UP | HOME

Homework 4

Logistics

  1. Create a git repository using the link provided in Canvas.
  2. Your homework will be submitted via this git repository.
  3. I will grade whatever is on the main branch when I clone the repository.
  4. Guidelines provides important information about what you should turn-in, including information about your README.md file.
  • The goal is to use the simulated robot to map an environment and then navigate within the map using the slam_toolbox

The Environment

  • Clone the package nubot, which is a package that simulates a mobile robot with a lidar in Gazebo.
  • Your repository will be next to nubot in the source space: e.g.(ws/src/nubot and ws/src/homework4).
  • I purposely did not put a README.md in nubot, so you will need to explore it to find out how it works.
  • HINT: There is one launchfile that runs the simulation and visualizes it in rviz.
  • HINT: There are two worlds to use, either will be accepted for the assignment but one is significantly smaller (and hence easier to explore).
  • Going through the First Time Setup Guide (through setting up the robot's footprint) will help you understand how the Nav2 stack works, and may be a helpful reference while completing this assignment.
  • Much of the setup, however, is already done for you in nubot.
  • By running the launch file in the nubot package you will be able to see what features required by the nav_stack are already implemented.

A Launch File

  • Your package will be called nubot_nav
  • In nubot_nav, write a launch file called manual_explore.launch.xml that:
    • Starts the nubot simulation.
    • Starts slam_toolbox in async mode.
    • Starts the nav2 stack (to create cost-maps and also enable navigation).
      • Make sure to set a parameter to increase the footprint of the robot so that it encompasses the whole robot
      • You may use a circular approximation of the robot
    • Starts rviz and displays:
      • The robot relative to the /map frame
      • The global costmap
      • The local costmap (with the costmap color scheme)
      • The planned path
  • Record a video of the robot mapping the environment until at least 1 corridor is mapped.
    • You can manually set waypoints using rviz2
  • HINT: You must enable simulation time when using the launchfiles to start navigation and SLAM

Important

  • In ROS 2 kilted as of v1.4.2, the nav2_bringup package's default nav2_params.yaml file has invalid parameters.
  • Use ros2 launch nav2_bringup navigation_launch.py --show-args to find the location of this launchfile, copy it into your own package, and modify all values that end in _ZONE_ENABLED to be false.
  • When using nav2_bringup navigation_launch.py in your own launchfiles, make sure to set the argument so it uses your custom configuration.

Autonomous Exploration

  • Write a node called explore that causes the robot to explore the environment autonomously while mapping the world.
  • The launchfile called explore.launch.xml should start all the nodes necessary to allow the robot to explore and map its environment
    • This launchfile should include manual_explore.launch.xml and then start the explore node.
  • Use the map_saver_cli in the nav2_map_server package to save the resulting map and place it within the maps directory of the git repository
  • Record a video of the robot exploring the environment in rviz (you need not capture what the robot is doing in gazebo)

Exploration Algorithm

There are two choices for exploration algorithm:

  1. Frontier Exploration.
    • See Robot Frontier for a friendly description, in particular Detecting Frontiers
    • This is the most established method and most difficult to implement.
    • Students who successfully complete this method will receive extra credit
  2. Random exploration: some variation of the following algorithm:
    • Choose a random waypoint
    • Travel to the waypoint
    • Repeat
  3. An algorithm of your choice:
    • The exploration does not need to be complete, perfect, or optimal, but it should be generic (so no hardcoding locations that work for this particular room).
    • The algorithm should be described or cited in the README.md
    • Extra credit may be received (at my sole discretion) based on whether it is qualitatively closer (in terms of effectiveness) to Frontier Exploration or Random Exploration.

Hints

  1. Subscribe to the /map topic to get information about the map

README

  • The README.md should contain your name, and a brief project description.
  • It should also describe the ros2 launch commands needed to run the manual and autonomous exploration
  • It should embed both of the required videos directly in the README so that they can viewed on github
  • It should clearly indicate which exploration algorithm you chose to implement and (if applicable) a high-level description of the algorithm.

Author: Matthew Elwin.