The Trials Of Quadcopter Graffiti

Last April, graffiti artist [KATSU] strapped a can of red spray paint to a Phantom quadcopter, flew it up against one of the largest billboards in New York City, and pressed a button. Now, [KATSU], [Dan Moore], and Adafruit’s [Becky Stern] are trying to perfect a flying can of spraypaint, and they’ve met with some success and surely many broken props.

The team used an Iris+ for this project instead of the Phantom used by [KATSU] earlier this year, but the principle of the entire endeavor remains the same: fly up against a wall, flick a switch, and watch paint come out of a spray gun. To get the can spraying paint, they modified a can gun to accept a micro servo. This servo is connected to the trigger mechanism of the can gun, and the entire unit is slung under the quad.

Getting a quadcopter to put paint exactly where you want it is hard, even indoors. Luckily, the Pixhawk inside the Iris has sensor inputs and an ‘altitude hold’ mode that can accept a sonar sensor and can be programmed to stay a set distance away from a wall. These sensors are susceptible to interference, and a proper, shielded cable had to be made, but the sensor did work.

Flying the quad did not go as smoothly. The swinging can of paint changes the center of gravity of the quad, and even flying indoors proved difficult. Still, if you’d like to give it a go, [Becky] put up the instructions for their build. You can see the hover attempts in the video below.

61 thoughts on “The Trials Of Quadcopter Graffiti

  1. yeah, smaller paint-cans, closed-loop control and succecissive development: first go with a laserpointer and maybe long-exposure photography. after that’s working, go with more payload?

  2. That whole things doesn’t sound to cool. Some idiot with this spraypainting skyscrapers will certainly not do the drone community any good. It’s not like the FAA likes drones in particular…

    1. A few ideas to help pull off this stunt come readily to mind.
      But I’m loath to post any suggestions as this scheme is so ill-advised :-(
      IMO, not a good idea to be associated with this project.
      Time will tell, I hope the drone community does not become shackled
      due to the fallout from this.

    2. You mad bro? I’ve been dropping porn pics over kindergartens und primary schools for 2 moths already with my hexacopter. Fun times and cops can’t do shit (:

    3. So should we ban baseball bats because they can be used for assault?

      Painting drones could conceivably be used to replace scaffolding and ladders, possibly preventing countless fall injuries. As for graffiti, the law should in all things hold people accountable for their actions rather than try and proactively constrain people into acting correctly.

      1. There is quite a bit of distance between should and will. Baseball bats are not banned (while clubs are) because they are used for a purpose many find desirable. Few find “drones” useful for any purpose and increasingly see them as destructive and potentially deadly as more projects of this sort become publicized.

      1. I’d suggest you try and open a window in a high building, you’ll very quickly understand why using a quadcopter for this is ill-advised…

        hint – strong wind and turbulence from the building

  3. I’m with the “not a good idea” crowd on this one. There may be one or two redeeming features of this effort, but I haven’t found them. Drones/quadcopters don’t need any more bad publicity, and if that is all get out of this, you’re extremely lucky. Nothing is going to attract harsh regulation like drones spraying paint all over the place.

    Remember: just because you CAN do something, doesn’t mean you SHOULD.

      1. I’m not suggesting banning attaching spray cans to quadcopters. I just don’t think it’s a good idea.

        I’m asking those who think it’s a good idea to take a step back and think about how quadcopter graffiti is going look to the authorities. They’re not going to be happy about it. Their response will most likely be harsh and overreaching. That will be bad for all quadcopter users, not just the graffiti types.

        Nobody’s abridging anyone’s freedom to fly a quadcopter yet. But I think quadcopter graffiti, no matter how technically challenging, isn’t a good way to demonstrate how quadcopters can be used responsibly. You may feel differently, and that’s fine. But public opinion, the media and the authorities may react differently, and it would be irresponsible or us not to consider this and act accordingly.

  4. So, what do they do? They damage other’s property without any care, without any feeling for responsibility, without asking the owners if they like that or not. This has nothing, really nothing, to do with any kind of art or technical break through, when somebody smears paint on a freshly renovated wall. I can see this every day, when I walk through the city: On every wall the same, useless, ugly signs are sprayed, nothing more but “I was here and damaged things”. some characters without meaning, like “PAUS:”. Better do not ask, what I would like to happen to such spraying guys.

    This is a really BAD use of technology.

    1. There’s also plenty of fantastic graffiti art in places where people want it and often on a scale where a quadcopter is more applicable than just somewhere you can reach with a spray can. Also what about quads roving round covering up unsightly tagging?

  5. The interesting things for me here are load carrying and accurate position holding. The position holding has to be mostly down to good sensing, the quadcopter can react very quickly but it needs to know exactly how far off the desired position it is.

    Carrying a heavy load slung under the quad like that is bound to upset things without changing the software to expect it but you could likely mitigate a lot of the issues by bringing the weight up in to the centre of the quad where the centre of gravity is expected to be. Either point the booms down to lower the props or even rebuild the quad with a hollow centre that the load can hang in.

