ros2-course

DevContainer setup for ROS2 course

Prerequisites

Linux

Install the following:

• Docker Engine
• VS Code
• VS Code extension: Dev Containers

Follow the listed steps to finish setup:

1. Create a json file in the .devcontainer dir named devcontainer.json and copy the content of devcontainer - linux.json into it.

Windows

Install the following:

• Docker Engine
• VS Code
• VS Code extension: Dev Containers
• XLaunch or other Xserver

Follow the listed steps to finish setup:

1. Create a json file in the .devcontainer dir named devcontainer.json and copy the content of devcontainer - windows.json into it.
2. Start a Xserver with 0.0 as DISPLAY port

Start DevContainer

1. Open this repository in VS Code.
2. crtl + shift + P to open command prompt
3. Enter Dev Containers: Rebuild and Reopen in Container

VS Code will start the Dev Container building the Dockerfile found in the /docker dir. This is a basic ROS2 Jazzy Container with the following ROS related binaries installed:

• ros-jazzy-plotjuggler-ros, a topic plotter package.
• ros-jazzy-moveit, binary install of the base MoveIt2 packages
• ros-jazzy-ros-gz, binary install of gz-sim11 Gazebo Harmonic ROS2
• '~nros-jazzy-rqt*', RQT for ROS2 Jazzy

Installing in Container

The ~/ros2_ws/src directory in the container is bound to the src directory in this repository. Which means source installs are persistent and saved when stopping and/or rebuilding the DevContainer. The opposite is true for binary installs, these are not persistent. Unless added to the Dockerfile.

ROS2 packages should be installed/cloned into the ~/ros2_ws/src folder. To install the dependencies for these packages use:

cd ~/ros2_ws
sudo apt update
rosdep update
rosdep install --from-paths src --ignore-src -y -r

This should install all package dependencies referenced in the package.xml files. To build the workspace use:

cd ~/ros2_ws
colcon build --symlink-install
source install/setup.bash

Start BFFP Demo Factory

After installing the rosdeps and building and sourcing the workspace (see Installing in Container), use: ros2 launch bffp_gazebo factory_moveit_full.launch.py to launch the demo. This launches a Gazebo simulation and a HC10DTP with full MoveIt2 support. Rviz2 is also launched to visualize ROS2 topics.

Start Motoman demo

TODO

Start Tiago simulation

Tiago 2D navigation using Nav2

Documentation : https://docs.pal-robotics.com/sdk-dev/navigation

1️⃣ Start Gazebo simulation of Tiago robot:

ros2 launch tiago_gazebo tiago_gazebo.launch.py is_public_sim:=True world_name:=house_pick_and_place

2️⃣ In a new terminal, start cartographer to run SLAM:

ros2 launch tiago_2dnav tiago_nav_bringup.launch.py slam:=True is_public_sim:=True

3️⃣ In a new terminal, start teleoperation from the keyboard:

ros2 run teleop_twist_keyboard teleop_twist_keyboard --ros-args -r /cmd_vel:=/key_vel

Then move the robot to accumulate map data. Close teleoperation with Ctrl+C and save map to file our_map using:

ros2 run nav2_map_server map_saver_cli -f ~/ros2_ws/src/pal_maps/maps/our_map/map

Close the SLAM terminal with Ctrl+C. If you have not used --symlink-install, workspace building and sourcing are needed so that the new map is installed. 4️⃣ Start 2D navigation by loading your our_map map, using:

ros2 launch tiago_2dnav tiago_nav_bringup.launch.py is_public_sim:=True world_name:=our_map

⚠️ Teleoperation and navigation both publish Twist robot commands to /cmd_vel: make sure they don't publish at the same time, your robot would receive contradictory commands. 5️⃣ In RViz use the arrows to control navigation:

1. Estimate 2D pose to initialize random particles to an approximate position around the actual robot position
2. Send a 2D nav goal to any point of the map: your robot must navigate there 6️⃣ Command navigation from Python code using the navigate_to_pose action service (see workshops instructions) 7️⃣ 4-waypoints robot patrol: Inspire from the goto code to add a new patrol node, endlessly patroling between 4 map poses.

Tiago arms and gripper manipulation using MoveIt2

Documentation: https://docs.pal-robotics.com/sdk-dev/manipulation 8️⃣ Command MoveIt using color handles from the GUI in RViz:

ros2 launch tiago_moveit_config moveit_rviz.launch.py

9️⃣ Command MoveIt from Python file ros2_ws/src/tiago_pick_and_place/tiago_pick_and_place/pick.py:

# Warning: before executing, first edit the pick.py file to uncomment the code that you want
ros2 launch tiago_pick_and_place plan.launch.py use_sim_time:=True