Slashdot Mirror
Army Eyes Anti-Sniper Robot
Hiawatha writes "iRobot has teamed up with Boston University to create a robot that can spot enemy snipers on a battlefield. Before the smoke of the shot clears away, the REDOWL robot should have the shooter in its sights." iRobot is the same company that brought you the popular Roomba robotic vacuum.
Besides, he said, it would be dangerous to have a weapon-toting robot that could open fire on its own.
''You need to have a man in the loop," he said.
The article says that the robot would not return fire, it would just pinpoint where the shooting is coming from. So, why does it need to be a robot exactly? Why wouldn't it just be a comptuer with some cameras and microphones?
One idea is that our soldiers could have a chip in their dogtags that the robot could identify so as to not shoot at them. Then you would have the problem of the enemy stealing people's dog tags, but maybe you could deactivate that code once you knew the enemy had the tag.
Avoid Missing Ball for High Score
A good sniper needs only one shot to have his job done.
Back in the old days, before Rod Brooks started iRobot, I worked part-time at a small MIT AI Lab spinoff making robots for kids called Turtles. The Turtle was an outcome of Logo, and, which itself was an outgrowth of Lisp, and so somehow the company's name got on a list of AI and Robotics vendors. (Sidenote: Lego Mindstorms also came from this same group of people at MIT and industry, though not this particular startup.)
Anyway, we got a letter from a defense contractor asking for "applications of our AI and robotics products to battlefield logistics" and gave a half-dozen or so areas for us to evaluate our products.
As you can imagine, puzzlement gave way to amusement, which quickly gave way to mayhem and by 3:05AM we had started writing our response, starting off with "The Turtle enjoys very low observability, due to a minimal radar cross-section and an almost non-existent infra-red signature."
The letter made the rounds on the photocopy/bulletin-board circuit (there was no electronic copy available outside), and somehow the response got published in an ACM journal. Through the magic of modern imaging, it is available for you to read today in PDF.
I admit that I know nothing about this technology, at all, but in my imagination, I would like to think that a technology such as the one used for hunting via the internet would be a little better. At least a real human would have the ability to decide when and where to shoot rather than relying on code to decide whether or not the target is friend of foe.
;-).
Just a thought, though I admit that a robot has a lot more of the geek quality many of us would like to see
My lame blog.
Think for a moment. This isn't for "snipers". It's to pick out the one person in the crowd of protesters who's shooting at you.
The US military is in the process of completely redesigning itself as a civilian occupation force. Iraq is just training.
"I assumed blithely that there were no elves out there in the darkness"
I few years ago I think one of the Discovery channels had something on an optical bullet tracker, I'm not sure if it was from Trex or not. Basically, it took advantage of the fact that bullets move much faster than anything in their vicinity, so the only major differences between successive images from a high-speed digital camera would most likely be the bullets. They had a cool demo system where the camera was pointing at a shooting gun in the distance, and it was highlighting the trajectories of the bullets it detected on a computer screen, with all of the lines converging on the gun. Couple this optical system with acoustic triangulation, and you have the best of both systems: zero in the camera quickly on the general source of the shots using sound, then pin-point the exact source of the bullets optically. Eventually this system could be made compact enough to fit on top of a rifle like a digital scope.
Ghillie suits utilizing evaporated aluminum minimize the IR signature of the person wearing them. Thwarting thermal vision required a change of tactics and slightly more specialized gear on the sniper's part - which is exactly what will happen with acoustic tracking devices. Muffling the report of the gunshot down to nearly nothing can be performed without significantly altering the ballistics of the bullet being fired by using conventional suppressors. Which means that this particular technology can be defeated more or less for 'free' where 'free' equals hundreds of hours in training to adjust to the subtle ballistic effects of a suppressor and the weight of said suppressor. What can't be reduced for free is the 'zipping' sound from the bullet exceeding the soundbarrier (all common bullets except .45ACP do this). To do that requires the use of special subsonic ammunition with extremely poor ballistics. No more one-shot kills on Taliban fighters from a mile and a half off (second from bottom).