    1. Or, don’t hang it. Lose the legs, hard mount the can, and use the can base as landing gear. Then the stability issue reduces to sloshing paint in the can, rather than the sloshing paint and swinging can.

  6. Nothing will change that a) there are people working on this, and similar, projects, and b) any new tech has good and bad uses (with a few exceptions… I can’t think of any good about nerve gas, for instance), and, throughout history, regulators and a certain portion of the public have focused primarily on the negative aspects. This type of project isn’t really the origin of the bad image. The tarnished image has come from things like people flying and crashing in inappropriate places, using the devices to take photos and film where they shouldn’t, and an assortment of other issues. Media outlets then magnify negative incidents for the public, without contrast or context, and, in the end, people that don’t have an understanding of the tech and the difference between appropriate and inappropriate use focus on the negative, and this perception leads to things like this weeks Dilbert storyline.

    The harsh regulation is already there, and has been for a while now. It isn’t strongly enforced, but there have been a number of changes over the last few years, some of which have been mentioned on HAD, and many of which have been discussed in the RC aircraft community and in the technical literature.

  7. I wonder when we’ll see the first sorties between graffiti drones like this and some building’s private defense drones. A disposable quadcopter or two dangling some long pieces of string could stop this craft. It would be entertaining to watch the whole thing go down. Watch out for falling drones!

  8. Nobody’s asked the basic question – is a spray can the best way to deliver a marking material to a surface from a semi-unstable, propwash-affected platform? They’re designed to apply a broadly dispersed, smooth surface coat under somewhat idealized conditions. Something like a paintball system or well designed stream-nozzle (think fire-extinguisher here) would be more effective since it delivers the marking material as an intact unit or linear stream to the target surface.

    1. Sure, but people seem to find it way more entertaining to be dick bags and damage property rather than doing something productive. Would love to spray crap all over these people’s cars/houses/apartments just to see the reaction. I’m betting it would not be positive.

    1. or make the control program accept dynamic weight loading. the weight should follow all principles of momentum and as such could be confided a pendulum as such its rate of swing can be predicted and compensated for.

      1. Unfortunately, not the case, due to the liquid in the can. First approx would be a double pendulum, which is chaotic long term, but short term predictable. The liquid sloshing in the can, though, is more complex than a simple pendulum, and very difficult to model on its own. Add the nonlinear interaction between the copter and a suspended can, and the problem becomes whooly mammoth hirsute. Hard-mount the can to the copter to remove a couple degrees of freedom simplifies it a LOT

  9. Geek graffiti. Great, let’s get the entire maker community looked at as criminals. Again.
    Didn’t we have enough trouble shaking that image off after the 70’s and 80’s? Am I the only one old enough to remember when “hacker” was a derogatory term?

    1. Yep. I’m with the “not a good idea” side on this one. There are so many cool and useful things to do, so why not do them? If you really think this will help develop an ability to safely paint high areas then just make that your project.

  10. Painting from a distance with a pressurised can sounds rather unstable because the PID loop will be constantly trying to fight the intermittent recoil from the spray.
    Maybe the the quadcopter could have feelers pushing against the wall being painted so that it’ll drift less, and also be more stable in the normal direction?

  11. I’m thinking the key might be to squirt the paint. Spray’s too light and easily deflected by the prop wash.

    It might also be possible to put the paint in a plastic bag (wine-in-a-box style) and squirt it using a small electric pump. This could mitigate the sloshing, and you might also come out ahead on weight by eliminating the need to carry pressurized propellant and a steel can.

    Seems to me I remember reading about quads being used to look for trees infested with some invasive pest. If found, they’re paint and GPS tagged. A ground crew is later sent to burn them, using GPS to find the approximate location, and the paint to identify the exact tree. Wish I could remember more detail. That at least is a useful application. Poorly executed graffiti is not.

  12. The first thing that occurs to me is to create an extended nozzle that points in line with the can. That would prevent it rotating around the center of the can’s gravity, and moving the spray head out past the rotors would allow more precision.

    The second thing would be to mount a camera & computer (with picture) that watches what has been sprayed already, and trigger the nozzle whenever it lines up with a ‘pixel’ that hasn’t already been sprayed.

  13. I reckon you’d need something like the nozzle extender on a WD40 can to extend the output of the can beyond the downdraft of the rotors, otherwise the can’s fighting the rotor downdraft.

  14. Proving once again that gigantic assholery can still be hilarious.

    I’m imagining popular targets for vandals buy “guard drones” that try to knock graffiti bots out of the sky. I *really* want to see this happen. If I’m going to live in a dystopia, it might as well be an awesome cyberpunk one.

  15. Those first few flips were do not look (or sound) like they were cause by the swinging paint can. Rather not using the proper orientation of propeller for each motor.

  16. Forget about graffiti. Use the tech to spray and remove wasp’s nests. Now that would be a productive and very satisfying purpose for making such a machine. . :-)

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.