Nov 2025

Altitude Sensor Fusion with GNSS

Kalman filter project fusing smartphone GPS and barometric data to produce cleaner altitude estimates than either sensor alone.

Final report

Kalman Filter
GNSSLogger
Phyphox
Python

Altitude heatmap over a 2D GPS path using Kalman-filtered altitude data

Altitude Sensor Fusion with GNSS hero image

Altitude comparison of barometer, Kalman filter, and GPS estimates over time

Building 310 benchmark disk and data collection spot

Campus elevation map with route colored by elevation

Altitude heatmap over a 2D GPS path using fused location provider data

Altitude heatmap over a 2D GPS path using GPS-only data

Summary

This project fused smartphone GPS and barometric pressure measurements with a Kalman filter to produce cleaner altitude estimates for campus elevation mapping. GPS provided an absolute altitude reference but was noisy and low-rate in the vertical direction, while the barometer tracked short-term elevation changes more smoothly but could drift with weather and calibration error. By calibrating the pressure-derived altitude and tuning the filter to use GPS as the long-term anchor, the final estimate preserved the barometer's local smoothness while remaining tied to a usable absolute reference for route interpretation. My role was designing the trials, running reliability tests, and comparing the fused data.