in the USA, and the pressure that could be brought to bear on Congress, I cannot see it refusing to fund production of such weapons if they are shown to work.
GWB or the next version of same would merely have to start braying that "Congress wants your sons and daughters to die on the battlefield!", and Congress would crumble.
Sorry if I'm very cynical about this.
Posted Tuesday 27th November 2007 22:31 GMT
they are flown by pilots on the ground who can in theory fly multiple uavs at once and when they need to go to the restroom or the end of their shift another pilot takes over. The other reason you will see uavs replace manned combat aircraft eventually is the fact the plane is already capable of doing more than the pilots body can handle. I worked on f-16s for the usaf and they had g-limit switch overrides only to be used in extreme emergency because the plane could pull g forces that would def cause the pilot to blackout but also possible cause structural damage. In the future they will build these planes for higher g tolerance from the start because the pilot is no longer your limiting factor...
Posted Tuesday 27th November 2007 22:31 GMT
"Call me irrational but no matter how good a robot is at piloting a plane I want a human at the helm.
Robots have no imagination or creativity and lack the ability to "think" illogically in a creative way. They can not be reasoned with, they can just run instructions and the way they perform is only as good as the instructions that they are given.
If the computer driving your docklands light rail train has a divide by zero error and you miss your next stop or come to a grinding halt you dont have to worry about the 30,000 ft of air between you and the ground and the crater that the 200 - 300 vehicle you are in is going to make when it reaches it.
A robot at the helm has no feelings either way about becoming one with the earth at high speed from a perpendicular direction however a human usually does and as a general rule of thumb will try to avoid it by all means possible.
The most successful and widely used fully autonomous aircraft used to date is a cruse missile and that is not the kind of cruise I wish to be boarding when I go on holiday."
I would rather have a robot pilot that will stay awake, rational and logical rather than having some schmuck who is either boozed out, horny at the stewardess, psychotic nutbag that decides that his life isnt worth living or any of the other hundred "pilot error" statistics sitting up in front.
Besides, if a robot does go awry for some reason, you can always have a human from the ground take control of the flight in an emergency.
Posted Tuesday 27th November 2007 22:31 GMT
The yanks are extremely good at friendly fire but how on earth are they going to train a robot to shoot at it's allies?
Posted Tuesday 27th November 2007 22:31 GMT
The X-47B is a bomber, so no, it will not blast anything out of the sky. Also, it is completely autonomous: its course and target(s) are preprogrammed, so there is no wetware in the control loop once the "go" button is pushed.
Once airlines begin offering reduced fares for flying on fully automated aircraft, pilot's days will be numbered.
BTW, why is this not in the RoTM category?
Posted Tuesday 27th November 2007 22:31 GMT
Paddy:
"if they do need a UCAV, why not design a controlling system that can fly an existing platform? reusability, anybody?"
That's been going on for decades for obsolete-aircraft-target drones, like you see in all those shoot-em-up SAM videos. Really, pilots were scarce!
But autonomous versions of current planes are flying and more proposed: an unpiloted-but-occupied F-18 is used to test the X-47's landing software today:
http://www.flightglobal.com/articles/2007/11/26/219820/northrop-grumman-begins-testing-x-47b-ucas-d-software.html
and a G550 UCAV is being offered for the U.S.N.'s BAMS program: Seehttp://www.spyflight.co.uk/IVsp.htm
look up RQ-37.
Posted Tuesday 27th November 2007 22:31 GMT
Not necessarily just for the big planes but also for for the possibility of having small autonomous aircraft that can deliver a few thousand pounds of cargo quickly to smaller or remote airports.
Posted Wednesday 28th November 2007 00:12 GMT
Aside from the whole no casualties if it gets shot down thing, the human being physically in the aircraft really is the weak link in terms of maneuverability. Many modern aircraft are capable of much greater maneuverability than we allow them to operate with. An aircraft could pull up in a 10 G climb, for example, and there wouldn't be concerns about a pilot blacking out. Even if the planes weren't completely autonomous, the pilot could be sitting in more or less a simulator, with a full cockpit only the aircraft being controlled is on the other side of the world.
