To Combat Drone or Not to Combat Drone?

[Kane Starkiller]

Maybe the biggest advantage of AI controlled fighter is that AIs abilities won't deteriorate if it doesn't fly and it doesn't get old. You can stick the UCAV in a bunker and after 15 years its abilities will be just the same (provided of course there isn't some mechanical or electronic malfunction).
You can have 5000 craft sitting in hangars with 10 or so doing "training missions" where people are trying out new attack modes and additional AI capabilities. When they are satisfied other 4990 craft simply get uploaded with Iceman_V2.3 program.

[Shroom Man 777]

Oh, and also, one advantage UCAVs have over pilots is that they don't get tired after flying over a dozen sorties non-stop overnight, and they don't need methamphetamines to keep them sharp. So unlike pilots, UCAVs probably won't fuck up and drop a fucking JDAM on friendlies because of combat fatigue, or strafe a whole bunch of civilians, or do those dickeries Israeli fighter jocks are known for.

Love the TOP GUN, Kane. UCAVs also won't buzz air towers OR waste time on homoerotic beach volleyball.

[aieeegrunt]

With the right AI they might

[Zixinus]

And also it just doesn't seem particularly design a weapon system that could potentially be countered by any sufficient amount of tech wanking, remote controlled fighters? Why not jam them? Ai controlled ones? Why not hack them? Having both in one? Increases complexity and possibility of errors and breakdowns no?

The main problem with hacking military stuff is that, unless the man in charge was an idiot, you will need extremely focused intelligence gathering to reverse-engineer the software and even more to find possible exploits. Then you'd probably need at least some semi-specialised equipment and/or personal to actually get those exploits/viruses/hacks to work.

In the end, a lucky moron with a RPG or just traditional anti-aircraft defences might save you some money and investment.

I see the flaw in that just as a non-professional.

[MKSheppard]

Answer me this -- how will you control them once your enemy uses ASATs to kill satellites that relay the orders to the UCAVs and the video back to home plate?

Answer me this -- what do you think will happen after the now 'rogue' UCAVs carry out their primary mission objectives and bomb strafe and kill a refugee column that their processors mis-identified as an enemy tank column, despite glaringly obvious red cross symbols on their vehicle roofs?

[open_sketchbook]

Nothing happens, Shep. Some hippy asshole writes about it, ten people read it and stage a pathetic little protest, and then the world moves on. Nobody really cares if civilians die, unless those civilians are from first-world countries.

[Shroom Man 777]

Answer me this -- how will you control them once your enemy uses ASATs to kill satellites that relay the orders to the UCAVs and the video back to home plate?

Then I guess the UCAVs get pre-programmed to strike designated targets prior to launch ala cruise missiles, or receive commands from the ground like normal planes and have fancy shmancy Starglider AI to allow them to carry out these orders like normal pilots? Obviously this would require fancy shmancy AI whatevers, but I don't think the air force would become all-UCAV until fancy shmancy AI whatevers get developed.

Answer me this -- what do you think will happen after the now 'rogue' UCAVs carry out their primary mission objectives and bomb strafe and kill a refugee column that their processors mis-identified as an enemy tank column, despite glaringly obvious red cross symbols on their vehicle roofs?

The UCAVs get patched to program them to not attack the Red Cross? Then the enemy will start painting Red Cross on their SAM sites to trick the UCAVs!

[Master of Ossus]

Lets put it in another way.

Say you had a weapon system that could win any war quickly and efficently and doing so caused not one of your own to die, but using it easily killed thousands or tens of thousands.

The argument is that it become political easy to resort to violance to solve problems because after all what can they do to strike back?

You have just described a preemptive nuclear strike--something that is anything but politically easy to employ.

Now the philosophical argument is just that a thought experiment, I feel that there are probably more valid criticisms tecnologically that would make completely AI/remote controlling your airforce as impractical and unwise.

People have repeatedly pointed out to you that, even by your reasoning that a military's goal should be broader than service to an individual nation, combat drones are not the "problem," but merely a small facet of the fact that militaries exist and hence inherently waste resources. Further, others have pointed out that virtually all technology makes war "less horrible" for the side that's using it, but that doesn't prevent us from employing that in warfare.

