Baltic Shoreline Pollution Monitor

What is this project about?

TheBaltic Shoreline Pollution Monitor is an autonomous, solar-powered buoy that acts as an intelligent environmental sentinel for coastal waters. Using computer vision, acoustic sensing, and edge AI, it detects pollution events like oil spills, microplastic clusters, and unusual underwater activity in real-time. Multiple buoys form a mesh network using LoRa radios, creating a distributed early-warning system for marine pollution.

Why did I decide to make it?

Growing up near the polluted shores of Lake Maracaibo in Venezuela, I witnessed firsthand how industrial runoff and neglect can devastate marine ecosystems. Now living near the Baltic Sea, a region once pristine, now increasingly threatened by shipping, industrial activity, and plastic waste, I've seen the same pattern emerge.

Traditional monitoring relies on expensive research vessels or satellite imagery, which often miss small-scale, fast-moving pollution events until it’s too late.

This project began with a simple idea: What if our coastlines had their own early warning system, a network of intelligent, self-deployable buoys that detect threats before they escalate?

Most commercial solutions cost hundreds of thousands of euros and require specialized operators. I wanted to design something modular, affordable (under €500/unit), and community-deployable and accessible to NGOs, universities, and local authorities alike.

How willit work?

Each buoy is a self-contained monitoring station built around a Seeed XIAO ESP32-S3 microcontroller. It combines multiple sensing modalities with edge AI processing:

• Vision: A Grove Vision AI V2 camera runs lightweight TinyML models to detect surface oil films, foam patterns, macro-debris, and water discoloration.
• Acoustics: An underwater hydrophone with preamp listens for unusual noise signatures such as illegal dumping, marine distress, or ship activity. It can also monitor marine life like whales.
• Positioning: An Air530 GPS module geotags sensor data and tracks buoy drift.
• Power: A solar panel and LiPo battery enable long-term, autonomous deployment.
• Connectivity: The Wio-SX1262 LoRa module enables long-range mesh communication using the Meshtastic protocol, allowing buoys to operate without cellular infrastructure.
• Storage: A MicroSD card provides local logging of high-resolution and diagnostic data.

Firmware Architecture

The firmware is being developed using FreeRTOS, enabling concurrent and reliable operation of all subsystems. A central SystemController singleton coordinates access to shared resources and handles inter-task communication.

Sensing Tasks

• Task_SenseAudio: Processes raw hydrophone signals and extracts acoustic features.
• Task_Vision: Runs TinyML inference on captured frames for pollution detection.
• Task_GPS: Polls GPS data for location tagging and drift monitoring.

Uplink Task

• Task_Uplink: Packages key events and transmits them via LoRa using a MeshLink driver compatible with the Meshtastic protocol.

Logging Task

• Task_Logger: Manages local file rotation and SD card logging for in-depth, timestamped data.

This modular, task-based architecture will ensure that the system is robust, scalable, and easy to maintain.Project Status & Disclaimer

Current Development Stage: While the core architecture, hardware selection, and software framework have been thoroughly designed and documented, the complete prototype is not yet ready for deployment. Due to time constraints, the final firmware integration and weatherproof enclosure design remain in development.

Why This Project Matters: Despite the current incomplete state, this project addresses a critical environmental challenge that affects not only the Baltic Sea but coastal regions worldwide. The combination of affordable hardware, edge AI processing, and mesh networking represents a paradigm shift in accessible environmental monitoring that communities desperately need

Commitment to Completion: The technical foundation is solid, and the environmental impact potential is too significant to abandon. This project will be completed and deployed, as coastal pollution monitoring cannot wait for perfect timing. The Baltic Sea will serve as our initial testbed, but the modular design ensures rapid expansion to other vulnerable coastal areas globally.

Next Steps

• Finalize waterproof housing and floatation system
• Complete Meshtastic firmware integration and networking tests
• Conduct shoreline trials in controlled environments along the Baltic coast

Future Work

This project serves as a powerful, extensible platform for environmental monitoring. Future versions could evolve into autonomous patrol vessels, offering mobile pollution detection. The data collected by swarms of these buoys can be used to train more advanced AI models, improving detection accuracy and enabling the identification of new types of pollutants. We also plan to build a cloud-based dashboard to visualize the real-time data from the entire network of buoys.