This paper proposes a fusion framework to locate trains travelling on track routes. The input data are gained from two independent sensor devices, namely, a Global Navigation Satellite System and an eddy current sensor device. The sensor data are fused by an Extended Kalman Filter gaining precise information about the train location. This positioning system combines the sensor devices with the fusion approach to perform a robust location even in the case of noise or when a sensor fault occurs. Additionally, some future fusion strategies to extend the existing location system are presented.