An Autonomous Search and Rescue System (ASARS) created for the Spring 2023 RBE 594 Capstone project.
Clone repository:
cd ~/
git clone https://github.com/poolec4/rbe594_asars_ws.git
cd ~/rbe594_asars_ws/
git submodule init
git submodule update
Manually install dependencies:
sudo apt-get install ros-noetic-husky-simulator
sudo apt-get install ros-noetic-teb-local-planner
sudo apt-get install python3-pcl pcl-tools
These may be needed as well:
sudo apt-get install ros-[distro]-tf2-sensor-msgs
sudo apt-get install ros-[distro]-navigation
sudo apt-get install ros-[distro]-costmap-converter
sudo apt-get install ros-[distro]-mbf-costmap-nav
sudo apt-get install libsdl-image1.2-dev
sudo apt-get install libsdl-dev
sudo apt-get install ros-move-base-msgs
Add exports:
export GAZEBO_MODEL_PATH=$GAZEBO_MODEL_PATH:~/rbe594_asars_ws/src/gazebo_models_worlds_collection/models
export GAZEBO_RESOURCE_PATH=$GAZEBO_RESOURCE_PATH:~/rbe594_asars_ws/src/gazebo_models_worlds_collection/worlds
Build catkin environment:
catkin_make
source ~/rbe594_asars_ws/devel/setup.bash
This will run the UAV part ASARS simulation. It will begin by generating the world and spawning the victims. Then, it will begin the UAV mapping and scanning operations. Once this is complete, the occupancy grid is generated:
roslaunch rbe594_asars main_asars_uav.launch
Then, the AGV section can be run which will begin the path planning process and the AGV will begin to deliver supplies to victims:
roslaunch rbe594_asars main_asars_agv.launch
This will run Hector UAV in a test world and start the teleop menu. It will also open the sensor visualization in RViz. This has the downward facing LIDAR for ground mapping:
roslaunch rbe594_asars hector_world.launch
This will run Husky AGV in a test world and start the teleop menu. It will also open the sensor visualization in RViz. This has the front facing HUSKY_UST10 2D-LIDAR for local planner:
roslaunch rbe594_asars husky_world.launch
Open the test environment with LIDAR unit:
roslaunch rbe594_asars my_world.launch