Technex Green Hackathon 2026 | Problem Statement 1
SortSmart AI is an AI-powered waste classification system that identifies waste from images, maps it to the correct disposal category, calculates carbon footprint savings, and displays sustainability analytics — all in real time.
India generates 62 million tonnes of waste annually. Less than 20% is correctly segregated. The rest ends up in landfills, releasing methane — 25x more potent than CO₂. SortSmart AI closes this gap by making correct waste disposal instant, effortless, and impactful.
- AI Waste Classification — Upload any image, get instant category detection
- 3 Category Mapping — Recyclable, Biodegradable, Hazardous
- Carbon Footprint Calculator — Auto-estimates CO₂ saved per item
- Trees Equivalent — Translates CO₂ into relatable real-world impact
- Sustainability Dashboard — Real-time analytics for citizens, campuses & municipalities
- Circular Economy Score — Second life suggestion for every waste item
Future Vision
- Community Leaderboard — Campus/city-wide rankings based on waste correctly segregated, encouraging healthy competition and sustained engagement
| Layer | Technology |
|---|---|
| AI Model | Roboflow — 99.1% accuracy, 52 waste classes |
| Backend | Python + FastAPI |
| Frontend | React.js + TailwindCSS + Recharts |
| Camera | OpenCV |
| Database | Firebase Firestore |
- Python 3.10+
- Node.js 18+
- Webcam (for live scanning)
python -m venv venv
venv\Scripts\activate # Windows
source venv/bin/activate # Mac/Linux
pip install fastapi uvicorn inference-sdk opencv-python numpy
cd backend
uvicorn main:app --reloadcd frontend
npm install
npm start| User | What They Get |
|---|---|
| Citizens | Personal carbon impact and disposal guidance |
| Campus Admins | Aggregate waste patterns and analytics |
| Municipalities | City-wide data to optimize collection routes |
- SaaS licensing to municipalities — ₹50,000/month per city zone
- Per-bin software licensing — ₹5,000/bin/year
- Carbon credit brokerage — 10% commission on verified credits sold
- Break-even at 50 smart bins deployed