Inspiration

• UN Development Goal #13: Climate Action
• When we focused on this goal, we wanted to provide a unique method of providing climate action through intelligent data collection.

How We Built It

FIRMS (Fire Information for Resource Management System)

• NASA’s API detects heat signatures via satellite imagery
• Provides real-time wildfire coordinates
• Front end selects specific wildfires by coordinates

OpenWeatherMap API

• Gathers weather data (cloudiness, humidity, precipitation, temperature, wind)
• Helps predict wildfire behavior using real-time conditions

AI Models Integration

• Cohere: Analyzes locations and environmental data to generate insights
• Gemini: Predicts wildfire spread and radius using real-time data

Database (MongoDB)

• Stores 1000+ wildfire documents
• Includes location, weather, and user data
• NoSQL design for efficient backend queries

Front End (React, Tailwind, Vite)

• Minimalist UI with smooth animations
• Login built with JavaScript
• Deployed on Vercel

Backend (Flask, MongoDB, APScheduler)

• Handles authentication, data storage, and updates
• APScheduler automates daily updates (24-hour delay for first update)

Automation & Data Updates

• Scheduler updates wildfire and weather data every 24 hours
• Initial 24-hour delay after startup

Security & Performance

• API keys and DB URIs stored in environment variables
• Optimized queries for performance
• Robust error handling for stability

Wildfire Spread Analysis

• Predicts spread using wind, humidity, temperature
• AI-based computation for accuracy

What's Next

• Expanded Input Support: Add videos, PDFs, and more file types for analysis
• Multi-language & Accent Support: Improve inclusivity and global adaptability

View on Devpost!