Space fighters
- Thread starter Evo Plurion
- Start date
mike wightman
SOC-14 10K
Ahh, you must mean Fire Fusion and Steel 
mike wightman
SOC-14 10K
Joking aside, FF&S allows you to build "spinal mount" fighters, and there is no 1 hardpoint per 100t rule either.
Another possibility, depending again mostly on campaign feel, is that a combination of the two might be optimal. A human pilot with some kind of AI co-pilot might be the way to go, perhaps interfaced with the pilot's brain if you wanted a cyberpunkish feel. Or for a not so cyberpunkish feel, there's the Skywalker/R2D2 combo that's demonstrably superior at blowing up BIG ships
.
^ I like that idea. Sort of a mix between Star Wars and that episode of BSG when Starbuck was flying the recon Viper (with C.O.R.A. for company). Computer systems in development right now for the Joint Strike Fighter and other aircraft are already performing all the house keeping functions that normally absorb a lot of the pilot's attention. We can extrapolate out to atleast AI and maybe even to 'thought' controlled craft.
mike wightman
SOC-14 10K
If they can do this now what are robotic/drone spacefighters going to be capable of thousands of years hence???
Also, note the use of the word "autonomously"
Also, note the use of the word "autonomously"
mike wightman
SOC-14 10K
Using book 8 I get a TL9 robot brain with ship's boat 4, gunnery 4, and ship tactics 4 to be 52 litres (0.052 cubic metres), mass 11.8kg.
It has an Int of 6 and an Edu of 6 thanks to it's 58 unit CPU (23 linear, 35 parallel) and 60 units of linear storage.
Problem is it costs 1,786,900 Cr.
It does get cheaper as TL increases though (by TL12 it's about half price).
What price human life???
That is a good point, but (the equivalent of) that is a remote-piloted vehicle. You still have a pilot, but the pilot's back on the carrier. (I didn't see "autonomously" in the article, but maybe I missed it.)Originally posted by Sigg Oddra:
If they can do this now what are robotic/drone spacefighters going to be capable of thousands of years hence???
Also, note the use of the word "autonomously"
Maybe, if there are some space savings, you could have a single pilot controlling several of these vehicles, and several pilots working together...
mike wightman
SOC-14 10K
Third paragraph down.
The human operator supervised but the vehicle did everything by itself.
Now autopilots can already take off, follow beacons/GPS waypoints and land.
The human "pilot" is only needed for oversight, therefore a Playstation generation operator could be all that's required. The computers would fly the plane, much like in some modern manned aircraft that can only be flown with computer assistance.
The human operator supervised but the vehicle did everything by itself.
Now autopilots can already take off, follow beacons/GPS waypoints and land.
The human "pilot" is only needed for oversight, therefore a Playstation generation operator could be all that's required. The computers would fly the plane, much like in some modern manned aircraft that can only be flown with computer assistance.
^ The single biggest issue with remote piloted craft is they are susceptible to loss of communications with the home base.
If any major power develops remote piloted craft as a weapon, then the opposition has no choice but to develop methods to negate their effectiveness. Whether this means subverting the signals going back and forth or inducing a failure of that gear onboard the craft, comms would be the weakest link.
A computer system can be hardened against electronic attack but any comms system it is tied to forces a breach that can not be protected thoroughly from electrical or electronic interference. EMP or high level RF energy will travel through the antennae and disrupt any interface.
Are dead ROV's better than dead piloted craft? No, but when communications to their base is disrupted, a pilot can still continue on with his mission or make decisions concerning the best course of action.
Maybe cyborgs are the way to go?
If any major power develops remote piloted craft as a weapon, then the opposition has no choice but to develop methods to negate their effectiveness. Whether this means subverting the signals going back and forth or inducing a failure of that gear onboard the craft, comms would be the weakest link.
A computer system can be hardened against electronic attack but any comms system it is tied to forces a breach that can not be protected thoroughly from electrical or electronic interference. EMP or high level RF energy will travel through the antennae and disrupt any interface.
Are dead ROV's better than dead piloted craft? No, but when communications to their base is disrupted, a pilot can still continue on with his mission or make decisions concerning the best course of action.
Maybe cyborgs are the way to go?
Evo Plurion
SOC-12
Artificial Intelligence.
mike wightman
SOC-14 10K
I think the whole point of the "autonomous" comment is to point out that the J-UCAS X-45 can carry out its mission without a pilot or operator 
The human "controller" is there to take over if the computer fails in its orders, and as the technology evolves that isn't very likely
As I said earlier, the autopilot to get it to its mission theatre is old tech, cruise missiles have been doing it for years now.
What the J-UCAS offers is a platform that can penetrate enemy space; deliver accurate, but cheap (in comparison), ordnance, and then return home to be re-armed and do it all again.
I wonder when we'll see a dogfighting version?
The human "controller" is there to take over if the computer fails in its orders, and as the technology evolves that isn't very likely
As I said earlier, the autopilot to get it to its mission theatre is old tech, cruise missiles have been doing it for years now.
What the J-UCAS offers is a platform that can penetrate enemy space; deliver accurate, but cheap (in comparison), ordnance, and then return home to be re-armed and do it all again.
I wonder when we'll see a dogfighting version?
mike wightman
SOC-14 10K
Evo Plurion
SOC-12
Egad! We are progressing aren't we?
mike wightman
SOC-14 10K
If I believed in Black Projects then I would bet that dogfighting AI controlled drone aircraft are already flying - but that's just silly, isn't it 
Evo Plurion
SOC-12
It begs the question: how narrow has the gap become between AI-controlled aircraft in computer simulations and on the battlefield?
Autnomous combat operations are still limited by the lack of pattern recognition overall.
