GNSS + IMU + Magnetometer Sensor Fusion for Wearable Motion Tracking

Author: Egbe Chigozie Michael
Role: Surveyor | Geospatial Specialist | Sensor Fusion Enthusiast | Data Analyst
Status: Simulated Concept — Work In Progress
Tech Stack: Python, Numpy, Matplotlib

---

🚀 Project Overview

This repository demonstrates a simulation-based approach to fusing multiple sensor data streams—GNSS, IMU, and magnetometer—for estimating the real-time kinematics (position, velocity, and orientation) of a moving object such as a professional footballer.

The goal is to simulate a simplified version of the sensor fusion pipeline used in elite athlete wearables to match FIFA EPTS tracking standards.

---

🎯 Key Features

• 🔁 Error-State Kalman Filter (ESKF) implementation
• 📍 GNSS simulation for absolute position updates
• 📈 IMU (accelerometer + gyroscope) for high-rate motion tracking
• 🧭 Magnetometer for orientation correction
• 🔧 Optional barometer input for altitude tracking (extension planned)
• 📊 Visual plots comparing raw GNSS vs fused estimates

---

📦 Folder Structure

sensor-fusion-football-tracker-simulation/
├── fusion_simulation.py
├── data/
│   ├── simulated_gnss.csv
│   ├── simulated_imu.csv
│   └── simulated_mag.csv
├── results/
│   └── fusion_output_plot.png
├── README.md
└── requirements.txt

---

📚 Learning Goals

This simulation is part of a learning journey to understand how high-accuracy motion tracking systems work in sports wearables. The next goal is to:

• Test the pipeline with real sensor logs from Android smartphones or wearable IMUs
• Port the algorithm to C++ for real-time performance

---

🤝 Let's Collaborate

I'm actively looking for remote sensor fusion roles in sports tracking, drones, autonomous navigation, or GIS.
Feel free to fork, suggest improvements, or reach out via www.linkedin.com/in/chigozie-michael or https://www.upwork.com/freelancers/~01a849aa50ac79b416.