Advanced urban public transportation system for Indian scenarios

Abstract

This article presents a part of the work being carried out under the Department of Science & Technology (DST)-funded project, titled Advanced Urban Public Transportation System. The solutions designed for real-time tracking of metro buses in urban areas of India, detecting bus-stops automatically, and predicting the arrival time of buses accurately are elaborated. The proposed system addresses the challenges related to GPS outage, unknown schedule and stoppages of buses, and unavailability of real time traffic information along the bus-route. The system is evaluated using multiple-trip data collected over a 32 kilometer long route during the peak and off-peak hour traffic conditions. The bus-stop detection accuracy of 75% (6 out of 8 bus-stops are accurately detected) could be achieved using an arbitrary set of trips conducted over the route. The arrival time prediction error of 7% (5 minutes) has been reported. The scalability assessment of the system shows that it can support the transit of more than ten thousand buses and over one million subscribers/commuters.