Community-Based Ocean & Coastal Cleanup Management System

Group 8 · Institute of Technology, University of Moratuwa.

A platform connecting citizens, volunteer divers, government authorities and organisations around one loop for Sri Lanka's coastline: report a polluted site → let the community verify it → an administrator approves it → the government authority approves it and it becomes a cleanup project owned by the reporter → clean it up and record the result.

Volunteer divers (with certificates) and organisations (with a website link) are verified by an administrator before they can sign in.

├── frontend/   React + Vite, built on the Tideline design system
└── backend/    Spring Boot 3.3 + JPA + JWT, H2 for dev and MySQL for the target deployment

Run both

cd frontend && npm install && npm run dev

cd backend && ./mvnw spring-boot:run

Frontend on http://localhost:5173, API on http://localhost:8080. The frontend reads and writes live API data — Vite proxies /api to the backend, so there is no CORS setup in development. Start the backend first; every seeded account uses password123, and the login screen lists them.

See frontend/README.md and backend/README.md for details, demo accounts and the full endpoint list.

Module coverage

#	Module	Frontend	Backend
1	User registration & profile	yes	yes (certificate upload, admin account verification)
2	Pollution reporting	yes	yes (photo and video evidence upload)
3	Community verification & trust	yes	yes (75% threshold, configurable)
4	Admin review & moderation	yes	yes
5	Government approval workflow	yes	yes
6	Location-based alerts	yes	yes (Haversine radius + escalation steps)
7	Project progress monitoring	yes	yes
8	Diver employment & opportunities	yes	yes
9	Reporting & analytics	yes	yes

Still to do

• Geocoding for location validation, and an embedded pollution map (screens link to OpenStreetMap for now).
• Scheduled reminders to volunteers (module 6) and a complaints channel (module 4).
• A hosted deployment — the backend's Dockerfile is ready (see backend/README.md), and the image still needs a run against real MySQL.