By any reasonable standards, since all militaries are trying to develop better ways of hurting the enemy while keeping their people alive, combat drones are obviously valuable and viable weapons systems.

[MKSheppard]

Then I guess the UCAVs get pre-programmed to strike designated targets prior to launch ala cruise missiles

Then they're nothing more than $50 million dollar cruise missiles that are reusable, as opposed to $1 million expendable cruise missiles.

or receive commands from the ground like normal planes and have fancy shmancy Starglider AI to allow them to carry out these orders like normal pilots?

Radio only goes so far, depending on line of sight, and the time of day -- e.g. you can send signals further at night due to atmospherics -- though you could redirect it through AWACS orbiting at 25-30,000 ft for longer ranges. And radios can be jammed -- the Soviets had a fairly comprehensive suite of jammers for every radio spectrum you could think of -- they even deployed special jammer vehicles based on BMPs or BTRs (I forget which) that were designed to jam VT fuzes so they would not airburst.

Nobody we've ever faced since the GODLESSS SOVIETS has ever deployed or developed such a comprehensive jamming strategy.

The UCAVs get patched to program them to not attack the Red Cross?

That would require some pretty good patter recognition skills, and would there be enough left over processing power to run that pattern recognition software in addition to all the other software?

[Shroom Man 777]

Then I guess the UCAVs get pre-programmed to strike designated targets prior to launch ala cruise missiles

Then they're nothing more than $50 million dollar cruise missiles that are reusable, as opposed to $1 million expendable cruise missiles.

or receive commands from the ground like normal planes and have fancy shmancy Starglider AI to allow them to carry out these orders like normal pilots?

Radio only goes so far, depending on line of sight, and the time of day -- e.g. you can send signals further at night due to atmospherics -- though you could redirect it through AWACS orbiting at 25-30,000 ft for longer ranges. And radios can be jammed -- the Soviets had a fairly comprehensive suite of jammers for every radio spectrum you could think of -- they even deployed special jammer vehicles based on BMPs or BTRs (I forget which) that were designed to jam VT fuzes so they would not airburst.

Nobody we've ever faced since the GODLESSS SOVIETS has ever deployed or developed such a comprehensive jamming strategy.

What's wrong with having reusable cruise missiles that can throw AAMs in the face of enemy fighters? If a UCAV ends up getting jammed and, like, the pilot can't remote control it anymore, having the UCAV revert to reusable cruise-missile mode would be pretty nifty. Hell, you can even program it to automatically attack/evade/dodge jammers in the event of jamming! When communications is restored, the pilot can see that his UCAV's killfucked the fuck out of an enemy ECM site. Wicked cool!

The UCAVs get patched to program them to not attack the Red Cross?

That would require some pretty good patter recognition skills, and would there be enough left over processing power to run that pattern recognition software in addition to all the other software?

Yeah, and a "red cross" can easily mean anything. Does this mean that a UCAV won't attack a Church because its cross is similar to a Red Cross cross? This puts a dampener on the plans for Operation Vatican Freedom! But if we allow the UCAVs to attack churches, then they might also blow up the Red Cross! Obviously if we're going to blow up churches and Red Crosses, then UCAV-controlled areas must be accessible only to the Red Crescent and our good Zionist friends in the Red Diamond!

Either that or the UCAV is remote controlled, but can have an advanced AI for whenever jamming happens or for when it's doing something too fast or fancy for the remote-control pilot to whatever.

[Starglider]

Radio only goes so far, depending on line of sight, and the time of day -- e.g. you can send signals further at night due to atmospherics -- though you could redirect it through AWACS orbiting at 25-30,000 ft for longer ranges. And radios can be jammed -- the Soviets had a fairly comprehensive suite of jammers for every radio spectrum you could think of -- they even deployed special jammer vehicles based on BMPs or BTRs (I forget which) that were designed to jam VT fuzes so they would not airburst.

How about a massive swarm of UAVs (of varying sizes) using laser comms to create a peer-to-peer network?

Damn if I lived in the US I could get a five year DARPA grant to research that idea.

[hunter5]

Oh, and also, one advantage UCAVs have over pilots is that they don't get tired after flying over a dozen sorties non-stop overnight, and they don't need methamphetamines to keep them sharp. So unlike pilots, UCAVs probably won't fuck up and drop a fucking JDAM on friendlies because of combat fatigue, or strafe a whole bunch of civilians, or do those dickeries Israeli fighter jocks are known for.

