Rail JU logo in white
European Union flag

A body of the
European Union

FP2 – R2DATO - Project Outputs

Railway Localization with Magnetic Field Distortions: Calibration, Sensor Fusion, and Evaluation

Jun 10, 2026

Output type:

ION GNSS+ 2026 conference

This paper proposes a method for achieving meter-level train localization using only speed measurements and disturbances in the Earth’s magnetic field caused by railway infrastructure. A Kalman filter combines wheel encoder data with a magnetic-field-based localization and calibration algorithm to estimate train position. Tested on real data from Deutsche Bahn’s TrainLab in Berlin, the approach demonstrates high accuracy, achieving a position error below 0.5 meters, showing the feasibility of magnetic field-based train localization.

Download output file

Europe's Rail