Skip to content

README_f1tenth_slam.md

Latest commit

 

History

History
31 lines (24 loc) · 1.74 KB

README_f1tenth_slam.md

File metadata and controls

31 lines (24 loc) · 1.74 KB

F1Tenth Mapping Algorithm: SLAM

This exercise is based on CMU 16663 - F1Tenth Course :: Lab 5.

I. Learning Goals

Here, we introduce SLAM, Simultaneous Localization and Mapping.

We are not going to implement SLAM ourselves, but instead use slam_toolbox to make a map of your surroundings. Please watch UPenn's F1Tenth lecture on Simultaneous Localization and Mapping (SLAM) and review CMU's F1Tenth lecture slide on Hector SLAM (slides are highly recommended).

II. SLAM

We won't have you implement SLAM yourselves in this tutorial (that itself can be a whole project). Instead, please use slam_toolbox to make a map of the Levine Hall. Then, save the map as levine_hall.pgm and levine_hall.yaml.

  1. Installing slam_toolbox 
    sudo apt install ros-foxy-slam-toolbox
  2. Before running slam_toolbox, make sure the odometry on your vehicle/simulator is tuned!
  3. Launcing slam_toolbox
    • Launch teleop in one window
    • Launch slam_toolbox in another window 
      ros2 launch slam_toolbox online_async_launch.py params_file:=/home/nvidia/f1tenth_ws/src/f1tenth_system/f1tenth_stack/config/f1tenth_online_async.yaml
  4. Visualization
    • Launch rviz2
    • Add /map by topic
    • Add /graph_visualization by topic
    • On top left corner of rviz, panels > add new panel > add SlamToolBoxPlugin panel
    • Once you're done mapping, save the map using the plugin. You can give it a name in the text box next to "Save Map". Map will be saved in whichever directory you ran slam_toolbox.