Love the TOP GUN, Kane. UCAVs also won't buzz air towers OR waste time on homoerotic beach volleyball.

Nah the friendly fire can still result from poor intelligence and or communication errors, but I do agree they are less like with drones.

[Kane Starkiller]

Answer me this -- how will you control them once your enemy uses ASATs to kill satellites that relay the orders to the UCAVs and the video back to home plate?

Provided they are not completely mindless and have even a rudimentary autopilot do a 180 and return to base but I was talking more about more capable AIs.

Answer me this -- what do you think will happen after the now 'rogue' UCAVs carry out their primary mission objectives and bomb strafe and kill a refugee column that their processors mis-identified as an enemy tank column, despite glaringly obvious red cross symbols on their vehicle roofs?

Seeing as how even civilian cameras can recognize human faces, detect smiles and even differentiate between different human faces somehow I doubt a bigass red cross will be a problem for the fancy military AI. I have no doubt it will be able to make out not only the cross but the exact make and model of the vehicle in question.
Can it still fuck up in unexpected ways? Of course but, as British tank crews learned the hard way, so can human pilots.

That would require some pretty good patter recognition skills, and would there be enough left over processing power to run that pattern recognition software in addition to all the other software?

Again cameras can detect smiles on a human face in a room. Battlefield will be more complex but so will the processing power available. Human pilot decisions are basically a list of IF->THEN statements.
IF enemy_craft_x APPROACHING(angle y, speed z, distance q) THEN UCAV_HEADING(y,z,q) AND FIRE_MISSILE(missile_r,enemy_craft_x)
This kind of statements are not all that demanding on processing power. Computers were doing it since Wing Commander.

[MKSheppard]

Provided they are not completely mindless and have even a rudimentary autopilot do a 180 and return to base but I was talking more about more capable AIs.

Then the enemy has effectively mission killed them.

Seeing as how even civilian cameras can recognize human faces, detect smiles and even differentiate between different human faces somehow I doubt a bigass red cross will be a problem for the fancy military AI.

Facial recognition software is in a very cooperative environment.

Not one with dust, smoke, fires, and enemy weapons shooting back at you; while the attacking platform is moving at 500 MPH or faster -- Predators move relatively slowly and have very huge drag inducing stabilized camera mounts -- a actual UCAV would be going much faster, putting more stress on camera mounts and software.

Human pilot decisions are basically a list of IF->THEN statements.

A human pilot can be given incredibly vague orders and then execute them based on his own judgement -- balancing contradictory orders and objectives -- he can for example abort an attack if he feels he can't execute the mission successfully, and then go roam the battlefield for something else that fulfills his objectives. Best example of this would be the B-2's mission to nuke road-mobile ICBM launchers in the Soviet Union in the event of WW3.

[Kane Starkiller]

Then the enemy has effectively mission killed them.

If the enemy has ASAT capability and has shot down all satellites in the area the US has much bigger problems than the performance of its tactical UAVs.

Facial recognition software is in a very cooperative environment.

Not one with dust, smoke, fires, and enemy weapons shooting back at you; while the attacking platform is moving at 500 MPH or faster -- Predators move relatively slowly and have very huge drag inducing stabilized camera mounts -- a actual UCAV would be going much faster, putting more stress on camera mounts and software.

I acknowledged battlefield situation will be more complex but you haven't acknowledged that civilian hand held cameras are not exactly top of the line processor or software. Can a human make out ground vehicles through the smoke?
Why would UCAVs have to drag cameras? Can't they use EODAS?
EODAS+Ghz_multicore_processors+no_cockpit+no_squishy_human=more stealthy+quicker reaction+greater maneuverability=dead enemy craft

A human pilot can be given incredibly vague orders and then execute them based on his own judgement -- balancing contradictory orders and objectives -- he can for example abort an attack if he feels he can't execute the mission successfully, and then go roam the battlefield for something else that fulfills his objectives. Best example of this would be the B-2's mission to nuke road-mobile ICBM launchers in the Soviet Union in the event of WW3.

