FishNet

About the Project — FishNet: Shedding Light on Hidden Oceans

We were inspired by the environmental sustainability track, driven by a shared goal to protect marine ecosystems from illegal fishing. Every year, millions of tons of seafood are caught outside the law, devastating fish populations and threatening coastal livelihoods. We wanted to build something that makes the invisible visible — a system that can expose vessels that “go dark” to hide illicit activity.

What We Learned

We dove deep into real-time maritime AIS (Automatic Identification System) data and learned how to combine AI-driven pattern analysis with graph-based network detection to identify suspicious vessel behavior. We also explored how small irregularities in transmission signals can reveal much bigger environmental crimes.

How We Built It

Our backend pipeline processes millions of AIS records, detects “dark events” (periods where vessels turn off their trackers), and correlates them with other nearby ships using proximity indexing and community graph analysis.

Challenges We Faced

We faced challenges with sparse and inconsistent data, where signals often dropped for legitimate reasons. Data cleaning — filtering noise and incomplete records without losing key patterns. Distinguishing normal from suspicious behavior, requiring careful calibration and validation.

Built With

• ais
• flask
• github
• javascript
• python
• react
• vscode
• wdpa
• websockets