Forget DIY Drones, How About DIY Armed Drones?


DIY armed drone

Remotely piloted drones are all the rage these days, and have become must-haves for discerning government agencies at the local, state and federal levels. Our fearless public servants acquire them for purposes of snooping and tracking and, of course, for one-upsmanship on other agencies. Under current law, private domestic use of drones is severely restricted until 2015, although hobbyists have taken to them. Of course, the use of armed drones, which has become so controversial and lethal in government hands, will remain beyond the reach of private parties— Oh wait. No, it won't. In fact, armed drones (of a sort) have already been created by DIYers. At least, one has. And yes, it's cool.

The folks at Dangerous Information constructed a DIY drone with six rotors as so many hobbyists have done. Then, just to see if they could, they mounted a paintball gun on the drone that could be fired by the operator. In his video demonstration, "Milo Danger" shows the drone flying and firing at stationary targets. Jay Stanley, a senior technology policy analyst with the ACLU reports that "Milo" called into a radio show on which he was a guest:

Although not shown on the video, on the radio call-in "Milo" said that he had used it to shoot not only at stationary targets but also at live human beings. In response to a somewhat incredulous question from the host, he assured us that they were fully consenting volunteers. In general, he appears to be making an effort to spark awareness and discussion of the technology's potential in a responsible manner.

Dangerous Information's drone is basically a proof-of-concept piece, but it's one that apparently works just fine, and points to a very … interesting future.

NEXT: Intelligence Officials Still Won't Detail Domestic Surveillance

Editor's Note: We invite comments and request that they be civil and on-topic. We do not moderate or assume any responsibility for comments, which are owned by the readers who post them. Comments do not represent the views of Reason.com or Reason Foundation. We reserve the right to delete any comment for any reason at any time. Report abuses.

  1. I hope Milo isn’t on federal probation.

    1. Double secret probation…

  2. Cops shoots unarmed man eleven times. Nothing else happens.

    1. Wife is lucky that guy emptied his clip into her husband, or she probably would have gotten it, too.

      1. Probably? She disobeyed a command! That’s a capital offense!

        1. with the cop saving the state the price of a lengthy trial.

    2. The comments don’t give me much hope in humanity.

  3. There better not be any foreign nationals reading Dangerous Information, because that would be a munitions export.

  4. Forgetting the Terrorists (who are almost certainly already working on this) I can imagine that there are some gentlemen in Sicilly and Russia with some scores to settle thinking about hiring themselves a geek to build something like this.

    It is afterall the perfect murder weapon, no prints, no need to be physically present would make it virtually impossible to prove who killed the target, even if they managed to capture the drone.

    1. I’m pretty sure the frequencies that civilian remote controlled vehicles are allowed to use have limited range and probably require line of sight, requiring the operator to be nearby and in plain view of any witnesses.

      1. My housemate is really into r/c planes and gave me the long course after I showed him the Burrito Bomber. The autopilot on these things has come a long way. Mulitple navigational checkpoints, RTB, the whole 9 yards. No LOS required as long as the GPS is functioning.

      2. Yes, you need LOS to the drone, not LOS to the target.

        Target is X, Drone is Y, Assasin is Z


        Now put Z inside a car sitting in a parking lot with several buildings between him and X

        1. That’s thinking!

          1. Comes from playing too much Shadowrun 🙂

      3. And I’m pretty sure that anyone willing to kill someone with an armed drone isn’t going to bother worrying about violating the local version of the FCC’s regulations.

        Also – set up the strike location in a place with decent free wi-fi and control the drone over the internet.

    2. “virtually impossible to prove who killed the target, even if they managed to capture the drone.”

      Drones have omerta too?

    3. Once the mission is accomplished, the drone could pretty easily be set up to self-destruct. Yeah, it wouldn’t disappear, but it could be reduced to a smoldering pile of melted plastic goo with metal wires here and there. Not a lot of fingerprints or other forensic information to be gleaned from it, and no new information, either.

      “He was killed by an armed aerial drone!” Yeah, no shit.

      And prior to “mission accomplished”, there’s no good proof of the drone’s intentions, especially if it is set up to fry its own memory quickly enough.

      1. But you run into the same problems that bomb builders have – its impossible to destroy everything and police have pretty good forensics when it comes to discovering the “signature” of the maker.

