SwarmDrone Package

Senior Design for University of Central Florida. Sponsored By Lockheed Martin

• Implements a drone swarm consisting of 4-5 drones. Incorporates path planning algorithms and object detection. The swarm traverses an urban environment simulated in Gazebo and implemented in ROS Noetic.

Requirements:

A. Install ROS and Environment:

# enable repo
sudo add-apt-repository universe
sudo add-apt-repository multiverse
sudo add-apt-repository restricted

# install dev libraries
sudo apt-get install libsdl-image1.2-dev

# setup source list
sudo sh -c 'echo "deb http://packages.ros.org/ros/ubuntu $(lsb_release -sc) main" > /etc/apt/sources.list.d/ros-latest.list'

# setup keys
sudo apt install curl
curl -s https://raw.githubusercontent.com/ros/rosdistro/master/ros.asc | sudo apt-key add -

# install
sudo apt update
sudo apt install ros-noetic-desktop-full

# install dependencies for building packages
source /opt/ros/noetic/setup.bash
sudo apt install python3-rosdep python3-rosinstall python3-rosinstall-generator python3-wstool build-essential
sudo apt install python3-rosdep
sudo rosdep init
rosdep update
rosdep install --from-paths src --ignore-src -r -y

# auto source whenever bash starts
echo "source /opt/ros/noetic/setup.bash" >> ~/.bashrc
echo "export GAZEBO_MODEL_PATH=$GAZEBO_MODEL_PATH:~/catkin_ws/src/gazebo_models_worlds_collection/models" >> ~/.bashrc
source ~/.bashrc

# setup gazebo
. /usr/share/gazebo/setup.sh

B. Set up workspace:

# Create catkin_ws
cd ~/
mkdir catkin_ws
cd catkin_ws/
mkdir src
catkin build

# clone gazebo models and repo
cd ~/catkin_ws/src/
git clone https://github.com/ros-geographic-info/unique_identifier.git
git clone https://github.com/ros-geographic-info/geographic_info.git
git clone https://github.com/kkerns2/swarmDrone.git

Take all the contents inside swarmDrone and insert it into the src folder, and delete the swarmDrone folder

C. Files that need to be changed:

• If your workspace isn't named catkin_ws, changed the following:

  1. Go to ~/catkin_ws/src/bash_script and modify the following files:

  <ul> <li>drone1_reset.sh</li> <li>drone2_reset.sh</li> <li>drone3_reset.sh</li> <li>drone4_reset.sh</li> <li>drone1_takeoff.sh</li> <li>drone2_takeoff.sh</li> <li>drone3_takeoff.sh</li> <li>drone4_takeoff.sh</li> </ul>

  Change this line with the name of your current workspace

  Before:                                       After:
  source ~/catkin_ws/devel/setup.bash     ->  source ~/name_of_your_workspace/devel/setup.bash 
  

  2. Go to ~/catkin_ws/src/swarm/scripts/ and modify the following files:

  <ul> <li>drone1_client.py</li> <li>drone2_client.py</li> <li>drone3_client.py</li> <li>drone4_client.py</li> </ul>

  Change this line with the name of your current workspace

  Before:                                                                                   After:
  with open('/home/mcp/catkin_ws/src/swarm/waypoints/waypoints_d1.txt', 'r') as f:    ->    with open('/home/mcp/name_of_your_workspace/src/swarm/waypoints/waypoints_d1.txt', 'r') as f:
  

D. Move the bash_script folder where your src folder resides:

catkin_ws/
    -> bash_script
       build
       devel
       logs
       src

<br>

Generate waypoints for phase 2:

Generate the waypoints for each drone using the map created from above explore-lite.

rosrun swarm Image_Tiler_Driver.py

Steps to run the SwarmDrone Package:

Phase 1: Autonomously Map the environment (SLAM):

Steps to run

1. Launch World Simulation with Four Drones into Environment with Hector SLAM enabled:

roslaunch swarm world.launch slam:=true

2. Launch bb8 into the world (arg x and y is for the spawn location of bb8):

roslaunch bb8 main.launch

3. Launch Four Drones Mid-Air:

cd bash_script/

./takeoff.sh

4. Reset Hector SLAM:

cd bash_script/

./reset_scan.sh

5. Launch YOLOv5 for object detection of bb8 for drone1, drone2, drone3 and drone4:

rosrun swarm yolo_slam.py drone1 drone2 drone3 drone4

6. Enable Movebase & Explore Frontiers to allow the Drones to Navigate around the Map autonomously:

roslaunch swarm start_frontier.launch

7. Once you're satisfied with the completed map, saved the map:

rosrun map_server map_saver --occ 100 --free 10 -f map:=/drone1/map ~/catkin_ws/src/swarm/maps/drone1_map

rosrun map_server map_saver --occ 100 --free 10 -f map:=/drone2/map ~/catkin_ws/src/swarm/maps/drone2_map

rosrun map_server map_saver --occ 100 --free 10 -f map:=/drone3/map ~/catkin_ws/src/swarm/maps/drone3_map

rosrun map_server map_saver --occ 100 --free 10 -f map:=/drone4/map ~/catkin_ws/src/swarm/maps/drone4_map

8. Merge your Saved Maps into a Master Map:

Copy the following maps and insert it into: catkin_ws/src/Map-Merge-Tool/build/figure/

<ul> <li>drone1_map.pgm</li> <li>drone2_map.pgm</li> <li>drone3_map.pgm</li> <li>drone4_map.pgm</li> </ul>

cd src/Map-Merge-Tool/build/

./DisplayImage figure/drone1_map.pgm figure/drone2_map.pgm figure/drone_a.pgm

./DisplayImage figure/drone3_map.pgm figure/drone4_map.pgm figure/drone_b.pgm

./DisplayImage figure/drone_a.pgm figure/drone_b.pgm result/urban_result.pgm

Copy drone1_map.yaml and rename it to urban_result.yaml, and leave it in catkin_ws/src/swarm/maps

Edit urban_result.yaml on line 1:

Before:                     After
image: drone1_map.pgm       image: urban_result.pgm

Copy urban_result.pgm and insert it in catkin_ws/src/swarm/maps

Lastly modify ~/catkin_ws/src/swarm/launch/drones.launch

Line 53:    <arg name="map_file" default="$(find swarm)/maps/urban_small.yaml"/>

Change 'urban_small.yaml' to 'urban_result.yaml'

Phase 2: Navigation

Steps to run

1. Launch World Simulation with Four Drones into Environment:

roslaunch swarm world.launch slam:=false

2. Launch bb8 into the world (arg x and y is for the spawn location of bb8):

roslaunch bb8 main.launch

3. Launch Four Drones Mid-Air:

cd bash_script/

./takeoff.sh

4. Launch YOLOv5 for object detection of bb8 for drone1, drone2, drone3 and drone4:

rosrun swarm yolo_nav.py drone1 drone2 drone3 drone4

5. Enable Movebase & AMCL to allow the Four Drones to Navigate the Map:

roslaunch swarm start_move.launch

<br/>