Satellite Tracker – AI-Powered Space Surveillance System

Satellite Tracker is an intelligent satellite monitoring platform that leverages Artificial Intelligence, Kafka, and microservices to provide predictive and real-time insights into satellite operations. It is designed to assist in managing satellite fleets by predicting their health status, end-of-life, and potential collision risks, and by visualizing their operational conditions through an interactive dashboard.

---

Table of Contents

• Overview
• Key Features
• System Architecture
• Microservices Overview
• UML Diagrams
  • Class Diagram
  • Sequence Diagram
  • Use Case Diagram
• Kafka Data Flow
• MongoDB Collections
• UI Mockups
• Project Folder Structure
• Future Improvements

---

Overview

The Satellite Tracker system connects to external APIs (like N2YO) to collect live satellite telemetry and orbital data. The data undergoes preprocessing before being sent to AI-powered microservices that analyze the information and return predictions.

These predictions are then routed to the backend services that store and serve the results to users via an interactive frontend.

---

Key Features

Health Monitoring

• Uses machine learning models to analyze telemetry data such as battery voltage, temperature, and data transmission rate.
• Predicts potential failures and recommends preventive maintenance.

End of Life Estimation

• Predicts the remaining operational lifespan of satellites based on orbital characteristics and motion analysis.

Collision Detection

• Continuously monitors distances between orbiting satellites.
• Generates alerts for possible close approaches or collision risks.

---

System Architecture

Main Technologies Used:

• Backend: Django (Python), Spring Boot (Java)
• Data Streaming: Apache Kafka
• Databases: MongoDB
• AI Models: Python-based ML models (scikit-learn)
• Frontend: Nextjs (or any frontend consuming API)

---

Microservices Overview

Microservice	Description
Data Processing Microservice	Fetches data from N2YO, cleans and prepares it.
Kafka	Message broker for asynchronously delivering data between services.
Models Microservice	Contains and executes AI models for health, EOL, and collision prediction.
Prediction Microservice	Stores and serves predictions for visualization.
User Data Microservice	Manages user accounts and linked satellites.
Frontend	Displays visual insights and predictions to end-users.

---

🧬 UML Diagrams

📘 Class Diagram

Represents the structure of MongoDB collections used in the project:

• User collection with NORAD IDs
• satellites, collision_alerts, Eol_collection, etc.

---

📗 Sequence Diagram

Demonstrates the data flow through the microservices(healthStatusFlow):

1. Data is fetched from N2YO
2. Sent to Kafka
3. Received by the Django service for model processing
4. Predictions sent to Spring Boot services
5. Data exposed to frontend

---

📙 Use Case Diagram

Illustrates how different actors (admin, user) interact with the system.

---

System Realization – Application Mockups

Below are interface mockups that represent the actual user experience of the Satellite Tracker platform. Each screen corresponds to a key functional component in the system, from user onboarding to predictive visualization.

System Realization – Application Mockups

Below are screenshots of the core interfaces that make up the Satellite Tracker user experience:

---

Sign In

Secure login page for existing users to access their satellite dashboard.

---

Sign Up

Registration form allowing users to create a new account and start tracking satellites.

---

3D Dashboard

Displays a 3D globe showing real-time positions of satellites and basic operational stats.

---

Health Status Page

Presents AI-predicted satellite health data, including:

• Battery voltage
• Solar panel temperature
• Attitude control error
• Transmission rate

---

End-of-Life Page

Shows machine learning-based predictions of the remaining operational lifetime for each tracked satellite.

---

Collision Alerts Page

Displays warnings for satellites at risk of collision, including:

• Distance to other objects
• Collision time estimate
• Severity level

Example of kafka running and the workflow of data

In Fetching Data Service in Spring Boot which fetches data from n2yo.com and Send it to Preprocessing service in Django

In the main server of django

Data sent by Fetching Data to Eol Service

---

Video Demo

https://drive.google.com/drive/folders/1QxA6MUNRYBNzJgKfu4MvqYYhY4_3JfhM?usp=sharing