USMA Swarm Capstone - Hivemine

Contributor

• Dongbin Kim ([email protected])
• Josiah Steckenrider ([email protected])

Requirements

This package requires two computers (one PC or Laptop and one Single Board Computer).

• Ground Station Computer (GSC): Ubuntu 20.04 LTS with ROS Noetic, PX4, MAVROS, QGroundControl

• Drone Onboard Computer (DC):

  • Software: Ubuntu 20.04 LTS with ROS Noetic, MAVROS
  • Hardware: ODroid XU4

• Check GCS_DC_Network_Setup.md to establish SSH connection over Wifi

Ground Station Computer (GSC) Preparation

• Install Ubuntu 20.04

• Install ROS Noetic (ros-noetic-desktop-full)

• Install QGround Control

• Install following packages to meet dependencies (PX4, MAVROS, etc)

    (PX4)
    git clone https://github.com/PX4/PX4-Autopilot.git --recursive
    bash ./PX4-Autopilot/Tools/setup/ubuntu.sh --no-sim-tools --no-nuttx
    sudo apt-get install protobuf-compiler libeigen3-dev libopencv-dev -y
    (MAVROS)
    sudo apt-get install ros-noetic-mavros ros-noetic-mavros-extras ros-noetic-mavros-msgs
    wget https://raw.githubusercontent.com/mavlink/mavros/master/mavros/scripts/install_geographiclib_datasets.sh
    sudo bash ./install_geographiclib_datasets.sh
    (OpenCV)
    sudo apt install python3-opencv
  

• Install following python packages (pip)

    sudo apt install python3-pip
    pip install pandas
  

• Reboot the PC

    sudo reboot
  

• Download and unzip the USMA_VRTeleM pacakge in your ROS workspace, then build the code

    cd catkin_ws/src
    unzip {the package}
    cd ..
    catkin_make
    cd /src/flight_test/scripts
    chmod +x 
    cd ../src/vision_processing/scripts
    chmod +x        
  

Drone Onboard Computer (DC) Preparation

• Install Ubuntu 20.04

• Install ROS Noetic (ros-noetic-desktop-full)

• Install following packages to meet dependencies (MAVROS, etc)

    (MAVROS)
    sudo apt-get install ros-noetic-mavros ros-noetic-mavros-extras ros-noetic-mavros-msgs
    wget https://raw.githubusercontent.com/mavlink/mavros/master/mavros/scripts/install_geographiclib_datasets.sh
    sudo bash ./install_geographiclib_datasets.sh
    (OpenCV)
    sudo apt install python3-opencv
    sudo usermod -a -G video $LOGNAME 
  

• Install following python packages (pip)

    sudo apt install python3-pip
    pip install pandas
  

• Reboot the PC

    sudo reboot
  

• Update and upgrade the debian pacakges

    sudo apt update
    sudo apt upgrade
  

• Download and unzip the USMA_VRTeleM pacakge in your ROS workspace, then build the code

    cd catkin_ws/src
    unzip {the package}
    cd ..
    catkin_make
    cd /src/flight_test/scripts
    chmod +x 
    cd ../src/vision_processing/scripts
    chmod +x 
  

Operation in the air

Make sure to have both GSC and DC are on the same network over Alfa Wireless USB!

Pixhawk Side (not DC)

• Open QGroundControl

• Set the following parameter

    MAV_1_CONFIG = TELEM2   # If TELEM2 is not an option enter 102
  

• Reboot Pixhawk, set the following paramters

    MAV_1_MODE = Onboard   # If Onboard is not an option enter 2
    SER_TEL2_BAUD = 921600
    MAV_1_RATE = 0
    MAV_1_FORWARD = Disabled
    MPC_LAND_SPEED = 0.6
  

GSC Side

• Open QGroundControl to check the drone status over Telemetry

• Open Terminal (T1) - Connect DC to Pixhawk over MAVROS

    ssh [email protected].{tail_number}
    source /opt/ros/noetic/setup.bash
    source ~/catkin_ws/devel/setup.bash
    
    (Do the below 3 lines to set the serial port to USB, once you do this, skip this command for the next operation)
    roscd mavros/launch
    sudo nano px4.launch
    **edit this line** <arg name="fcu_url" default="/dev/ttyUSB0:921600" />
    (Save and exit)
    roslaunch mavros px4.launch
  

• Open Terminal (T2) - Send the command to the drone (1: Arm, 2: OFFBOARD (Takeoff), 3: Landing, 4: Disarm, 5: Mission)

    ssh [email protected].{tail_number}
    source /opt/ros/noetic/setup.bash
    source ~/catkin_ws/devel/setup.bash
    rosrun flight_test hm24_da_mode
  

• Open Terminal (T3) - Receive the commmand from hm24_da_mode, control the drone

    ssh [email protected].{tail_number}
    source /opt/ros/noetic/setup.bash
    source ~/catkin_ws/devel/setup.bash
    rosrun flight_test hm24_response
  

• Open Terminal (T4) - Run Lissajous Pattern flight missions. You can also adjust the flight altitude

    ssh [email protected].{tail_number}
    source /opt/ros/noetic/setup.bash
    source ~/catkin_ws/devel/setup.bash
    (You can adjust the home position)
    roscd flight_test/scripts/
    nano Lissajous_pattern.py
    **Line 26-28 (below)** 
    self.home_lat = 41.3909084
    self.home_lon = -73.9529137
    self.home_alt = 15.0
    (Change, Save, and exit)
    rosrun flight_test Lissajous_pattern.py
  

• Open Terminal (T5) - Run computer vision processing

    ssh [email protected].{tail_number}
    source /opt/ros/noetic/setup.bash
    source ~/catkin_ws/devel/setup.bash
    (Sanity check! You have to comment "imshow" because Odroid does not have display connected)
    roscd vision_processing/scripts
    nano img_process_node.py
    **Line 63, add '##' to comment the line** ##cv2.imshow("images", combined_frame) 
    (Save and exit)
    python3 img_process_node.py
  

• When all the terminal are running without any problem, go to T2, do the following

  • Type 1 and enter: Arm the drone
  • Type 2 and enter: OFFBOARD mode and Takeoff
  • Check if drone is hovering at the home position with desired altitude after Takeoff
  • If the desired altitude is not matched to what you see from QGroundControl, go to T4, adjust the height by adding numbers. This will increase/descrease the altitude in (m)
  • If the altitude looks good, go back to T2, type 5 and enter! The drone will begin the Lissajous pattern mission
  • Make sure to keep the setmode stick to Position Hold and throttle stick in the mid-way on your transmitter.
  • After the mission, use your transmitter to bring back the drone for safe landing.

ENJOY!