🌕 Lunar.OS – Lunar Orbital Reconnaissance & Navigation System

Lunar.OS is an AI-powered lunar terrain navigation system that uses a U-Net segmentation model to detect hazardous terrain and the A* algorithm to generate safe navigation paths. The project features a FastAPI backend and a React + Vite frontend, fully containerized with Docker.

🔗 Live Demo

• Live UI: https://lunar-navigation-system.vercel.app
• Backend API docs: https://sajidftw-lunar-backend.hf.space/docs

Note: the hosted demo only has a curated subset of sectors loaded (IDs: 1, 151, 301, 451, 601, 751, 901, 1051, 1201, 1351, 1501, 1651, 1801, 1951, 2101, 2251, 2401, 2551, 2701, 2851). Other sector IDs will return a 404 on the live demo. Run locally via Docker (below) for access to the full dataset.

🚀 Quickstart: Run via Docker (Recommended)

You don't need Python, Node.js, or any local dependencies installed to run Lunar.OS. You can run the entire pre-built system using Docker.

1. Create a file named docker-compose.yml anywhere on your computer and paste this configuration:

services:
  backend:
    image: sajid1108/lunar-backend:latest
    ports:
      - "7860:7860"

  frontend:
    image: sajid1108/lunar-frontend:latest
    ports:
      - "5173:5173"
    depends_on:
      - backend

2. Open a terminal in that same folder and start the system:

docker compose up

3. Access the application:

• Frontend UI: http://localhost:5173
• Backend API Docs: http://localhost:7860/docs

🛠️ Tech Stack

• Backend: FastAPI, PyTorch, U-Net, A* Pathfinding
• Frontend: React, Vite
• Deployment: Docker, Docker Compose

📦 Docker Images

• Backend: sajid1108/lunar-backend

• Frontend: sajid1108/lunar-frontend

  FINALY YEAR PROJECT FOR B TECH