        Though that does require the maker to create multiple drones used in crimes to be effective.

        And surviving components can be traced back to manufacturers with purchase records.

        1. Ted Kaczynski had a nearly-20-year career as the Unabomber, and was finally found because his brother led the FBI to him.

          And he deliberately left “signatures”.

          I have no plans to use drones to attack anyone, nor to follow in Ted Kaczynski’s footsteps. However, if someone did, I wouldn’t count on the TV show forensics stuff to stop him too soon.

          There’s a difference between forensics used to prove someone’s guilt in a court case, with little time crunch, and what it takes to stop someone like him before he claims many victims.

          Bottom line? Someone who did it would likely be found guilty in court, once caught. And he’d be likely to be caught, because, like Kaczinski, someone would turn on him eventually. However, the police might have a very difficult time actually stopping him quickly.

  5. Distributed murderdrone technology. Now I can go armadillo for real.

    1. “Now I can go armadillo for real”

      Squished on a Texas highway? I’d rather you didn’t do that, H&R needs all the good commenters it can keep.

      1. Heh. We’ve got ’em here in north FL, too. I saw a whole family earlier this year. Mama and 4 mini-tanks. But more seriously, I took the term from one of V. Vinge’s short stories in Across Realtime called “The Ungoverned”, in which a bunch of anarchist farmers and mercenary cops (as in, contracted as a backup defense force by the anarchist farmers) take on an army and win. Farmers holed up and fighting back are referred to as having gone armadillo.

  6. Why would I want one of these? My first solo project at work was helping a start-up program their controllers for these.

    1. Am I wrong in thinking that the recoil from firing an actual gun round from this sort of platform is probably going to tear the wings off it?

      1. I’m sure it wouldn’t be that difficult to rig up something, perhaps some springs, to absorb the bulk of the energy.

      2. As long as you have a reasonably heavy handgun shooting a 9mm Luger or so, the recoil won’t be much greater than a lightweight paintball gun. And the drones being discussed don’t have wings.

        1. You can put recoil compensators on there, too. Either gas ports, or go with one o’ them counterweight thingies. Or both.

          1. Between a big muzzle brake and a suspension system, it’s no sweat.

            Manageable .50 BMG handguns have been built. It still transfers a lot of recoil to the shooter, but similar systems on low-recoil varmint rounds like .223 or .204 Ruger would make for a long-range weapon a small drone could handle.

            Yes, I know about firearms. I did write .50 BMG, I did mean .50 BMG, and I know exactly what the .50 BMG cartridge is. 🙂

            Here’s an example: http://www.youtube.com/watch?v=Wb9D4Gb5dMI

  7. These guys are 8 years behind. San Andreas had homemade killer RC planes back in 2004.

  8. There’s going to be a market for these soon enough, and it won’t be like traditional RC airplanes. Hooking one of these to an iPhone and utilizing a nav system in conjunction with an elevation program and you’re all set without having to be anywhere near line of sight.

    Hell, I bet there are people right now on here that could set it up. Now the trigger mechanism might be a little tricky with the recoil, as mentioned above, but I’m sure there’s a simple solution. How about one that let loose a bomb of sorts as opposed to using a gun? That would make it a lot easier to use and likely more accurate.

    1. Why can’t the drone BE the weapon? Load the thing up with C4 or something, fly it to your target, and blow the whole kit and kaboodle.

      1. Drop some ricin in a few spots on the way.

        Really makes you think about the money we’re wasting on ruining our lives with the TSA…

      2. ‘Cause then its not a drone, its a missile.

    2. or don’t worry about it. kamikaze.

  9. There’s really nothing new about this. People have been remotely operating electric powered model planes via live video feed for years now. And I’m certain some of them realized they could carry out an assassination by using those planes in a kamikaze mode with an explosive device.

    1. And I’m certain some of them realized they could carry out an assassination by using those planes in a kamikaze mode with an explosive device.

      More like every single one of them, within minutes of getting off the ground.

      1. More like that was the motivation to build every single one of them in the first place.

    2. Yeah, back in the 1970s, when I was a kid and a member of the NAR (National Association of Rocketry, a club for model rocket hobbyists that offers liability insurance among other things), you had to sign that you would not put a warhead on any rocket, to be a member (and especially to get the insurance).

