Author(s)

M. A. Dagli, O. Ozdemir, E. Dede, L. Akcil

Abstract

In the public transportation industry remarkable Big Data is formed. The data which are generated from tracking the vehicles shall be used for different kinds of purposes. The arrival time of the vehicles is one of the most desired data for the passengers. In this study, big data which is produced for tracking the vehicles at the Command Centre are used. The T1 tram line of Istanbul City is segmented into 300 parts using RFID sensors and each part is observed for estimation processes. Also, for computing the location accurately, odometer data of the train is gathered. Elapsed time of each train at each part of the line is inspected for the previous 2 hours. Then the average of the elapsed time of each part is calculated. To determine the remaining time of the closest trains, the location of all trains is observed. Finally, according to the distance between the trains and the station, the arrival time of a train is calculated. As test laboratory generated data for each station are displayed on a web environment. For observation of estimation accuracy, estimated times are compared with the real arrival times. As a result, ninety percent accuracy is obtained. This study will generate the data for the information display on the train platforms.

Keywords

tram tracking, Big Data, estimation of arrival, railway vehicle, RFID, passenger information system