An algorithm for determining in which public transport vehicle a passenger is traveling

Abstract

This thesis presents an algorithm, that can determine in which public transport vehicle a passenger is traveling. The algorithm makes use of data such as Global Positioning System (GPS) location, vehicle length, bearing, speed, acceleration, Wi-Fi Positioning System (WPS) and planned itinerary. The algorithm collects the passenger location every few seconds, and queries all the nearby vehicles. Every vehicle is given a vote based how well the data match the data of the passenger. After a series of iterations the vehicle consistently appearing as a likely candidate will be returned. Tests show that the algorithm successfully can determine in which vehicle a passenger is traveling, in less than 1 minute. These tests were performed on an implementation of the algorithm that only uses GPS data. It can thus be concluded that the idea for the algorithm is sound.