      Definitely not a new idea…

  10. Let me know when the drone packs a nuclear-powered MW UV laser. Interesting tech in Ing’s stories.

    Though I have wondered why groups like, e.g., organized crime, haven’t taken Mad Scientist’s idea and run with it for certain hard targets? Then again, you don’t read about too many assassinations with EFP’s, and those aren’t supposed to be that hard to make. I guess you could combine the two concepts, sort of a DIY ATGM.

    1. “I don’t want hear Wales dead. I want to see Wales dead.” Something like that.

      Not as much of a concern to Al Qaeda as to La Cosa Nostra.

  11. Just wait until they have miniature drones the size of a fly, equipped with cameras… celebrities will have nowhere to hide.

  12. I’ve told this story a bunch of times here but I’ve seen a prototype at a table at the Knob Creek machinegun show of a .22 belt fed solenoid fired MG for mounting on lightweight drones.

  13. I’ve seen several videos of quad-rotors being used to pick up building blocks and place them into structures.

    So an assassination bot only needs to be able to pick up an explosive device, loiter in a target area, and then drop the explosive device at the opportune moment.

    1. “The creation of perfection is no error.”

    2. Problem with explosives is that in order for them to be truely effective the end up being large enough to cause significant collateral damage.

      A drone firing a bullet however just needs to be a stable firing platform with a sufficient targeting system to reliably hit the target from say 100 meters

      1. You have a huge latency issue with sending video to a remote operator who makes the decision to shoot and then commands the drone to fire a weapon.

        I don’t see an effective way for a small drone to stabilize it’s attitude and heading well enough to sight a rifle on a target and then have a remote operator fire a weapon.

        1. Didn’t you see the drone in the picture?

          It is a hover drone, it can literally hover in one spot.

          As far as latency, Not an issue, we’ve had real time data feeds for quite a while now, not really all that hard to do.

          You also don’t worry about sighting the firearm, you simply design it such that the targeting camera points exactly where the weapon does then let the controler aim it for himself

          1. That doesn’t even account for the facial recognition and other “augmented reality” technology that’s readily available and so easy that Facebook does it for giggles.

            1. (Not saying I’d trust facial recognition, but the code to lock onto a known object visually, using a camera, is readily available. A remote operator could easily find and designate the target, and the computer could do the rest, even with a moving target and a moving drone.)

          2. so many errors, so little time.

            To hit a target from 100+ meters, the aiming accuracy has to be small fractions of a degree in azimuth and elevation.

            A small drone, in any realistic atmospheric conditions, will not be able to hold still enough to steadily point a gun even if there was zero latency in the system.

            With a human in the loop, you have to transmit video (high bandwidth requirements), present the video to the “shooter”, let the “shooter” command the drone to line up the weapon on the target, and then “pull the trigger”. The human element of the system will add latency of around a second all by itself (3/4 of second between the decision to move a muscle and the muscle actually moving). The total lag time between the moment when the gun is lined up on the target, the video is streamed and presented to the shooter, shooter commands the gun to fire, and the time the drone receives command to shoot will be seconds. The gun will no longer be pointed in the proper direction.

            This is not doable.

            1. You’re wrong.

              A human can’t do it, but a computer can “lock on” easily, using a video camera.

              You’re assuming no computer assistance. But the fact is, we’ve been flying fighter planes that could not be flown by human pilots, since the 1970s, using active systems to stabilize them. The F-117 is obsolete and they’ve been decommissioned, but they flew many successful combat missions, where they destroyed defensive RADAR systems, during the Gulf War 20 years ago already.

              ABS and stability control are standard features in cars.

              Active stability systems driven by computers (with clock speeds in the Gigahertz now, nothing like a human) are old technology, and available cheap.

              1. You’re wrong.

                It’s possible. But my 27 year career in engineering safety critical systems (including a 7 year stint writing aircraft simulation software) gives me some insight to the problem.

                1. It won’t work well in high winds, that’s true.

                  But in somewhat favorable conditions, stabilizing the aircraft, plus stabilizing the gun barrel, can keep the gun on target long enough to get a shot off.

                  A human can’t do it because of the latency involved, not to mention human reaction time. You’re right. But it’s entirely possible to stabilize the platform and gun long enough to fire a shot, especially since the range involved needn’t be extreme.

