Automated People Mover System (APMS) at Soekarno-Hatta airport is an automated passenger carrier that serves between terminals by train based on the Communication Based Train Control (CBTC) system. This system allows the train to run automatically without the driver. To achieve this state of automation, the control system requires information on the position and speed of the train continuously. Although any sensors used to measure this data are quite accurate, it still produces error in the measurement. In this paper, a sensor fusion method with Extended Kalman Filter (EKF) will be proposed. It will combine measurement data from several sensors to get better result than what would possibly be achieved when these sensors were used individually. The sensors that will be fused in this study are tachometer and accelerometer. Real time experiment results on the APMS running on its track at the airport showed that the EKF method gives the Root Mean Squared Error (RMSE) value of 5.89. This method provides better result when compared to the tachometer whose RMSE value is 7.12 and accelerometer whose RMSE value is 7.41.