Skip to content

Autonomy User Manual

Kacper Marchlewicz
16.10.2025
1.0.0

Accessing ROS

ROS is running within its own Docker container. To access it, use:

sudo docker exec -it rover-ros-1 bash

Launching the Visualization Web App

To launch the simple web app for data visualization and goal sending, run:

ros2 launch vizanti_server vizanti_rws.launch.py

You can then access it in your browser at:

http://<host_ip>:5000

Note: The <host_ip> depends on your connection method — it may be via VPN or local network.

Make sure to switch to the map frame. Topics will appear once nodes are launched.
If the connection becomes unstable, try restarting (either use “Force Reload” in the app or restart the vizanti_server node).

Autonomous Mode Setup

Before launching autonomous mode, a map of the environment must be prepared.
The robot’s odometry currently has limited accuracy, so mapping requires a static environment (avoid mapping when dynamic obstacles are present as they will be added to map as static ones).

Steps

  1. Launch sensors and odometry:

    ros2 launch firo_navigation sensors.launch.py

  2. Launch data fusion nodes:

    ros2 launch firo_navigation localization.launch.py

  3. Launch SLAM mode:

    ros2 launch firo_navigation slam.launch.py

After launching, add the map topic, robot frame, and 2D point cloud to display in the app.
Drive the robot around to build the map — it updates every ~0.45 meters or radians of change.

If the robot completes a loop (e.g., a circle or rectangle), loop closure might not occur immediately. Give it time, continue driving.

Avoid long, featureless areas (e.g., “infinite corridors”) — they reduce SLAM quality. Try approaching them from other directions.

Saving the Map

When the map quality is satisfactory, save it using:

ros2 run nav2_map_server map_saver_cli -f map_name

Then, close the data fusion and SLAM nodes (CTRL+C in their terminals), but keep sensors running.

Edit the navigation configuration file to set the map name:

nano install/firo_navigation/share/firo_navigation/config/navigation_params.yaml

Find and update the following section:

map_server:
  ros__parameters:
    yaml_filename: "test_map.yaml"

Replace "test_map..." with your own map name.
Save and exit with CTRL+X, then Y, then Enter.

Launching Navigation

With the sensors still running, start the navigation system:

ros2 launch firo_navigation navigation.launch.py

Launch may take some time. If it fails on the first attempt, try again.

In the web app, display:

  • Map
  • Global Costmap
  • Local Costmap

Before driving, set the initial pose (green arrow → click, hold, and rotate for orientation).
Move the robot slightly forward/backward to improve localization accuracy.
Once localized, you can send navigation goals (purple arrow → set pose and orientation).

Monitor the terminal running navigation — it will show status messages (e.g., receiving goal, recovery behavior).

Docking Station

Docking station coordinates are used by the docking server for automatic docking and undocking (using a PID controller).

Start detection when the robot is about 0.5 m from the docking station.
Initiate docking when about 1.5 m away and facing the station.

Steps

  1. Detect the docking station shape:

    ros2 launch firo_navigation dock_save_pose.launch.py

    You can visualize the detected candidate shape in the web app.

  2. Save the dock position:

    ros2 service call /save_dock_pose std_srvs/srv/Trigger "{}"

  3. Move the robot ~1 meter back and save the approach pose:

    ros2 service call /save_approach_dock_pose std_srvs/srv/Trigger "{}"

  4. Approach the docking station:

    ros2 service call /send_approach_dock_pose std_srvs/srv/Trigger "{}"

  5. Perform docking:

    ros2 launch firo_navigation dock_action_docking_launch.py

  6. Perform undocking:

    ros2 launch firo_navigation dock_action_udocking_launch.py