Repository for package containing OfficeMate robot description and launch files
- Install ROS2 Humble following steps in URL: https://docs.ros.org/en/humble/Installation/Ubuntu-Install-Debians.html
Small tip: execute line to automaticaly source ros when terminal is being opened:
echo "source /opt/ros/humble/setup.bash" >> ~/.bashrc - Install dependencies:
sudo apt install ros-humble-gazebo-ros-pkgs
sudo apt install ros-humble-xacro ros-humble-joint-state-publisher-gui
sudo apt install ros-humble-ament-lint-auto
sudo apt install ros-humble-ros2-control ros-humble-ros2-controllers ros-humble-gazebo-ros2-control
sudo apt install ros-humble-foxglove-bridge
sudo apt install ros-humble-slam-toolbox
sudo apt install ros-humble-navigation2 ros-humble-nav2-bringup ros-humble-turtlebot3
sudo apt install ros-humble-twist-mux
- Prepare workspace
mkdir -p dev_ws_office_mate/src
cd dev_ws_office_mate/src
- Clone repositories
git clone <link_to_repo>
git clone https://github.com/lukbieli/teleop_tools.git
- Rename folder to office_mate (Important!)
mv OfficeMate-RosMateBot office_mate
cd ..
- If not already installed install colcon
sudo apt install python3-colcon-common-extensions
- Build package with colcon
colcon build --symlink-install
To launch high detailed model exported from onshape: Warning: It can cause errors during launch or lag in simulation. Especially if used world is also large
source install/setup.bash
ros2 launch office_mate sim.launch.py
To use prepared world use argument modifier
source install/setup.bash
ros2 launch office_mate sim.launch.py world:=<path_to_world_file>
To use LTTS kitchen 3D scan as a model, before calling a launch command change directory to officemate/gazebo where the jpg with textures is located. (That the only way (for now) to make textures visible in gazebo.
source install/setup.bash
cd src/office_mate/gazebo
ros2 launch office_mate sim.launch.py world:=<path_to_world_file>
To workaround issues with lagging simulation or errors simplified model is available:
source install/setup.bash
ros2 launch office_mate sim_simple.launch.py
Launch with model of a kitchen
source install/setup.bash
cd src/office_mate/gazebo
ros2 launch office_mate sim_simple.launch.py world:=<path_to_world_file>
Launch with joystick steering and model of a kitchen
source install/setup.bash
cd src/office_mate/gazebo
ros2 launch office_mate sim_simple_full.launch.py world:=<path_to_world_file>
That's option enabled by default in sim_simple.launch.py file by calling joystick launch. To disable it comment out joystick in last return.
First possibility to steer the model manually is via teleopt_twist_keyboard package
ros2 run teleop_twist_keyboard teleop_twist_keyboard --ros-args -r /cmd_vel:=/diff_cont/cmd_vel_unstamped
Second one possibility to steer the model manually is via joistick from foxglove.
First you need a foxglove studio app. The debian package is avalaible in website:
https://foxglove.dev/download
Download it and follow the instrucions there or here. Do in location where you downlad it:
sudo apt install ./foxglove-studio-*.deb
sudo apt update && sudo apt install foxglove-studio
Now you need a websocket to connect joystick with ros2. Download it:
sudo apt install ros-humble-foxglove-bridge
Anytime you will run foxglove you will need a connection bridge to communicate it with ROS. Do:
ros2 launch foxglove_bridge foxglove_bridge_launch.xml
Open foglove studio app. On the left side are Open Data Sources. Click on Open conection.. Then chose Foxglove WebSocket, WebSocket URL left with default settings and click open. Now click on the user in top right corner and chose "Extensions". Find the Joystick extension and install it. In the Joystick panel change Data source from "Subscribed Joy Topic" to "Interactive". Change Publish to on and In display section - display value to Custom Display. Now you need a toystic node and translation node to x, z axes. Run:
ros2 launch office_mate joystick.launch.py
Now you can steer a robot with joystick.
There is a lunch file for starting rviz
source install/setup.bash
ros2 launch office_mate start_rviz.launch.py
Now it's time to launch our slam_toolbox and see how mapping works:
ros2 launch slam_toolbox online_async_launch.py slam_params_file:=src/office_mate/config/mapper_params_online_async.yaml use_sim_time:=true
or call a launch file:
ros2 launch office_mate online_async_launch.py
To save scanned map and reuse it in the future, on the top side of the Rviz window click Panels > Add New Panel and select SlamToolBoxPlugin. Next to Save Map and Serialize Map give a name for your saved files and click on the save buttons. (Note: give a different name to these files. Example: name_save, name_serialize) We only care about Serialize Map, but it won't hurt to save both.
*Note: you can find premade map of simulated 3D enviroment in folder "Premade Map"
In /config/ folder open mapper_params_online_async.yaml file and search for "mode" which should be located at line 18 and change parameter from "mapping" to "localization". And just a few lines down, uncomment line 23: map_file_name and provide a path to your saved map which has extension ".data", in my example it would be:
First lets run nav2_bringup:
ros2 launch nav2_bringup navigation_launch.py use_sim_time:=true
or use custom launch file:
ros2 launch office_mate nav2_navigation_launch.py
In the top section of Rviz you will find a button called "2D Goal Pose", click on it and choose a destination on the map.
*Note: don't forget to change Fixed Frame to "map", otherwise robot might get stuck and abort his given commands!
If you need more description with video representation of all above steps, please follow these links: SLAM_TOOLBOX: https://www.youtube.com/watch?v=ZaiA3hWaRzE NAV2: https://www.youtube.com/watch?v=jkoGkAd0GYk