                  1. A computer may be able to lock on, true, but the coarseness of the control system, never mind sensor limitations—range finding and wind doping are going to be fun—are probably going to preclude MOA accuracy from a hovering drone. Depending on the range we’re talking about, this may not be an issue. If we’re talking anything over, WAG, 200 m, it will be.

                    1. On the other hand, if you’re shooting from 200 m then you really don’t need the drone in the first place.

                    2. Exactly. You’ve got a remotely piloted drone. Fly that thing up to like 20 feet away, and open fire. The azimuth angles get a lot bigger.

            2. a 3′ target area would be sufficient for a 90%+ kill probability with a .50bmg aimed at center mass, that translates to a targeting accuracy of 0.1 degrees at 300 meters and 0.35 degrees at 100 meters and 1 degree at 60 meters.

              That said even if it were not possible to have a direct human link it should not be difficult for the operator to highlight the target point, issue the order to fire and then have the onboard computer “pull the trigger” once it detects the aim point to be within a maximum distance of the aim point.

              There is no technological reason why you could not generate 99% accuracy at 100 meters with this type of arrangement using off the shelf commercial technology and I still maintain 300 meters should not be difficult.

  14. I saw HD camera drones at Costco for $300.

  15. Actually this entire thread shows just how lacking in either imagination or honesty the US Military Brass is.

    So the entire goal behind their murder drone program is supposedly to kill Al Quaeda leaders and all those innocents killed are supposed to be unfortunate accidents right?

    Given that there is no legitimate reason why they could not make a drone similar to the one pictured above that fires a .50 BMG accurately enough to hit a target from 300 Meters, then have that drone be delivered by standard Predators.

    The mission profile would work like this.

    Predator is loitering on station at 20,000 feet with “Drone Pods” attached to the weapons Pylon. A target is identified and the Predator is dispatched. At the range of 5 miles from the target a drone pod is dropped. The pod is designed to slow shed speed to a safe speed for drone deployment and then deploy it. The Sub drone is then guided to a firing position via dataink with the predator. Once in firing range the target is acquired and killed. As an enhancement the drone can carry 5 rounds if multiple targets need to be killed in a single attack.

    Once the controler determines that the target is either dead or escaped and is no longer a viable target the drone self destructs.

    Realistically there is no reason why a system like this should cost more than a $4 or $5 thousand dollars per drone which compared very favorably to the $68,000 a Hellfire missile costs and risks of collateral damage would be all but eliminated.

    1. But they’re not just interested in murder, they’re interested in creating terror in the general population. They’re not murder drones, they’re terror drones.

  16. 300 meters is close enough that you could use standard steel core AP .30-06 ammunition for even hardened targets. A sufficiently-accurate .30-06 barreled action is a few hundred bucks, tops, and with a big muzzle brake, wouldn’t have any recoil to speak of (an issue with a lightweight drone). .50 BMG hardware is expensive, but not necessary.

    Granted, it would cost the Pentagon $250K per mini-drone, but a creative hobbyist could do it all for a grand.

    1. Previous post @Rasilio

    2. Yeah I specified .50 BMG more to rachet up the probability of a kill on a peripheral shot because I assumed that the accuracy would not be good enough to reliably hit inside a 6 inch area in center mass.

      If the accuracy were greater then obviously 30-06, .308, or even possibly .223 would be preferable

      1. .50 BMG? You trying to take out an armored vehicle?

  17. Heh. Nice cite of Dean Ing, whose novels tracked reality
    with eerie accuracy. Also, a quote from William Gibson:
    “I have seen the future of armed response, and it is not human.’

    The drones will probably use gyrojet projectile weapons for their
    zero recoil and light weight, and more ingenious means of attack
    against armored vehicles.

    Twenty-five years ago, on the Byte Information Exchange (BIX)
    this topic on one thread included the use of a shaped charge
    explosive introduced into the barrel end of a tank main gun
    to explosively weld a circumferential copper ring to the bore.
    Naturally, the delivery drone was named ‘Spike’. 🙂

  18. is there a link to the actual kit / plans on how you built the drone? I am with an airsoft team, and we would love to have something like this in our arsenal. if it is possible you can view my email from the back end, and send to me privately, I would not release any of the specs / design…

Please to post comments

Comments are closed.