Skip to content


Subversion checkout URL

You can clone with
Download ZIP
Code for messing with an AR drone
JavaScript CoffeeScript C++
branch: master

Merge pull request #2 from terenced/master

Bug creating the camera stream
latest commit 9b414a71c9
@contra authored
Failed to load latest commit information.
.gitignore clean
package.json droneit

This is a collection of utilities related to the AR drone


drone augmentation can be found in lib/


Adds landed, landing, hovering, flying, batteryChange, and altitudeChange events to the drone. Being pulled into ar-drone core so will be removed


returns battery level to callback


disables emergency mode on start and adds error handles to prevent your app from crashing and having your drone fly off into space


Records raw video data to location (HD with higher framerate)

You'll need to run ffmpeg -i recording.h264 -vcodec copy recording.mp4 to convert it into an mp4. record will append video to your existing one so you can keep a log of your drone's entire history easily


Adds controls via stdin to the drone - look below for these


Adds a processing step to PNG frames for running facial recognition. drone.faces will emit a data event with a PNG buffer that has circles drawn around faces


Depends on opencv for facial recognition

Run coffee to control the drone from the command line with WASD


w - forward
a - left
s - backward
d - right

arrow up - up
arrow left - spin left
arrow down - down
arrow right - spin right

<space> - land/takeoff
x - hover
r - disable emergency mode
1/2 - flipAhead/flipLeft

You can view the camera from localhost:8080

Installing node on the drone

(thanks felixge for doing the hard parts on this)

Run this command while connected to the drone's network

npm run-script deploy


Do you want your drone to control itself? Install ar-drone on it with

npm run-script setup

Now you can go into /data/video/labs where you have ar-drone and a repl.js to launch a repl. Run node repl.js and you're on your way. You can write any ar-drone scripts and run them from the drone itself now.

Now you can telnet into the drone and type node to run the repl

Something went wrong with that request. Please try again.