MATE 2026 ROV

UConn Underwater Robotics Club

Architecture Overview

┌─────────────────────────────────┐          ┌──────────────────────────────────┐
│     LattePanda (Onboard)        │          │     Laptop (Control Station)     │
│     Ubuntu 22.04 + ROS 2        │          │      Ubuntu 22.04 + ROS 2        │
│                                 │   LAN    │                                  │
│  rov_onboard package:           │◄────────►│  rov_control package:            │
│    • camera_node      ───────────publish──►│    • camera_viewer_node          │
│    • sensor_node      ───────────publish──►│    • dashboard_node              │
│    • status_node      ───────────publish──►│                                  │
│    • thruster_node    ◄──────────subscribe─│    • gamepad_node                │
└─────────────────────────────────┘          └──────────────────────────────────┘

ROS 2 Topics

Topic	Type	Direction
/rov/camera/image_raw	sensor_msgs/Image	LattePanda → Laptop
/rov/camera/image_compressed	sensor_msgs/CompressedImage	LattePanda → Laptop
/rov/sensor_data	rov_msgs/SensorData	LattePanda → Laptop
/rov/status	rov_msgs/RovStatus	LattePanda → Laptop
/rov/thruster_command	rov_msgs/ThrusterCommand	Laptop → LattePanda

---

File Structure

mate_rov_ws/
├── README.md
├── src/
│   ├── rov_msgs/                          # Shared custom messages
│   │   ├── CMakeLists.txt
│   │   ├── package.xml
│   │   └── msg/
│   │       ├── ThrusterCommand.msg
│   │       ├── SensorData.msg
│   │       └── RovStatus.msg
│   │
│   ├── rov_onboard/                       # Runs on LattePanda
│   │   ├── package.xml
│   │   ├── setup.py
│   │   ├── setup.cfg
│   │   ├── resource/rov_onboard
│   │   ├── config/
│   │   │   └── onboard_params.yaml
│   │   ├── launch/
│   │   │   ├── onboard_launch.py          # Full system launch
│   │   │   └── camera_only_launch.py      # Camera-only for testing
│   │   └── rov_onboard/
│   │       ├── __init__.py
│   │       ├── camera_node.py
│   │       ├── thruster_node.py
│   │       ├── sensor_node.py
│   │       └── status_node.py
│   │
│   └── rov_control/                       # Runs on Laptop
│       ├── package.xml
│       ├── setup.py
│       ├── setup.cfg
│       ├── resource/rov_control
│       ├── config/
│       │   └── control_params.yaml
│       ├── launch/
│       │   ├── control_launch.py          # Full control launch
│       │   └── camera_viewer_only_launch.py
│       └── rov_control/
│           ├── __init__.py
│           ├── camera_viewer_node.py
│           ├── gamepad_node.py
│           └── dashboard_node.py

---

STEP-BY-STEP SETUP INSTRUCTIONS

STEP 1: Network Configuration (BOTH machines)

Both machines must be on the same local network (e.g. connected to the same router/switch, or a direct Ethernet cable).

On the LattePanda, find its IP:

ip addr show
# Look for your ethernet/wifi adapter, note the IP (e.g. 192.168.1.100)

On the Laptop, find its IP:

ip addr show
# e.g. 192.168.1.101

Verify they can reach each other:

# From laptop:
ping 192.168.1.100

# From LattePanda:
ping 192.168.1.101

STEP 2: ROS 2 DDS Network Setup (BOTH machines)

ROS 2 uses DDS for communication. By default, nodes on the same LAN with the same ROS_DOMAIN_ID will auto-discover each other.

Run this on BOTH machines (add to ~/.bashrc for persistence):

echo 'export ROS_DOMAIN_ID=42' >> ~/.bashrc
source ~/.bashrc

Use any number 0-232. Just make sure both machines use the same number. Using a non-zero ID avoids conflicts with other ROS 2 users on the network.

To verify DDS works across machines, on one machine run:

ros2 topic pub /test std_msgs/String "data: hello"

On the other machine:

ros2 topic echo /test

You should see hello appear. Press Ctrl+C to stop both.

STEP 3: Install Dependencies (BOTH machines)

Both machines need the same workspace (so custom messages are available). Copy this entire mate_rov_ws folder to both machines at ~/mate_rov_ws.

You can use scp to copy from one machine to another:

# From the machine that has the files, copy to the other:
scp -r ~/mate_rov_ws user@192.168.1.XXX:~/mate_rov_ws

Or use a USB drive, git repo, etc.