Pretty slick stuff...
Posted Wednesday 28th November 2007 01:43 GMT
That way we get extra service out back whilst the plane controls itself and in the event of an emergency the plane can page the pilot to come and help out and can take over till the plane's computer corrects itself!
To be honest though even in an emergency id trust a computer to land an aircraft more safely then a lot of pilots!
Posted Wednesday 28th November 2007 09:43 GMT
Given what drives commercial airlines (costs and sales), it will be interesting to see how the market changes when it's cheaper to insure a robot airliner, and it the fuel consumption is a lot better, but older airlines push the 'would you trust your life to a toaster' marketing angle.
On the topic of fully autonomous fighters/bombers (vs UAVs), given that modern game AI has to be toned down to not waste anything with the enemy marker above it's head in 3 seconds flat, I'd hate to be the sucker from a small central American country that has to go up against a stealthy, hight speed, highly maneuverable, well armed aircraft that's about as smart as a wasp (which you could probably simulate neuron by neuron). Or for that matter, I'd hate to be the US pilot flying an F35 going up against a 'swarm' of Sukhois.
Posted Wednesday 28th November 2007 10:07 GMT
Modern fly-by-wire aircraft are already flown by computer, or rather five or so of them all voting on the best course of action. When the pilot makes an input, each computer calculates the adjustments necessary, adjusts them to fit the flight envelope for the aircraft considering passenger comfort and so on, then does something like what the pilot says.
This was quite noticable with one of the airbus aircraft (I forget which) when they got the flight envelope wrong, it used to strike its tail on takeoff as the flight envelope didn't allow the aircraft to pull up enough at rotation.
Make the aircraft robot controlled, and give the pilot a copy of Flight Sim, see if anyone notices....
Posted Wednesday 28th November 2007 13:33 GMT
The difference between cruise missles and planes is that cruise missles use the terrain for a lock and path to a static target. ROVs use comms to perform visual sweeps of a battlefield looking for DYNAMIC targets, i.e., things that move or can employ countermeasures. Until we can put something that can target as well as a human operator in the plane itself, it will always need a good comms link for target identification, as well as post-strike damage assessment (i.e., do I need to hit it again?).
Posted Thursday 29th November 2007 01:15 GMT
Robot Pilots:
Reminds me of a quote from Stargate
Sam "We've designed it for every foreseen scenario"
O'Neill "Yeh, but Its the unforseen scenarios that worry me"
Robots cant improvise as yet so a human is still a better option in an emergency.
Posted Thursday 29th November 2007 01:15 GMT
So this is where the arms race begins.
Q: How do you take out large volumes of robots.
A: With a neutron bomb.
So the next time we see a nuke used I wonder who will drop it, the side with lots of robots, or the side without.
Thankfully Neutron bombs are not as easy to build as your classic nuke.
Posted Thursday 29th November 2007 09:14 GMT
In Iraq, Australian pilots reportedly earned the ire of the yanks, because pilots would refuse to attack targets if they did not believe the target was correct or that they could not meet international obligations, eg. minimising civilian casualties.
The major issue I have is that someone sitting a 1000Kms away can designate a target, the robot attacks regardless. Yet will the person who tasks the robot appear before the military tribunal. Will they be military. One programmer can task thousands of robots. Where are the checks and balances?
One of the great disincentives for war is the loss of your own troops, and the political consequences. That is why non-miltary powers normally negotiate. The United States without that disincentive might consider force as the first option - no recent precedents, of course ;-)
In autonomous mode, how will that American robot distinguish the forces of allied forces, something the real Pilots could not achieve, with devastating results. There is no such thing as friendly fire especially with a fully armed robot that cannot refuse an order.
Back to the article
