🚢 Illegal Fishing Hotspot Detection Using Machine Learning

Mining AIS Big Data to Identify Hidden Fishing Activities in the Mediterranean Sea

---

📌 Project Overview

Illegal, unreported, and unregulated (IUU) fishing threatens marine ecosystems and global food security. This project uses Machine Learning + AIS (Automatic Identification System) vessel data to detect hidden fishing patterns and identify possible illegal fishing hotspots in the Mediterranean Sea.

We process AIS data, detect suspicious behavior (like identity gaps), cluster dangerous zones, and visualize hotspots on an interactive map using Streamlit, Folium, and Heatmaps.

A KNN-based prediction system is also included to check whether a new vessel belongs to any hotspot.

---

🧠 Key Features

✔️ AIS Data Preprocessing

• Sorting by MMSI and timestamp
• Calculating gaps between consecutive signals
• Detecting “identity hiding gaps” (AIS turned off)

✔️ Suspicious Vessel Detection

• Identifying vessels with abnormal behavior
• Filtering real fishing-related points

✔️ Hotspot Clustering (DBSCAN)

• Density-based clustering to detect illegal-fishing zones
• Automatic centroid extraction
• Cluster quality metrics (Silhouette Score, Davies–Bouldin Score)

✔️ KNN Prediction Model

Predicts the cluster of a new vessel based on:

• Latitude
• Longitude
• Speed
• MMSI

If the vessel is too far from known clusters → returns Cluster = -1 (Safe / Not suspicious)

✔️ Interactive Streamlit Web App

• ⚡ Heatmap visualization
• 📍 Cluster centroid markers
• 📄 Raw data display
• 🤖 Live vessel cluster prediction
• 🖤 Black theme with custom UI
• 📷 Optional Mediterranean Sea background image

---

🗂️ Tech Stack

Machine Learning

• Python
• Scikit-Learn
• DBSCAN
• K-Nearest Neighbors (KNN)
• Clustering metrics

Data

• AIS Big Data (simulated dataset of ~30,000 rows)

Backend

• Python
• Pandas
• NumPy

Frontend / Visualization

• Streamlit
• Folium
• HeatMap Plugin
• Custom HTML/CSS styling

---

📌 Architecture / Pipeline

AIS Big Data → Clean & Sort Data →
Detect AIS Gaps → Extract Suspicious Points →
Cluster Using DBSCAN → Generate Hotspot Centroids →
Train KNN Model → Deploy Streamlit Dashboard →
Predict Vessel Hotspot Risk

---

🚀 How To Run This Project

1️⃣ Clone the Repository

git clone https://github.com/your-username/illegal-fishing-hotspots.git
cd illegal-fishing-hotspots

2️⃣ Install Dependencies

pip install -r requirements.txt

3️⃣ Run the Streamlit App

streamlit run app.py

4️⃣ Open in your browser

Usually at:

http://localhost:8501

---

🔮 Future Improvements

• Integrate real-time AIS streaming
• Use satellite imagery for validation
• Implement LSTM-based vessel behavior forecasting
• Add anomaly detection (Isolation Forest)
• Build a full maritime risk dashboard

---

👨‍💻 Author

Kevin Manjila
Machine Learning Developer

---