IF mission_sucess<0.6 THEN EXECUTE(mission_opportunity1,mission_opportunity2)
mission_sucess=f(number_friendly_craft,number_enemy_craft,number_SAM, friendly_craft_type, enemy_craft_type, SAM_type, anti_missile_inventory)
mission_opportunity1(search_grid((a N, b E)),(c N, d E),( e N, f E),( g N, h E)), search_pattern="RT-2UTTKh")
mission_opportunity2(search_grid((a N, b E)),(c N, d E),( e N, f E),( g N, h E)), search_pattern="S-300VM")

[Simon_Jester]

Answer me this -- how will you control them once your enemy uses ASATs to kill satellites that relay the orders to the UCAVs and the video back to home plate?

Answer me this -- what do you think will happen after the now 'rogue' UCAVs carry out their primary mission objectives and bomb strafe and kill a refugee column that their processors mis-identified as an enemy tank column, despite glaringly obvious red cross symbols on their vehicle roofs?

An enemy capable of ASAT launches is probably capable of hunting down random drones hovering over its own territory, I'd think. Conversely, if our air campaign has progressed to the point of tank plinking, we've probably knocked out their ASAT launch capability.

[MKSheppard]

You mean like Iran? They have the capability to place payloads into orbit now, and thus have defacto ASAT capability.

[MKSheppard]

And if you love UCAV killfuckyness. Go read this story The Measure of a Man

What's not to love about the Boeing F-124 Grizzly and it's Packard engines?

[Simon_Jester]

You mean like Iran? They have the capability to place payloads into orbit now, and thus have defacto ASAT capability.

Yes.

If we are fighting Iran, and they are still capable of launching ASATs (which they can only do from the pads or silos that fire space-capable rockets, which are not small), we haven't gotten to the point where we should be sending drones hovering over their territory to attack enemy armored columns.

Now, if they shoot down the satellites in the opening round of the war, the drones are grounded, and this is bad. It's an argument for not relying entirely on drones, unless we can come up with a way to transmit orders to the drones from something else, such as an AWACS. But it's not an argument against the drones as such, for two reasons:

-There are plenty of enemies we might use the drones against that can't shoot down satellites, and
-Worst case, they're weapons we built that cost billions, but which we end up not using. It's not as if the US military doesn't have a history of doing that. The only difference is that this time it would be related to enemy action instead of the weapons just being useless for the war we actually ended up fighting. Big deal.

Also, as Kane pointed out, if we lose all satellites in a given area the UAVs aren't our main problem. Losing GPS is a problem for everything, losing the massively complex networking capability is a problem for everything, losing the orbital reconnaissance is a problem for everything. We're already dependent on our satellites to fight a war, and making ourselves more dependent isn't going to change much.

[MKSheppard]

If we are fighting Iran, and they are still capable of launching ASATs (which they can only do from the pads or silos that fire space-capable rockets, which are not small)

Hmm.....

Course, that's for more lower altitude stuff. For higher altitude, you'd need something like:

There are plenty of enemies we might use the drones against that can't shoot down satellites, and

Like who? Taliban in caves? Then you don't need high tech UCAVs, but Predators with turboprops.

[adam_grif]

Answer me this -- how will you control them once your enemy uses ASATs to kill satellites that relay the orders to the UCAVs and the video back to home plate?

Answer me this -- what do you think will happen after the now 'rogue' UCAVs carry out their primary mission objectives and bomb strafe and kill a refugee column that their processors mis-identified as an enemy tank column, despite glaringly obvious red cross symbols on their vehicle roofs?

If the enemy has started blowing up all of your satellites, you're probably about fifteen minutes away from nuclear war so you shouldn't really stress about things like that.

[MKSheppard]

If the enemy has started blowing up all of your satellites, you're probably about fifteen minutes away from nuclear war so you shouldn't really stress about things like that.

I didn't know you were all for massive and total disproportionate retaliation. It's nice to see more people who think like me.

[Blayne]

Thank you shep, I luvs you. I was dying here as I don't have the knowledge to give a suitable argument and was hoping I could have others argue for me

[Stark]

Yeah, he thinks like you, which is why he thinks Iran can take out an array of satellites based on nothing.

[Blayne]

Theoretical capability yes, do they have practical asats probably not yet.