lonewolf

Repository for the bachelor thesis: ROS simulated world for ATV and finding optimal SLAM algorithm for generated point cloud

If you are running this on NVIDIA AGX Xavier, head over to the .devcontainer for the setup guide.

Prerequisites

• ROS 2 Foxy installed on Ubuntu Linux 20.04 or newer

• You have installed Gazebo on your computer: sudo apt install gazebo11

• You have installed the required packages:

  • Ceres-solver:

    # Install Dependencies
    sudo apt update -y && sudo apt upgrade -y
  
    sudo apt-get install libgoogle-glog-dev libgflags-dev -y
    sudo apt-get install libatlas-base-dev -y
    sudo apt-get install libeigen3-dev -y
    sudo apt-get install libsuitesparse-dev -y
  
    # Download
    cd /tmp
    CERES_VERSION="ceres-solver-2.0.0"
    CERES_ARCHIVE="$CERES_VERSION.tar.gz"
    wget http://ceres-solver.org/$CERES_ARCHIVE
    tar xfv $CERES_ARCHIVE
  
    # Install
    cd $CERES_VERSION
    mkdir build
    cd build
    NUM_CPU_CORES=$(grep -c ^processor /proc/cpuinfo)
    cmake ..
    cmake --build . -j $NUM_CPU_CORES
  
    sudo apt install checkinstall libssl-dev -y
    sudo checkinstall --pkgname ceres-solver
  

  • g2o

    git clone --recursive [email protected]:RainerKuemmerle/g2o.git
    cd g2o
    mkdir build
    cd build
    cmake ../
    sudo make install
  

  • gtsam.

  sudo add-apt-repository ppa:borglab/gtsam-release-4.0
  sudo apt update
  sudo apt install libgtsam-dev libgtsam-unstable-dev
  

  • Necessary ROS Foxy packages:

  sudo apt install ros-foxy-gazebo-ros-pkgs
  sudo apt install ros-foxy-velodyne 
  sudo apt install ros-foxy-teleop-twist-keyboard
  sudo apt install ros-foxy-ros2-control
  sudo apt install ros-foxy-ros2-controllers
  sudo apt install ros-foxy-velodyne-gazebo-plugins
  sudo apt install ros-foxy-slam-toolbox        #For 2D slam
  

Downloading the simulator

In your home directory:

source /opt/ros/foxy/setup.bash
git clone --recursive [email protected]:sigridmellemseter/lonewolf.git
cd lonewolf/
colcon build

If you get any opencv2 errors, enter the following:

sudo cp -r /usr/include/opencv4/opencv2/ ../

Place the models in the folder gazebomodels inside the .gazebo/models/ folder:

cp -r ~/lonewolf/gazebomodels/. ~/.gazebo/models/

If you don't have a .gazebo/models/ folder, you can just make one with mkdir ~/.gazebo/models/

The first two times you build you may get some errors. Just build again and they should dissapear.

Launching the simulator

Open a terminal and write the following.

source /opt/ros/foxy/setup.bash
cd ~/lonewolf
source install/setup.bash
ros2 launch atvurdf <world_name>.launch.py

You should now see the ATV and the world in Gazebo.

Steering the ATV

You can steer the robot through keyboard commands using the ros2 application teleop_twist_keyboard.

source opt/ros/foxy/setup.bash
ros2 run teleop_twist_keyboard teleop_twist_keyboard

This application tells you how to steer the robot.

How view point cloud in rviz2:

While the simulator is running, open a new terminal and run rviz2

source /opt/ros/foxy/setup.bash
rviz2 

In rviz2 you have to define the correct frame. Write velodyne as the Fixed Frame under Global Options.

Now you have to add the PointCloud2 in rviz:

• Click Add in the bottom left corner
• Choose By Topic
• Choose PointCloud2 under /points_raw

Now you should see the point cloud in rviz. To get a better visual choose the Style "Flat Squares", Size (m) to 0.03 and set Color Transformer to "AxisColor"

Launching 3D SLAM (Generating maps)

While the simulator is running, open a new terminal and write the following.

source /opt/ros/foxy/setup.bash
cd ~/lonewolf
source install/setup.bash
ros2 launch scanmatcher lio_bigloop.launch.py

You can also view the maps in rviz2.

• Open rviz2 as described above
• Set the fixed frame to map
• Add the topics /map, /path, /modified_map and /modified_path.

To see the data more clearly, you can choose the following styles:

• /map : Flat Squares, Size (m) to 0.03, Color Transformer to AxisColor
• /path: Color white
• /modified_map: Flat Squares, Size (m) 0.03, Color Transformer to ?
• /modified_path: Color green

Launching SLAM (2D) (Navigation)

Slam Toolbox is a set of tools and capabilities for 2D SLAM built by Steve Macenski. You can read more about the functionalities of this toolbox here.

To start the 2D slam, open a new terminal and write the following.

source /opt/ros/foxy/setup.bash
ros2 launch slam_toolbox online_async_launch.py

Only online_async (asynchronous modes of mapping) is testet with this simulator.