bebop2_autonav

Autonomous control of the Bebop 2 drone to showcase the possibilty of automated material delivery on the shop floor. Authors: Aakash Yadav, Yuhao Zhong, Aditya Yalamanchili and Ravi Srivatsa

Environment setup

Install the following operating system and other packages. It is highly recommended to perform the setup in given order and to test Python/OpenCV before installing ROS

Ubuntu 16.04 LTS Xenial
Python 3+
OpenCV 2
ROS Kinetic Kame

Installation

Setting up environment.

Visit the official bebop page bebop-autonomy

# Install the required packages
$ sudo apt-get install build-essential python-rosdep python-catkin-tools
# Create and initialize the workspace
$ mkdir -p ~/bebop_ws/src && cd ~/bebop_ws
$ catkin init
$ git clone https://github.com/AutonomyLab/bebop_autonomy.git src/bebop_autonomy
# Update rosdep database and install dependencies (including parrot_arsdk)
$ rosdep update
$ rosdep install --from-paths src -i
# Build the workspace
$ catkin build

The build should complete without any errors

Takeoff and Landing using ROS

Caution : Before performing the take off please follow all the safety requirements.

# launching the bebop driver
$ roslaunch bebop_driver bebop_node.launch

Open a new terminal

# takeoff drone
$ rostopic pub --once bebop/takeoff std_msgs/Empty
# land drone
$ rostopic pub --once bebop/land std_msgs/Empty
# emergency command to stop all the operations and free fall drone
$ rostopic pub --once bebop/reset std_msgs/Empty

ArUco generation and camera caliberation

# capture the checkerboard images from drones camera and run the caliberation script
$ cd ~bebop2_autonav/photos_checker_calibrate
$ python CalibrateCamera.py
# this will generate a pickle file to be used while aruco detection
# generating aruco markers
$ cd ~bebop2_autonav/aruco_generation
$ python GenArucoMarker.py

Creating the package (1st time only, not required if this package is already existing)

$ cd ~bebop2_autonav/src/bebop_autonomy
$ catkin_create_pkg motion_plan2 rospy roscpp std_msgs geometry_msgs sensor_msgs
$ cd motion_plan2
$ mkdir scripts
$ catkin make

Running a python script

# place your ROS python scripts at cd ~bebop2_autonav/src/bebop_autonomy/scripts
# install dos2unix to resolve any character mismatch conflict
$ sudo apt-get install dos2unix
$ dos2unix my_file_name.py
# make the python file executible (not required for cpp)
$ cd ~bebop2_autonav/src/bebop_autonomy/scripts
$ chmod +x my_file_name.py
# run the script (launch the bebop driver first)
$ rosrun motion_plan2 my_file_name.py
# for example use below to capture the images 
$ rosrun motion_plan2 imagOR.py

Related Publication(s):

Image-based flight control of unmanned aerial vehicles (UAVs) for material handling in custom manufacturing

@article{zhong2020image,
  title={Image-based flight control of unmanned aerial vehicles (UAVs) for material handling in custom manufacturing},
  author={Zhong, Yuhao and Wang, Zimo and Yalamanchili, Aditya V and Yadav, Aakash and Srivatsa, BN Ravi and Saripalli, Srikanth and Bukkapatnam, Satish TS},
  journal={Journal of Manufacturing Systems},
  year={2020},
  publisher={Elsevier}
}

benbenji/bebop2_autonav