Courses / programming

ROS 2 Navigation Advanced

advanced~15 hours12 lessonsEnvironment: docker

Prerequisites: ros2-navigation, kalman-filters

Start: AMCL: particle filters on a mapPractice questions

Modules

Module 1 — Localization
  1. 1AMCL: particle filters on a map
  2. 2Tuning AMCL
  3. 3Global vs local localization failures
Module 2 — Sensor fusion
  1. 1robot_localization and the EKF node
  2. 2Fusing odometry and IMU
  3. 3Frames done right: map, odom, base_link
Module 3 — Behavior trees
  1. 1Why Nav2 uses behavior trees
  2. 2Reading the default BT
  3. 3Customizing navigation behavior
Module 4 — Robust autonomy
  1. 1Recovery behaviors
  2. 2Dynamic obstacle handling
  3. 3Project: robot that survives a hostile warehouse

Textbooks

Video resources

  • Nav2 deep-dive talks — ROSCon talks