SAFE is a high-fidelity wildfire intelligence platform designed for real-time fire behavior prediction, regional risk assessment, and emergency response coordination. It utilizes the Rothermel Surface Fire Spread Model integrated with real-time environmental data to provide actionable safety analytics.
The platform features a Hybrid Intelligence System designed for maximum availability and low latency during emergencies:
The intelligence server implements a sequential failover chain across multiple state-of-the-art AI providers to bypass individual quota limits:
- Groq (Primary): Sub-second technical responses using Llama 3.
- Google Gemini: High-fidelity environmental reasoning.
- OpenAI / Mistral: Robust technical documentation analysis and fallback.
- Local Fallback: Instant technical summary if all external APIs are throttled.
- Local Layer: Instant response for core platform identity and greetings using word-boundary regex matching.
- Cloud Layer: High-fidelity technical analysis routed via the Multi-API server.
SAFE features the Concord Engine, a state-of-the-art modular simulation suite:
- 3D Topographic Mapping: High-performance rendering of terrain with procedurally generated elevations and regional vegetation zones.
- Physics-Based Spread: Implementation of the Rothermel model accounting for wind vectors, fuel Surface-Area-to-Volume (SAV), moisture damping, and topographic slope.
- Dynamic Mitigation: Interactive tools for modeling Helitack water drops and defensive Fireline construction.
- Frontend: Vite, React, Three.js (React Three Fiber), Framer Motion.
- Backend: Node.js, Express, LangChain, OpenAI/Gemini/Groq SDKs.
- Mapping: Google Maps JavaScript API.
- Simulation Logic: Modular Concord TypeScript engine.
LANDING PAGE
LIVE MAP SIMULATION
UNCONTROLLED-CONTROLLED SIMULATION
RESOURCES
ASSISTANT
Developed for the "Reboot the Earth" Hackathon.