Real-time adaptive arrival prediction system for public transport

Abstract

This paper introduces a low-cost, real-time prediction system for public transport arrival times. Using a predefined roads and stops network, arrival times are predicted for each public transport line. After the predictions have been made, the time for future arrivals is adapted in order to reflect current road conditions and any possible delays of future arrivals. The entire system is controlled using a dedicated online application. Sensitive location information can be encrypted and stored on a separate server.