ROS 2 Demo Catalog

Complete Run Your First ROS 2 Bridge before using these commands. All native examples assume that /opt/ros/jazzy/setup.bash has been sourced. Use gui:=true only with WSLg or another desktop display; otherwise use the headless form shown where available.

Demo	Purpose	Extra package	Launch command
TurtleBot3	First visible mobile-robot bridge demo	`ros-jazzy-turtlebot3-description`	`tb3_demo.launch.py`
Navigation	Mobile robot navigation topics and actions	None	`nav_demo.launch.py`
Panda arm	Seven-joint trajectory control	None	`arm_demo.launch.py`
Attach	Attach/detach a nearby box through ROS services	`ros-jazzy-turtlebot3-description`	`attach_demo.launch.py`
Fleet ROS GUI	100 robots with batched Fleet ROS commands	None	`fleet_demo.launch.py`
RMF office	RMF patrol and delivery integration	`ros-jazzy-pybullet-fleet-rmf` plus `rmf_demos` overlay	RMF quickstart

TurtleBot3

sudo apt install -y ros-jazzy-turtlebot3-description
ros2 launch pybullet_fleet_ros tb3_demo.launch.py \
  model:=burger gui:=true rviz:=false

For headless verification, replace the final arguments with gui:=false rviz:=false. The launch starts one robot named tb3_0.

Panda arm

Start the fixed-base Franka Panda demo. It exposes joint_states and the follow_joint_trajectory action under panda0:

ros2 launch pybullet_fleet_ros arm_demo.launch.py gui:=true rviz:=false

For RViz visualization, install ros-jazzy-rviz2 and omit rviz:=false.

From another terminal, send a seven-joint trajectory and keep the feedback on screen until the action completes:

ros2 action send_goal --feedback \
  /panda0/follow_joint_trajectory control_msgs/action/FollowJointTrajectory \
  "{trajectory: {joint_names: [panda_joint1, panda_joint2, panda_joint3, panda_joint4, panda_joint5, panda_joint6, panda_joint7], points: [{positions: [0.2, -0.5, 0.2, -1.8, 0.0, 1.3, 0.6], time_from_start: {sec: 3}}]}}"

Attach and detach

This demo starts tb3_0 beside three pickable boxes. Attach box_A, move with it attached, then detach it at the destination:

sudo apt install -y ros-jazzy-turtlebot3-description
ros2 launch pybullet_fleet_ros attach_demo.launch.py gui:=true

ros2 service call /tb3_0/toggle_attach std_srvs/srv/SetBool "{data: true}"

ros2 topic pub --once /tb3_0/goal_pose geometry_msgs/msg/PoseStamped \
  "{header: {frame_id: odom}, pose: {position: {x: 2.0, y: -1.0, z: 0.01}, orientation: {w: 1.0}}}"

After the robot reaches the goal, detach the box:

ros2 service call /tb3_0/toggle_attach std_srvs/srv/SetBool "{data: false}"

Fleet ROS GUI demo

This native demo creates a fleet scene with the Fleet ROS API enabled, then accepts one navigation request for every robot. It is the representative GUI demo for the fleet ROS path:

ros2 launch pybullet_fleet_ros fleet_demo.launch.py \
  robots:=10 robot_model:=simple_cube gui:=true target_rtf:=1.0

For a lower-overhead WSLg view, install ros-jazzy-rviz2 and launch with gui:=false rviz:=true. It renders lightweight cubes from the single /fleet/states stream rather than opening the PyBullet GUI.

From another terminal, inspect the batched state and move all ten robots while preserving their grid spacing:

ros2 topic echo /fleet/states --once
python3 -m pybullet_fleet_ros.fleet_nav_demo \
  --robots 10 --dx 0.5 --dy 0.5 --transport service

The bridge quickstart also includes a literal 10-robot ros2 service call suited to a live API demonstration. Set robot_model:=tb3_burger or tb3_waffle after installing ros-jazzy-turtlebot3-description. The same launch works in Docker; see the Docker bridge guide.

Docker-only model catalog

The model catalog launches every model available in the fully provisioned Docker image, including optional ROS descriptions. It is not an APT preview demo:

cd docker
docker compose run --rm bridge \
  ros2 launch pybullet_fleet_ros model_catalog_demo.launch.py gui:=true