TerraNova: New Earth
A symbol of renewal - one scan, one action, one better Earth.

˙⋆✮ TerraNova – Scan. Sort. Save Earth.

Inspiration

Every day, millions of people want to do the right thing for the planet, to recycle, compost, or dispose of waste responsibly, but they don’t always know where their trash really belongs.
Mismanaged waste contributes to overflowing landfills, ocean pollution, and environmental degradation.

We built TerraNova to bridge this gap.
TerraNova means “New Earth,” and that’s exactly what we want to create.
Our application makes responsible waste disposal as simple as taking a picture, empowering everyone to take small, everyday actions that add up to meaningful environmental change.

What It Does

TerraNova transforms responsible disposal into a simple and educational experience. It helps users identify the correct waste category through ML-powered image recognition, rewarding them for every eco-friendly action and making sustainability fun, rewarding, and effortless.

How We Built It

Frontend Implementation

Framework: HTML, CSS, JavaScript
Purpose: Build an interactive, client-side web experience, all running directly in the browser.

Key Features

Responsive Design: Built with CSS Grid and Flexbox
Sections: Inspiration, How It Works, Scan Demo, Map, Galaxy, Disposal Guide, Team
Dark Theme: Smooth layout with card containers and contrast text for readability

Camera & Scanning

Uses the device camera via navigator.mediaDevices.getUserMedia()
Captures still frames onto a <canvas> element for image analysis

Machine Learning Integration

Loads MobileNet from TensorFlow.js
Classifies captured images
Matches them to disposal categories: plastic, glass, compost, etc.
Displays disposal instructions and confidence score directly in the browser

Map Integration (Leaflet.js)

Interactive map with mock recycling centers
Uses OpenStreetMap tiles (no paid API)
Geolocation support to show nearby drop-off sites

Reward Galaxy System

Adds a fun star counter and particle animation using HTML <canvas>
Rewards users for correct recycling actions
Progress stored in localStorage to save stars across sessions

Backend Implementation

Type: Fully client-side (no backend server)

Client-Based Logic

TensorFlow.js runs locally for image recognition
User progress (stars) stored via localStorage
Map locations are hard-coded in JavaScript (future-ready for API integration)
100% privacy-focused: No user data collected or sent to servers

Challenges We Ran Into

Getting the model to detect different material types through the camera correctly
Adjusting the star animation to scale smoothly with higher counts
Attempted to use a training model to make predictions, but ultimately decided on TensorFlow.js

Accomplishments We’re Proud Of

Built a functional, pitch-ready prototype in under 24 hours
Designed a clean, intuitive interface that provides instant recycling guidance
Created a user-friendly experience that encourages daily eco-friendly habits

What We Learned

Instant feedback increases user engagement
Rewards help reinforce positive environmental habits naturally

What’s Next for TerraNova

AI Recommendations: Suggest disposal tips based on item and location
Expanded Database: Cover more products and recycling facilities
Advanced Gamification: Add streaks, badges, and leaderboards
Social Features: Share eco-achievements with friends
Community Partnerships: Collaborate with schools, cities, and recycling programs
Impact Visualization: Display cumulative environmental benefits over time