Multi-Hop LoRa Mesh Network & AI-Powered Oceanographic Monitoring Control Center

---

🌊 Project Overview

FishTrack is a state-of-the-art research project designed for marine life tracking and oceanographic telemetry monitoring. By leveraging LoRa Multi-Hop Mesh Networks and Edge Computer Vision (YOLO), the system enables cost-effective, long-range underwater surveillance and ecological monitoring in coastal zones.

The project is developed in collaboration with research data and taxonomy guidelines from the Bureau of Fisheries and Aquatic Resources (BFAR) Regional Fisheries Office - MIMAROPA (Provincial Fishery Office, Boac, Marinduque, Philippines).

📡 Multi-Hop Mesh Topology

Because high-frequency radio waves degrade rapidly over open water and physical obstructions, FishTrack implements a custom multi-hop routing protocol. Edge buoy nodes float on the ocean, capture marine data, and relay messages through each other until they reach the land-based ATLAS Base Station.

Figure 1: Conceptual visualization of the multi-hop LoRa buoy mesh network communicating with the coastal ATLAS Base Station.

---

🚀 Key Features

• Self-Healing Ad-Hoc Routing: Dynamically maintains neighbor tables, handles packet TTLs, deduplicates messages, and supports relayed transmissions.
• Edge YOLO Fish Detection: Smart buoys execute object detection on underwater cameras to classify fish species.
• Gemini AI Taxonomical Verification: Integrates with Gemini AI to map raw YOLO classifications to official BFAR MIMAROPA taxonomy and local common names (e.g., mapping Epinephelinae to "Lapu-Lapu").
• Acoustic Sonar Scanning: Captures sonar data to detect underwater movements and map seabed profiles.
• Reliable Chunked Transmission: Implements compression (zlib) and chunked packets with custom ACKs to transmit larger data streams (like images or sonar sweeps) over low-bandwidth LoRa.
• Real-time Flask Dashboard: Sleek Web GUI featuring network stats, live telemetry (GPS & Battery), active commands control center, and audit logging.

Figure 2: Simulated underwater camera feed illustrating the YOLO detection bounding box matched with local BFAR taxonomic labels.

---

📊 System Architecture & Data Flow

Below is the network message flow tracing a detection event from a buoy to the Base Station web dashboard:

sequenceDiagram
    autonumber
    participant Buoy as Buoy Node (e.g., BUOY-POSEIDON)
    participant Relay as Buoy Relay Node (Optional)
    participant Base as ATLAS Base Station (main.py)
    participant AI as Gemini AI API
    participant Web as Web Dashboard (Flask Server)

    Note over Buoy: Camera triggers YOLO detection
    Buoy->>Buoy: Classifies species & compresses telemetry
    Buoy->>Relay: Sends LoRa packet (TTL=5, MSG_TYPE_DATA)
    Note over Relay: Validates checksum & decrements TTL
    Relay->>Base: Forwards packet to Base Station
    Note over Base: Reassembles chunks & decompresses JSON
    Base->>AI: Sends YOLO label for taxonomic mapping
    AI-->>Base: Returns BFAR MIMAROPA local common name
    Base->>Base: Saves file on disk & updates cache
    Base->>Web: Renders live event via Web API
    Note over Web: Operator views "Lapu-Lapu" detection on HUD

Loading

---

📂 Codebase File Structure

The Base Station system is organized into modular Python files:

File	Description
main.py	System Entry Point. Starts the main thread loops: LoRa serial listener, Flask web server, scheduler, and network maintenance loop.
config.py	Central Configuration. Hardware settings (COM ports, baud rates), message types, directories paths, and command registries.
hardware.py	LoRa Interface. Manages half-duplex serial connection to the Raspberry Pi Pico + LoRa transceiver, handling TX/RX mode switching.
logic.py	Network Logic. Manages the NetworkState class, tracks active buoys, monitors battery logs, routes/forwards packets, and reassembles incoming chunks.
scheduler.py	Command Dispatcher. Schedules and queues manual or periodic commands to buoys (e.g., requesting sonar scans), handling retry logic and queues.
analytics.py	Analytics Engine. Bridges computer vision outputs with the BFAR MIMAROPA database and queries Gemini AI for taxonomical verification.
server.py	Web Server. Flask implementation serving API routes, tracking audits, and rendering dashboard templates.
utils.py	Utilities. Support functions for packet zlib compression/decompression, SHA256 checksum validation, and unique message ID generation.

---

⚙️ Installation & Setup

1. Prerequisites

• Python 3.8 or higher installed.
• A connected LoRa transceiver module (via USB/Serial, e.g., a Raspberry Pi Pico acting as a LoRa controller).

2. Install Dependencies

Install the required Python modules:

pip install flask requests pyserial

3. Hardware Configuration

Open config.py and verify the following variables match your physical setup:

• SERIAL_PORT: Set to your LoRa controller COM port (e.g., "COM4" on Windows or "/dev/ttyUSB0" on Linux).
• BAUD_RATE: Set to match your controller firmware (default 115200).
• ACTIVE_BUOY_LIST: Ensure expected buoy codenames are registered.

4. Running the Base Station

Run the main script to bootstrap the system:

python main.py

This command initializes the LoRa controller, spawns background threads for maintenance and command scheduling, and starts the local web server.

5. Accessing the Dashboard

Once running, open your web browser and navigate to:

http://localhost:8080

---

📜 Official BFAR MIMAROPA Citation

This research project adheres to data guidelines and taxonomy records provided by:

Republic of the Philippines Department of Agriculture Bureau of Fisheries and Aquatic Resources (BFAR) Regional Fisheries Office - MIMAROPA Provincial Fishery Office, Boac, Marinduque Noted by: Joel G. Malabana, OIC, PFO

💖 Donate & Support

If you find this project useful and would like to support the deployment of FishTrack buoys for coastal fishing communities, donations are greatly appreciated!

Bitcoin Address: 13zWnp2ty3NPzAXX9QxwEeoPSKhN5tPzic