Current automation is very narrowly defined parameters, but fully autonomous execution until parameters breached.
Think of it this way: we can set a target profile fairly easily; most computerized targeting systems suse a combination of RF return and IR Sig to ID targets. They do not do this visually, and have no decision making capability outsidde the "is the target in spec and in the target zone?"
AACV's have been on the drawing boards for years; they are now being used successfully against static cround targets and in non-friendly-operations zones. They can not, at present, ID a target by paint scheme alone, and few could do it at all.
Depending upon the IFF dependance, and other parameters, it ranges from easy to slightly hard to spoof them. Dogfighting is still a ways off; the decision to dogfight, however, is even trickier. A German shepherd is a more valid decision tree for friend versus foe in the abscence of clear IFF...
One could, for example, declare a no fly zone, and program it for all items with a radar cross section of range x, and an IR signal orf range y, within a designated no fly zone. That signature is going to be centered on known agressor squadrons, but due to the commonalities of most fighter-interceptors and fighter-bombers, on a bad day, a wandering friendly will get splashed. he AACV will fire on the first target that meets the criteria, unless given an ignore X targets command, or an IFF check. It could easily be possible to acquire tone, and fire AAMs.
Likewise, it is possible now to say find a target with a shillouette in this range Z, on the ground with a vector in range A, which is winthin ops range B, and destroy it. We can narrow down the shillouettes to a few similar models... but the bradly and one of the BMP have nearly identical visual and IR sigs... likewise certain US tanks and Russian tanks have nearly identical profiles (fortunately, those were not in Iraqui service). So, either it has an IFF, or there is a significant chance of a freindly kill if a friendly is in the target zone.
optical recognition is a very weak area for computing technologies; target identification is weak, at best.Target location, however, is strong...
In simulations, the enemy is given a benefit: the computer knows which are friendly and which are foe; the agressors are intentionally dumbed down to prevent action on this knowledge; in the real world, it is in fact a far different issue. Simulations can be programmed very effectively, in ways that real applcations of autonomy can not afford weight, power, or EMF wise.
The strongest methodology, however, is close controlled AACV's with human oversight. I suspect this to be the ideal model for long term fighters... 1 in, say, 10 is a control fighter, with full gravitic protections, and the others are AACV's which rely upon a go-ahead from the human operator.
You won't eliminate the need for man-on-the-scene in space warfare until the lag times are down below destruction times. You can, however, reduce the number of men in HARMs way. (Yes, that is an intentional pun.)
Current automation is very narrowly defined parameters, but fully autonomous execution until parameters breached.
Think of it this way: we can set a target profile fairly easily; most computerized targeting systems suse a combination of RF return and IR Sig to ID targets. They do not do this visually, and have no decision making capability outsidde the "is the target in spec and in the target zone?"
AACV's have been on the drawing boards for years; they are now being used successfully against static cround targets and in non-friendly-operations zones. They can not, at present, ID a target by paint scheme alone, and few could do it at all.
Depending upon the IFF dependance, and other parameters, it ranges from easy to slightly hard to spoof them. Dogfighting is still a ways off; the decision to dogfight, however, is even trickier. A German shepherd is a more valid decision tree for friend versus foe in the abscence of clear IFF...
One could, for example, declare a no fly zone, and program it for all items with a radar cross section of range x, and an IR signal orf range y, within a designated no fly zone. That signature is going to be centered on known agressor squadrons, but due to the commonalities of most fighter-interceptors and fighter-bombers, on a bad day, a wandering friendly will get splashed. he AACV will fire on the first target that meets the criteria, unless given an ignore X targets command, or an IFF check. It could easily be possible to acquire tone, and fire AAMs.
Likewise, it is possible now to say find a target with a shillouette in this range Z, on the ground with a vector in range A, which is winthin ops range B, and destroy it. We can narrow down the shillouettes to a few similar models... but the bradly and one of the BMP have nearly identical visual and IR sigs... likewise certain US tanks and Russian tanks have nearly identical profiles (fortunately, those were not in Iraqui service). So, either it has an IFF, or there is a significant chance of a freindly kill if a friendly is in the target zone.
optical recognition is a very weak area for computing technologies; target identification is weak, at best.Target location, however, is strong...
In simulations, the enemy is given a benefit: the computer knows which are friendly and which are foe; the agressors are intentionally dumbed down to prevent action on this knowledge; in the real world, it is in fact a far different issue. Simulations can be programmed very effectively, in ways that real applcations of autonomy can not afford weight, power, or EMF wise.
The strongest methodology, however, is close controlled AACV's with human oversight. I suspect this to be the ideal model for long term fighters... 1 in, say, 10 is a control fighter, with full gravitic protections, and the others are AACV's which rely upon a go-ahead from the human operator.
You won't eliminate the need for man-on-the-scene in space warfare until the lag times are down below destruction times. You can, however, reduce the number of men in HARMs way. (Yes, that is an intentional pun.)
Evo Plurion
SOC-12
Or until true artificial intelligence, if it's at all possible, removes the need for constant communication between the mothership and its fighter drone squadrons or missiles.
I still think small combat craft will be limited to missiles. Combat aircraft exist because their speed and maneuverability - and the fact that they are aircraft - provide tactical advantages over ground and naval units. Even so, it could be argued that all combat aircraft are support units; they supply, enhance or protect the operations of ground units. In space, all units operate in the same environment and share the same theoretical limits on speed and maneuverability. A space fighter has no tactical advantage over a starship. In fact, its relative cost, limited payload and limited range make it a particularly disadvantageous combat unit. It could be said that all spacecraft are support units; they supply, enhance or protect the operations of units on a world's surface.
