Advanced Features
Filtering log output
In config.sh you can set the level of logs you want outputted into the terminal. By default it is set to "info" to allow all logs. Possible values are: "debug", "info", "warn", and "error". Setting it to "warn" filters out all debug and info messages. Additionally, to filter out specific lines you can add the phrase you want filtered inside log_blacklist.txt and setting the log_level flag to "warn" or "error" will start filtering out all phrases found in the blacklist.
Aruco detection
If you want to add the TF links between the cameras and the ArUco markers without running the gpss command, you can run the following command. This is primarily useful for debugging, as gpss runs this as well.
./control.sh aruco_detection
Finally, to view the bird's-eye perspective from each camera, run the following command and open rviz. Then, navigate to the left panel and under "Camera" change the Topic /static_agents/camera_XXX/image_projected to visualize the corresponding camera feed:
Running ROS2 commands:
To avoid conflict between ros nodes in the same network, after each build a new random ROS2 domain is created. This means that you need to adjust this random domain before running any ros2 commands. For convenience you can use
./control.sh cmd YOUR_ROS2_COMMANDS
Here is an example of extracting tf2 hierarchy
./control.sh cmd ros2 run tf2_tools view_frames
These features and commands are under development and not fully supported yet and therefore are subject to change.
Cartographer: With both Gazebo and rviz running you can start creating a map of a robot's surroundings.
To start: ./control.sh cartographer
Teleop
If you want to control any robot (pallet truck, humanoid, etc.) manually you can run the following command. Remember to specify what robot you want to control by adding its namespace as argument, i.e. ./control.sh teleop pallet_truck_1
./control.sh teleop ${YOUR_ROBOT_NAMESPACE}
To move the humanoid around in the simulator
./control.sh teleop ${YOUR_HUMANOID_NAMESPACE}
Scenarios
In scenarios.sh you can run predefined scenarios of the project. At the bottom of the file, commands are shown how to run it; otherwise each scenario is referenced by a number. Currently there are 3 scenarios and to run them, run this command in the shell
Before running scenario 1 that uses GPSS cameras, make sure that CAMERA_ENABLED_IDS in config.sh has the list of GPSS cameras.
Note: A previous successful setup where all scenarios worked used these cameras: '160 161 162 163 164 165 166 167 168 169 170'. If robots are not moving in your simulation, it might be that there is no camera looking at the robot.
./control.sh build
./scenarios.sh 1 # navigation using GPSS cameras (real time factor need to be updated)
./scenarios.sh 2 # Humanoid and robotic arm
./scenarios.sh 3 # Humanoid navigation without GPSS cameras (navigate_w_replanning_and_recovery_robot_agent_X.xml need to be updated )
./scenarios.sh 4 # Record the video demo
If you want to see the planning route for all agents, load scenario_3_planning.rviz in Rviz.
Keep in mind to change real_time_factor in simulation/simlan_gazebo_environment/worlds/ign_simlan_factory.world.xacro to small values to slow down the simulator (e.g. 0.05-0.1) before building the project. ./control.sh build
Jackal
If you want to control the Jackal, you add the following lines into the control.sh:
elif [[ "$*" == *"jackal_teleop"* ]]
then
ros2 launch dyno_jackal_bringup keyboard_steering.launch.py
And then run:
./control.sh jackal_teleop
These ones did not work, so we put them here in issues
elif [[ "$*" == *"move_object"* ]]
then
ros2 run object_mover move_object
elif [[ "$*" == *"scenario"* ]]
then
ros2 launch scenario_execution_ros scenario_launch.py scenario:=simulation/scenario_manager/scenarios/test.osc
ISSUES
gazebo_ros2_control
[gzserver-1] [ERROR] [1727951819.671190014] [jackal.gazebo_ros2_control]: controller manager doesn't have an update_rate parameter
No solution.
OpenAL
[gzserver-1] [Err] [OpenAL.cc:84] Unable to open audio device[default]
Related to support for audio inside a Docker container. It will not be resolved.
Command failed: docker compose
If you have any issue with docker incompatibility (e.g. Error: Command failed: docker compose ...), make sure that docker compose or docker-compose is set correctly in the settings.
In docker-compose.yaml, uncomment the factory_simulation_nvidia section:
factory_simulation_nvidia:
<<: *research-base
container_name: factory_simulation_nvidia
runtime: nvidia
deploy:
resources:
reservations:
devices:
- driver: nvidia
count: "all"
capabilities: [compute,utility,graphics,display]
and update .devcontainer/devcontainer.json:
{
"name": "ROS2 RESEARCH CONTAINER",
"dockerComposeFile": "../docker-compose.yaml",
"service": "factory_simulation_nvidia
...
Missing nvidia docker runtime
Solution: https://stackoverflow.com/questions/59008295/add-nvidia-runtime-to-docker-runtimes
Docker nvidia-container-cli: requirement error
If you get this error when building the docker container:
nvidia-container-cli: requirement error: unsatisfied condition: cuda>=12.6, please update your driver to a newer version, or use an earlier cuda container: unknown
A solution to try is first: re-install nvidia-container-toolkit. Or, if that does not work, update your nvidia-drivers.
GLIBC_2 issue
Sometimes the wrong nvidia-container-toolkit results in this issue below:
$/usr/lib/x86_64-linux-gnu/libc.so.6: version `GLIBC_2.38' not found (required by /usr/lib/x86_64-linux-gnu/libGLdispatch.so.0)
$/usr/lib/x86_64-linux-gnu/libc.so.6: version `GLIBC_2.38' not found (required by /usr/lib/x86_64-linux-gnu/libGLX.so.0)
If so, try downgrading nvidia-container- toolkit. You can use these commands:
$sudo apt-get remove --purge nvidia-container-toolkit nvidia-container-toolkit-base libnvidia-container*
$sudo apt-get update
$sudo apt-get install nvidia-container-toolkit=1.17.4-1 nvidia-container-toolkit-base=1.17.4-1 libnvidia-container1=1.17.4-1 libnvidia-container-tools=1.17.4-1
docker issues
Make sure docker is installed correctly by following these two instructions: