Lab 2: Coordinate Frames and Simulation

By the end of this lab, you will learn how to run a full 3D simulation on your own computer and how to interact with the ROS tf2 library using CLI tools and the Python API.

Learning Objectives:

  • Complete the official ROS 2 Jazzy TF2 Python tutorials.
  • Run the TurtleBot 4 Gazebo simulation.
  • Inspect the TurtleBot 4 tf tree using CLI tools.
Note

In this lab, you will be testing the VM’s limits on your computer hardware. A lot of your effort might be focussed on the tuning and debugging involved in setting up your environment to run Task 3 (Gazebo simulator). As every student’s computer has a different configuration, please use online resources, LLMs, class discussions, and office hours to help get the Gazebo simulator up and running.

Task 1: Complete the ROS TF2 Tutorial

Before starting the (simulated) robot-specific task, you must complete the following official tutorials for ROS 2 Jazzy. These will help you understand how to use ROS tf2 and provide the boilerplate code you need for Part C.

  1. Learning about TF2: Overview of view_frames and tf2_echo.
  2. Writing a Static Broadcaster (Python): How to define a frame that doesn't move.
  3. Writing a Listener (Python): Learn the Buffer and TransformListener pattern.

Task 2: Run the Gazebo Simulation for Turtlebot4

In your ROS workspace, clone the simulation git repo:

git clone https://github.com/ras-mobile-robotics/ras598_sim.git

Build the ROS workspace and launch the TurtleBot 4 simulation:

ros2 launch ras598_sim turtlebot4_gz.launch.py

Wait 1-3 minutes until the log output in the terminal stabilize and you see the message below:

============================================================
       SIMULATION READY FOR STUDENTS
       The robot is undocked and ready for commands.
============================================================

Note: Due to race conditions or specific hardware limitations, the robot might not automatically undock. If after 1-3 minutes the terminal output is stable and not changing, run the command below to force an undock:

ros2 action send_goal /undock irobot_create_msgs/action/Undock "{}"

It should reply with the message "Goal finished with status: SUCCEEDED" and the simulated robot should be undocked.

Note

If you had issues with runnning the simulation, the simulation being slow, or seeing a low of warning messages, checkout the VM guide for Enable 3D Hardware Acceleration.

Task 3: Robot Teleoperation

Run the teleoperation node:

ros2 run teleop_twist_keyboard teleop_twist_keyboard --ros-args -p stamped:=true

Move the robot around to verify it is working.

Task 4: The Visual TF Tree

Run the following command to generate a PDF of the robot's hierarchy:

ros2 run tf2_tools view_frames

This generates a PDF of the entire tf tree. Take a look at it! If you are not sure how to open the file, this is a good opportunity to explore your VM interaction skills.

Task 5: Measuring the Offset

The camera is located away from the Center of Mass (CoM) of the robot. Your goal is to find this offset.

In the frames list from Task 2, identify the frame name for the OAK-D Lite camera. It will likely contain the keywords oakd and link (ROS terminology usually combines these terms). Pick one of these frames for the command below.

Use the CLI to find exactly what that offset is:

ros2 run tf2_ros tf2_echo base_link <YOUR_CAMERA_FRAME_NAME>

Deliverables

In Canvas, submit a short screen recording (~10 secs) for each of the following:

  1. The turtles moving from Task 1.
  2. Turtlebot moving around in Gazebo from Task 3.
  3. tf tree and the terminal window printing the offest from Task 5.