Inspiration
India generates millions of tons of waste every year, but a large portion of it is not properly segregated due to lack of awareness and easy tools. Many people want to recycle correctly but don’t know which bin an item belongs to or where to dispose it nearby.
GreenLens India was inspired by this everyday problem and the need for a simple, AI-powered solution for waste segregation.
What it does
GreenLens India helps users identify waste items using a camera scan powered by Gemini AI.
The app:
- Detects the type of waste (plastic, paper, organic, e-waste, etc.)
- Suggests the correct disposal method
- Provides eco-tips for sustainable alternatives
- Shows nearby waste collection centers using Google Maps
- Supports English and Hindi for better accessibility in India
How we built it
- Used Gemini 3 Flash for image understanding and waste classification
- Built the frontend using React + TypeScript
- Implemented camera-based item scanning
- Integrated Google Maps to show nearby recycling and collection centers
- Designed a clean, mobile-friendly UI focused on ease of use
Challenges we ran into
- Handling different types of waste with limited training data
- Designing prompts that give accurate and helpful responses from Gemini
- Making the UI simple while still informative
Accomplishments that we're proud of
- Successfully integrating Gemini AI for real-world environmental use
- Creating a functional end-to-end prototype during the hackathon
- Making the app multilingual (English + Hindi) for inclusivity
What we learned
- Prompt engineering for multimodal AI
- Practical integration of AI into frontend applications
- Designing user-focused sustainability tools
What's next for GreenLens India
- Support for more Indian languages
- User accounts and recycling history
- Rewards and gamification for eco-friendly behavior
- Collaboration with municipal waste services
Built With
- camera
- css
- gemini-3-flash
- google-ai-studio
- google-maps
- html
- react
- typescript
Log in or sign up for Devpost to join the conversation.