Modeling Variations in Bus Schedule Adherence at the Stop Level on an Urban Arterial Road

Abstract

The purpose of this study is to assess the adherence of buses to their schedules when they are operating in specific conditions made by separating buses from fast general traffic by median islands, but still sharing curbside lanes with slow traffic (e.g., cars and motorcycles) on an urban arterial road with heavy traffic in Taiwan. This study uses real-time traffic data (e.g., travel time, speed, and traffic volume) collected for a 3-month period in 2010 from buses equipped with a global positioning system (GPS) and inductive loop vehicle detectors buried under the road. The ordered probit and multinomial logit models are applied to examine variables contributing to trip time between two adjacent stops being later than the scheduled trip time during different time periods of the day. The findings indicate that variables significantly influence variations in adherence to schedule, including buses operating in mixed-traffic curbside lanes, the number of bus routes passing through a stop, bus dwell time at a stop, bus headway, traffic volume, the number of intersections without traffic signals, spacing of bus stops with fewer than 300 m between them in densely populated areas, and rainy weather.