Applications of Automatic Vehicle Location systems towards improving service reliability and operations planning in London

Abstract

Technological advances in the transit industry, such as the introduction of Automatic Vehicle Location (AVL) systems, have provided agencies with robust data collection and measurement systems and enabled the development of comprehensive planning and operations tools. This thesis reviews the impact of an AVL installation in London, and demonstrates how data recorded by this system may be used to improve service reliability and operations planning on the London bus network. In particular, this thesis focuses on capabilities which would have been impossible prior to the installation of AVL. Service reliability has traditionally been measured from an operations perspective despite a major objective of the transit agency being to provide high quality service to passengers. A framework for a service reliability analysis is developed which explores new passenger-centric measures used to describe the quality of transit service, which are measured with AVL data. First, an analysis of trends in service reliability and factors that contribute to service reliability is performed in order to gain a better understanding of the environment in which transit services are operated. Three new passenger-centric measures of reliability are then introduced which describe the entire bus passenger experience. These measures are evaluated for a set of origin/destination pairs on six bus routes, and the differences in the perception of reliability between the new measures and the traditional measures are identified. The analysis demonstrates that while the traditional measures of reliability are relevant, the new measures provide additional insight. Recommendations are made with regard to implementing these new measures, providing better passenger journey information by applying these measures, and improving service control practices by monitoring these measures. AVL data provides for the development of more robust operations plans since these data allow for a greater degree of accuracy when measuring point to point running times. A framework is presented for how running times measured with AVL may be used to assess the efficacy of current operations plans and aid operations planners in vehicle scheduling. Recommendations are made with regards to how AVL data should be used to evaluate the effectiveness of existing schedules and develop more robust schedules.