$10 Multicopter Safety System

A simple parachute deployment system for machines having "brick-like" flying characteristics when losing power.

Similar projects worth following
As an aeromodeller, I witness lots of crashes. Sometimes, with my own models. Results vary from minor scuffs to major destruction.
With an airplane, you still can try to land without much damage if something goes wrong with the motors. With a multicopter, this can get a little trickier as the machine will drop like a stone if it ever got to lose power. Autorotation may not be enough to stop the thing, so I started to think about something safer: parachutes.
Commercial systems are available, but they seem way too expensive, plus I wanted to make something simpler. Besides, I needed a project to put my 3D printer to good use.

I want to use a device which doesn't require the needs of pyrotechnics, something fast to reset. I'm thinking about a spring-loaded ejector, the hatch being closed by a 9g servo. The servo itself should be triggered by a simple ATMega328 checking an accelerometer.

Here's a small presentation video, in the most terrible english you've ever heard.

  • 1 × Tricopter I'll be testing my device on a MAKE: Tricopter built as described in the "Maker Hangar" series
  • 1 × 3D Printer I use a K8200 Velleman printer, available as a kit
  • 1 × Arduino Pro Micro Or some sort of AtMega328 clone board, with the Arduino bootloader
  • 1 × Accelerometer Probably the MMA7361
  • 1 × Lithium battery This thing has to be totally independent from the main system, hence a small li-po battery

View all 6 components

  • First enclosure

    M.daSilva08/13/2014 at 22:41 0 comments

    The project isn't moving on as fast as I wanted to, others things got in the way...

    I finished the print of the first casing for the parachute, based on rough dimensions I took from the parachute, and I made it waaaay too big... Lesson learned.

    It took 18 hours to print, but it allows me to make sure my system works correctly before making the definitive casing.

  • First parachute tests, description video

    M.daSilva08/07/2014 at 12:36 0 comments

    I just came back from my testing facility (ok, I just tossed the parachute from the highest window of my house), and the results seem pretty interesting: with a 500g load, the parachute slows down to about 2.5m/s, and with a kilogram, the speed is still decreased to 6-7m/s, which is slow enough to avoid major damage.

    Now that I know how big the parachute needs to be, I'll fold it up correctly and size the compartment. 

    I also made a small presentation video to hit the requirements for the Hackaday Prize. Gosh, listening to myself makes me realise how terrible my english actually is...

  • Parachute tests

    M.daSilva08/07/2014 at 11:24 0 comments

    After sizing the parachute I needed, I took some plastic sheet I had lying around and cut it into a circle of 45cm in diameter.

    I then attached 4 strings, making them into eight parachute lines:

    The mock-up is ready for some testing. It will allow me to check whether I have enough suface to slow it down, and size up the compartment that will hold it when folded.

  • First thoughts

    M.daSilva08/01/2014 at 19:46 0 comments

    I want a system that has the following requirements:

    -Be able to install easily on a any kind of multicopter (sub-750g)

    -Be lightweight (Under 50g would be great, I still need to figure out how light this actually needs to be)

    -Be cheap (If not under $10, at least under 30 bucks)

    -Be totally independent from the main system: I want something that deploys the chute when it detects an abnormal drop in altitude (or a g-variation, I have both sensors)

    The first thing I have to figure out, is how big the chute needs to be. If I follow a matching table (on page 17 of this PDF: ), I need a spherical parachute of 43.8cm in diameter to slow down a 1kg body at 10m/s. Now, we will be braking only 750g, but I want it to be slowed down a bit more, so I'll round  this diameter up to 45cm.

    With this in mind, I'll first make a mock-up of the chute using simple trash bags, to get a feeling of how big the compartment needs to be.

View all 4 project logs

Enjoy this project?



zakqwy wrote 08/06/2014 at 14:40 point
If you can do this for $10 (or even ten times that) and it works somewhat reliably and doesn't weigh much, I'll integrate it with my Prize entry ASAP. Great idea, there is a HUGE need for this!

  Are you sure? yes | no

Liberty1100 wrote 08/06/2014 at 14:01 point
Well, maybe minimize your strings. This is a emergency parachute, so success would to prevent major damage. Having a chute that is more cloth with less strings might be more reliable to deploy than say the traditional chute.

  Are you sure? yes | no

Adam Fabio wrote 08/04/2014 at 05:06 point
Thanks for entering your Safety System in The Hackaday Prize! As the author of droning on, this is right in my wheelhouse! People have been working on parachute systems for years, but without any widespread success - it seems simple enough though. It seems like the hard part is getting the parachute out fast enough to avoid it getting tangled in the props.
One suggestion - you can monitor the radio system for pulse train. Depending on setup, a lack of a pulse train (radio off), or a preset failsafe value could be used to fire off the chute.

  Are you sure? yes | no

M.daSilva wrote 08/05/2014 at 09:29 point
Thanks a lot for the input! I was thinking about linking it onto the receiver in some kind of way, this is a great idea!

Yeah, my main concern is about the ejection itself: I'm afraid of getting a big ball of strings instead of a nice deployment...

  Are you sure? yes | no

flashcactus wrote 04/22/2016 at 00:13 point

re: gwtting a ball of strings. look at how real emergency chutes are packed, in paragliding for example. Do the same. Also, I suggest you use (synthetic) fabric for the chute. It's lighter & packs more tightly, unfolds faster and has almost identical, if not better, aerodynamical characteristics compared to plastic.

  Are you sure? yes | no

flashcactus wrote 04/22/2016 at 00:10 point

Prop avoidance will be a bit hard, yes, but not too hard, especially if you add some sort of motor-cutting function and activate it, say, 200ms before deployment.

  Are you sure? yes | no

Similar Projects

Does this project spark your interest?

Become a member to follow this project and never miss any updates