Environment setup

Development container setup

To make it possible for you to develop your own code on the drone, you need to set up a docker workspace from the source code of the repository pidrone_pkg.

1. Run ssh duckie@[yourdrone].local from your base station to ssh into your drone. The password is quackquack.

2. Clone the repository to your Duckiedrone’s SD card:

   mkdir -p catkin_ws/src && cd catkin_ws/src
   git clone https://github.com/h2r/pidrone_pkg && cd pidrone_pkg
   

   Note

   The next step will take a long time because it has to download all the dependencies to build the image. Make sure your Raspberry Pi is plugged into an external USB power supply.

3. Finally build the Docker image needed to run the software (which is back on an older version of ROS, ROS Kinetic).

   rake build && rake create
   

4. You also need to download the raspicam_node package to be installed later:

   cd ~/catkin_ws/src/
   git clone https://github.com/UbiquityRobotics/raspicam_node 
   

5. You can start the container and go inside it by running, from the pidrone_pkg directory:

   cd ~/catkin_ws/src/pidrone_pkg
   rake start
   

6. Once in the container, you need to install the pidrone_pkg and raspicam_node ROS packages. To do this execute:

   cd ~/catkin_ws
   catkin_make && catkin_make install
   

7. Now that all the packages are installed, to access the workspace you will use to control the drone, run:

   screen -c pi.screenrc
   

This will start a screen session with each of the ROS nodes needed to run the drone and make it fly.

Note

Anytime you’ll want to interact with the drone software you will first need to start the container by running rake start from the ~/catkin_ws/src/pidrone_pkg directory and then, inside the container, running screen -c pi.screenrc.