ecovision

Inspiration Our team was inspired by the growing disconnect between people's desire to live sustainably and their uncertainty about how to do so effectively. We noticed three key problems:

Confusion about waste sorting
Difficulty tracking sustainable habits
Lack of motivation

What it does EcoVision Tracker is a comprehensive sustainability companion app with four core features:

AI Trash Scanner: Upload photos of waste items for instant AI classification (recycle, compost, or landfill) with educational tips and "Buds" rewards
Eco Actions Tracker: Log daily sustainable actions like using reusable containers or taking public transit
EcoWallet: Earn "Buds" (eco-currency) for sustainable actions and redeem them for badges, merchandise, or donations
AI Chatbot: Get instant answers to sustainability questions, receive tips on eco-friendly living, and learn how to get the most out of EcoVision's features.

How we built it We built EcoVision using a modern tech stack:

Frontend: React with TypeScript, Vite, and Tailwind CSS for a responsive, intuitive UI
Backend: Flask API with Google's Gemini AI integration for image classification
Database: Supabase (PostgreSQL) for user data, actions, and transactions
Authentication: Clerk for secure user management
State Management: React Query for efficient data fetching
UI Components: shadcn/ui and Radix UI primitives for accessible, beautiful components

The trash scanner was our most technically challenging feature, using Google's Gemini AI to analyze images and provide accurate waste classification with educational context.

Challenges we ran into

AI Integration: Configuring the Gemini API for accurate waste classification required extensive testing and prompt engineering
Database Schema Design: Creating a flexible yet efficient schema for tracking diverse eco actions and rewards
CORS Issues: Resolving cross-origin resource sharing problems between our frontend and backend
Deployment Complexity: Ensuring our Flask backend worked correctly in production environments
Transaction Consistency: Maintaining consistency between user stats and transaction history in the database

Accomplishments that we're proud of

Creating an AI-powered waste classification system that's both accurate and educational
Designing a gamified reward system that makes sustainability engaging and motivating
Building a clean UI that makes complex sustainability concepts accessible
Implementing a backend that handles diverse data types and user interactions
Successfully integrating multiple technologies into a cohesive application

What we learned

AI Integration: How to effectively use generative AI for image classification and educational content
Full-Stack Development: Coordinating frontend and backend development for a seamless user experience
Database Design: Creating efficient schemas for complex, interrelated data
User Experience: Designing interfaces that make sustainability accessible and engaging
Deployment: Configuring applications for production environments

What's next for EcoVision Tracker EcoVision is committed to scaling its platform and impact. In the future, we would want to introduce community features to foster collaboration, integrate with local recycling programs, provide advanced analytics on environmental impact, and launch mobile apps for accessibility. We will also enhance our AI with more waste data and regional recycling details.