To complete this workshop please arrive with:

• A basic understanding of Ubuntu Linux
• A basic understanding of Python and C++
• A laptop running either:
  • Ubuntu 18.04 and ROS Melodic or
  • Ubuntu 20.04 and ROS Noetic
• Please follow the installation instructions here
  • Ask a mentor if you get stuck, and we'll be happy to help
• Access to the Internet (you'll need to download 1-2 GB)
  • Ask a mentor for Wi-Fi access details
• Access to this GitHub repository (consider adding a SSH key)

Note: While any development environment can be used for this course, for beginners we recommend the free VSCode source code editor.

This course material has been developed in our spare time. As you work through the workshop, please consider contributing any suggestions or edits back -- it's easy! Please fork the repository you want to update and do a “pull request“ (PR) when you're ready for us to review. For the full process, read up on “Git Flow” here.

Topics:

• Filesystem Architecture
• Computational Graph
• Community
• References

Topics:

• Directory structures for Catkin workspaces
• Creating packages
• Nodes and topics
• Bag files
• Viewing rqt_graph
• About tf and tf trees
• Parameters and services
• Writing a publisher and subscriber nodes
• Publishing and subscribing in the same node
• Using rviz
• Using roslaunch

Topics:

• Creating a new package by hand
• Managing dependencies
• Running ROS across multiple machines
• Writing launch files for large projects
• Creating a URDF
• Visualising a robot in rviz
• Visualising a robot in Gazebo

Make sure you can answer these questions before proceeding:

1. What's a node and a topic?
2. Have you recorded a bag file, and what does the rosbag play --clock parameter do?
3. Have you used rqt_graph to check what topics a node is subscribed to?
4. Do you know what a URDF is?
5. What's the "tf tree" and have you used rqt_tf_tree for introspection?
6. How would you check how many messages a node is sending per second?
7. How to create custom messages using ROS message description language?
8. How to add external packages' nodes in a launch file?
9. How to edit your package's CMakeLists.txt file to add the source of ROS nodes, messages, services, etc. for compilation? 
10. How do you remap a topic name when starting a node in a lanch file?
11. What is an workspace "overlay" in Catkin?
12. What is the difference between a ROS package and a Debian (Ubuntu) package?

• Whenever you open a new terminal window remember to run source devel/setup.bash from the workspace/src directory
• After you have changed a file within a package you need to run catkin build from the workspace/src directory
• Often packages cannot be installed with sudo apt-get install as they might not have a rosdep key! If this is the case, clone the package manually and build within your workspace
• If rosrun does not find your package, use --force-discover

The second part of this workshop consists of four application areas that are worked on over four days. The workshop material is cloned from these repositories:

• Monday: Sensor Integration and URDFs
• Tuesday: SLAM & Navigation
• Wednesday: Perception
• Thursday: Manipulation

The fifth day (Friday) integrates these application areas to build a simulated robot that can navigate its environment, finding and picking up cubes:

• Friday: Mobility Plus Manipulation

Across these five days, you will incrementally build a catkin workspace called workshop_ws. Each day will build on the previous, so make sure you end each day with a working solution!

mkdir -p ~/workshop_ws/src  # Creates a workspace directory named workshop_ws.
cd ~/workshop_ws/src

For each day, clone the repository linked below into the src/ directory of the workspace workshop_ws.

cd ~/workshop_ws/src
git clone https://github.com/ros-workshop/sensor-integration.git

cd ~/workshop_ws/src
git clone [email protected]:ros-workshop/sensor-integration.git

Build and then source the workspace.

cd ~/workshop_ws
catkin build
source devel/setup.bash

Topics:

• Creating a URDF
• Visualising your robot
• Fixing a broken URDF
• Adding a sensor to a robot
• Controlling a simulated robot
• Detecting an obstacle and stopping the robot

Topics:

• Creating a map using a lidar
• Simultaneous Localisation and Mapping (SLAM)
• Using move_base for navigation
• Finding an object by navigating around a map

Topics:

• Using a camera to detect Apriltags
• Using a real camera with ROS
• Camera calibration
• Fusing lidar and camera/DNN data for person detection and localisation

Topics:

• Creating a Moveit configuration package
• Moving Your robot in rviz
• Using the Moveit class in a node
• Creating a OctoMap using a depth camera

Topics:

• How to integrate multiple ROS nodes together
• How to create a robot in a Gazebo world that finds and picks up as many cubes as it can