On BOTH machines, install required system packages:

# Source ROS 2 (adjust if you use a different distro than Humble)
source /opt/ros/humble/setup.bash

# Install dependencies
sudo apt update
sudo apt install -y \
    python3-opencv \
    ros-humble-cv-bridge \
    ros-humble-image-transport \
    python3-pip

# For the gamepad node (control laptop only):
pip3 install pygame

STEP 4: Build the Workspace (BOTH machines)

cd ~/mate_rov_ws

# Source ROS 2
source /opt/ros/humble/setup.bash

# Build all packages
colcon build --symlink-install

# Source the workspace
source install/setup.bash

Tip: Add these lines to ~/.bashrc on both machines so you don't have to type them every time:

echo 'source /opt/ros/humble/setup.bash' >> ~/.bashrc
echo 'source ~/mate_rov_ws/install/setup.bash' >> ~/.bashrc
echo 'export ROS_DOMAIN_ID=42' >> ~/.bashrc

STEP 5: Test Camera Streaming (Quick Test)

On the LattePanda:

cd ~/mate_rov_ws
source install/setup.bash
ros2 launch rov_onboard camera_only_launch.py

You should see output like:

[camera_node] Camera opened: 640x480 @ 30fps (index 0)
[camera_node] Published 150 frames

If the camera doesn't open, check:

# List available cameras:
ls /dev/video*

# If your camera is at a different index, change it:
# Edit src/rov_onboard/config/onboard_params.yaml
# Set camera_index to 2, 4, etc. then rebuild: colcon build

On the Laptop:

cd ~/mate_rov_ws
source install/setup.bash
ros2 launch rov_control camera_viewer_only_launch.py

An OpenCV window titled "ROV Camera Feed" should appear showing the live camera feed from the LattePanda!

STEP 6: Launch Full System

On the LattePanda (all onboard nodes):

cd ~/mate_rov_ws
source install/setup.bash
ros2 launch rov_onboard onboard_launch.py

On the Laptop (all control nodes):

cd ~/mate_rov_ws
source install/setup.bash
ros2 launch rov_control control_launch.py

STEP 7: Verify Everything is Running

On either machine:

# List all active nodes:
ros2 node list

# List all active topics:
ros2 topic list

# Check camera topic publish rate:
ros2 topic hz /rov/camera/image_compressed

# View sensor data:
ros2 topic echo /rov/sensor_data

# View ROV status:
ros2 topic echo /rov/status

---

TROUBLESHOOTING

"No frames received on laptop"

1. Check both machines have ROS_DOMAIN_ID=42 set:

   echo $ROS_DOMAIN_ID

2. Check firewall isn't blocking DDS multicast:

   sudo ufw allow in proto udp to 224.0.0.0/4
   sudo ufw allow in proto udp from 192.168.1.0/24

3. If on WiFi, some routers block multicast. Use a direct Ethernet connection or set up a Cyclone DDS config with unicast peers. See the "Cyclone DDS Unicast" section below.

Cyclone DDS Unicast Config (for problematic networks)

Create ~/cyclonedds.xml on BOTH machines:

<?xml version="1.0" encoding="UTF-8"?>
<CycloneDDS xmlns="https://cdds.io/config">
  <Domain>
    <General>
      <Interfaces>
        <NetworkInterface autodetermine="true"/>
      </Interfaces>
    </General>
    <Discovery>
      <Peers>
        <Peer address="192.168.1.100"/>  <!-- LattePanda IP -->
        <Peer address="192.168.1.101"/>  <!-- Laptop IP -->
      </Peers>
    </Discovery>
  </Domain>
</CycloneDDS>

Then add to ~/.bashrc:

export RMW_IMPLEMENTATION=rmw_cyclonedds_cpp
export CYCLONEDDS_URI=file:///home/$USER/cyclonedds.xml

Install Cyclone DDS:

sudo apt install -y ros-humble-rmw-cyclonedds-cpp

Camera index wrong

# Try different indices:
v4l2-ctl --list-devices
# or
ls /dev/video*

Low framerate over network

Edit src/rov_onboard/config/onboard_params.yaml:

• Reduce frame_width/frame_height (e.g., 320x240)
• Lower jpeg_quality (e.g., 30)
• Reduce fps (e.g., 15)

Then rebuild: colcon build --symlink-install

---

TODO

• Wire thrusters: Edit thruster_node.py with the actual motor driver code (serial to Arduino, direct PWM, etc.)
• Add real sensors: Edit sensor_node.py to read from the depth sensor, IMU, etc.
• Leak detection: Add a leak sensor node to rov_onboard
• Second camera: Add another camera_node instance with a different camera_index
• GUI dashboard: Upgrade dashboard_node.py to a PyQt5/tkinter or ideally html graphical display
• Autonomous features: Add new nodes for computer vision, PID control, etc.