--Ryvar
What if the robot is shot? Even presuming it had a gun it could rotate and fire back nearly instantaneously, if a supersonic round was used, the robot would be lying in pieces on the ground before it picked up sound waves from anything. If the detection system was optical-based, it MIGHT have a chance...
Please help metamoderate.
Just using sound won't do it for detection. Silicing is possible if you keep your bullet from moving more than ~1100fps, and long range accurate fire is still possible with that muzzel velocity if the ballistic coefficient is high enough (think sptitzer/boattail .308 around 275 grains). Even if you stay wtih "standard" ammo, you can alter the sound of it using "silencer" technology, to the point where it isn't really recognizeable as a gunshot.
Don't blame me, I voted for Kodos
It depends on what you mean by "win." You see, for the case in Iraq at the moment their fighters are culturally influenced by Asian and Steppe fighting methods (the history of this stems from Sun Tzu and other cultural writings being moved along the Silk Road and also from Mongol invasions) which posits that running away is in fact winning in their mindset. The western mindset, which you so eloquently put in your example of winning, is about fighting the enemy in a decisive battle and if the enemy runs that is a "win" in our mindset. So at a tactical level our western mindset might see it as a win but for the enemy it is part of a longer term strategic culture that champions running away to fight again another day.
In terms of government hands, there are no doubt many cases where it would be useful for INDIVIDUALS to quickly find out who shot first, so as to lay blame at the very least (1970, 1976, and I thought there was an incident during the revolutionary war, but I can't find it at the moment).
And there are no doubt certain cases where people WANT the government to know where the shot(s) came from (1963).
If I wasn't such an AC I would mod that insightful. So when Bush keeps telling us that our latest offensive has the Enemy of the Day "on the run," which he says quite often, what he's really telling us is that we're still totally screwed in Iraq. Interesting.
The reasons that so many guerilla movements use the AK-47 are because it is cheaper than any other gun (of similar effectiveness), it is easier to maintain (they are notoriously rugged), any idiot can fire one (if I can do it, how hard can it be?), and they are everywhere. You make it sound as though they had a purchasing department. I would be very surprised if they did.
I realize that not all of the potential enemies we might face fall into the guerilla category, but that seems to be our enemy right now.
I cried real tears when Li Mu Bai died.
A good sniper's ranges are listed in miles, not feet.
.50 cal sniper rifles can shoot a maximum of 1500-1600 meters while remaining accurate enough to provide one shot one kill accuracy.
Are you kidding me? "in miles" suggests that you mean to imply that a good sniper can shoot more than one mile. In actuality, a snipers range is measured in meters not miles or feet and several
The military sniper relies on being a sneaky bastard by using natural cover and remaining hidden from view. A good sniper will take one shot, hit his target, and remain perfectly still until he determines that it is safe for him to move or unless he determines that he must move and risk giving away his position.
Granted that they have studied Sun Tzu but it is not culturally or historically embedded enough in the US military to be of much use. It has only been in the last fifty years that Sun Tzu (Or Clausewitz) has been studied seriously by the US military. John Boyd and his acolytes have talked extensively about this. If anything the US military is culturally influenced by the Jominian or "scientific" way of warfare. Jomini was widely read during the Civil war and influenced the Generals at that time and made it essential reading for future officers. That method of warfare sees its evolution into the Operations research and RAND game theorists of the cold war which were all based on a Jominian way of war.
"The Western military mindset is to take and hold territory. A retreat that keeps a force intact as a threat to maintaining control of territory is not a victory for the other side."
Like I said, the enemy thinks otherwise. You may think you are winning by your sets of standards and values, in this case holding territory but for an eastern fighter they really couldn't care that you are holding ground. The enemy has a different set of values, objectives and cultural influence to the western world.
If you're interested in the subject of snipers in general you might want to read this story (and elsewhere) about a WWII Russian sniper named Vasily Zaitsev, and his duel with a German sniper sent to eliminate him. Various versions have the two adversaries stalking each other through the rubble-strewn streets of Stalingrad for days or weeks. The tale is often disputed and could have been Soviet propaganda, but it's a good story. I think it was also the basis for the movie "Enemy at the Gates."