This guide provides detailed operational instructions for running and managing the Autonomous Underwater Vehicle (AUV) system using the top side laptop. In the current configuration, all nodes except inference nodes are deployed on the AUV.

Ensure the following prerequisites are completed on the top side laptop:

• Static IP address set to 192.168.2.1/24

• IP routing to Jetson Nano via Raspberry Pi configured

• FastDDS configuration completed

• auv_ws workspace built with auv_interfaces sourced

Run the following command on the top side laptop:

sudo ip route add 192.168.2.4 via 192.168.2.2

On the top side laptop, add the following to each terminal session:

export ROS_DISCOVERY_SERVER=192.168.2.2:11811
export FASTRTPS_DEFAULT_PROFILES_FILE=~/auv_ws/src/auv_ros2/super_client_configuration_file.xml
ros2 daemon stop && ros2 daemon start

ssh [email protected]

Default password: raspberry

Start the container:

docker run --rm --network host --name pi-ros-main -it \
  --mount type=bind,src=/home/pi/trees,dst=/trees \
  ghcr.io/offset-official/pi-ros-full

If a container is already running:

docker container rm --force pi-ros-main

• The container starts with a tmux session and automatically runs pi.launch.py in the first window.

• Access the first window using Ctrl + b, then press 0.

Refer to docker/README.md for more information.

All code resides in the /auv_ws directory.

To rebuild packages:

cd /auv_ws
colcon build

To source the workspace:

source /auv_ws/install/setup.bash

ssh [email protected]

Default password: nano

Start the container:

sudo docker run --rm --network host --name nano-ros-main -it \
  --device=/dev/ttyACM0 \
  --device=/dev/video0 \
  --device=/dev/video2 \
  ghcr.io/offset-official/nano-ros-full

If a container is already running:

sudo docker container rm --force nano-ros-main

• A tmux session will launch with nano.launch.py running in the first window.

• Access with Ctrl + b, then press 0.

Refer to docker/README.md for details.

Code is located in /auv_ws.

To rebuild:

cd /auv_ws
colcon build

To source:

source /auv_ws/install/setup.bash

Run the inference container:

docker run --rm -it --network host --runtime nvidia \
  -v /home/nano/models:/models \
  --name jetson-ml-final \
  jetson-ml-final:latest

In a different tmux window inside the ROS container:

ros2 launch auv_bringup qualify_inference_launch.py

The top side laptop is responsible for sending commands and monitoring the AUV.

To issue movement commands:

ros2 run auv_controller control_cmd_cli

To check pressure sensor output:

ros2 topic echo /current_depth

To send a depth action (example: descend to -0.3 meters):

ros2 action send_goal /depth_descent auv_interfaces/action/DepthDescent \
  '{target_depth: -0.3}' --feedback

To manually arm thrusters:

ros2 service call /mavros/cmd/arming mavros_msgs/srv/CommandBool "{value: true}"

To run heading test with custom parameters:

ros2 run auv_mav_utils heading_test --ros-args \
  -p target_depth:=-0.8 \
  -p linear_speed:=1.5 \
  -p enable_angle_correction:=false \
  -p movement_duration:=15.0

To calibrate the depth sensor:

ros2 service call /calibrate_depth_sensor std_srvs/srv/Trigger

To actuate the dropper:

ros2 service call /dropper_trigger auv_interfaces/srv/DropperTrigger "{enable: true}"

Use rqt_image_view:

ros2 run rqt_image_view rqt_image_view

Select the desired camera topic from the GUI.

To change LED color (example: white):

ros2 service call /set_color auv_interfaces/srv/SetColor "{color: '#ffffff'}"

To turn off LEDs:

ros2 service call /set_color auv_interfaces/srv/SetColor "{color: 'off'}"

To run the full communication sequence:

ros2 action send_goal /read_comm_sequence auv_interfaces/action/ReadCommSequence {} --feedback

To run experiment nodes, for example the qualify experiment:

ros2 run auv_experiments qualify