From Space to Street — Detecting Bengaluru's Hidden Waste Crisis
LitterLens uses satellite imagery and AI to detect illegal garbage dumps across Bengaluru and auto-reports them to BBMP — no site visit, no manual photos, no hassle.
Built for INNOVATEX 4.0 — AppForge | SDG 6: Sustainable Cities & Communities
- Open the app — satellite map loads at your location
- Long-press any area to scan it
- AI (YOLO model via Roboflow) analyzes the satellite tile
- Detected dumps appear as color-coded pins on the map
- Tap a pin — see waste type, area, confidence, ward info
- One tap — pre-filled complaint sent to BBMP via WhatsApp or Email
Detection to government notification in under 30 seconds.
| Layer | Technology |
|---|---|
| Framework | React Native + Expo SDK 54 |
| Language | TypeScript |
| Maps | react-native-maps (Satellite View) |
| AI Detection | Roboflow Serverless API (YOLO) |
| Satellite Tiles | Google Maps Static API |
| Backend | Firebase (Firestore + Auth) |
| Location | expo-location |
| Reporting | WhatsApp + Email Deep Links |
# Clone
git clone https://github.com/0xMoni/LitterLens.git
cd LitterLens
# Install
npm install
# Add your API keys in src/constants/config.ts
# - Roboflow API key (free at roboflow.com)
# - Google Maps API key (console.cloud.google.com)
# - Firebase config (console.firebase.google.com)
# Run
npx expo startScan the QR code with Expo Go on your phone.
Other garbage apps need you to walk up to trash and take a photo. LitterLens finds dumps remotely from satellite view — proactive detection at city scale, not reactive point-and-shoot.
- Satellite-first detection (no physical presence needed)
- Full pipeline in one app: Detect → Map → Report → Track
- Hyper-localized for Bengaluru with BBMP ward mapping
- Zero hardware — uses existing Google Maps satellite imagery
MIT