Real-time monitoring system for university buses using WiFi networks and travel time prediction

A real-time bus monitoring system helps in providing useful bus information such as location and estimated travel time to users. Many universities also provide free bus transportation service and bus monitoring systems inside their campus for students and faculty members. This thesis proposes utilizing existing on-campus WiFi network to provide connection of the bus monitoring system. This eliminates the need for cellular network which reduces implementation and maintenance cost. However, some areas in the university may not have WiFi coverage and can be identified by doing experiments to create a coverage map along each bus route. This leads to an analysis of bus data to find relevant factors such as traffic condition, correlation between neighboring sections, and historical data for predicting bus travel time and location along sections without data connection. From the results, travel time prediction using historical averaging or support vector machine can be applied in areas where there is no internet connection.