Identifying Causes of Performance Issues in Bus Schedule Adherence with Automatic Vehicle Location and Passenger Count Data

Abstract

Automatic vehicle location (AVL) and automatic passenger counting (APC) systems can provide rich archived databases for analysis. Previous work has focused on using AVL–APC data to evaluate system performance using various quantitative performance measures and data visualization methods. Given the large volume of data, there is a benefit to automating the creation of performance measures and data visualizations and “pushing” interesting information to users, rather than requiring users to create the performance measures and figures and sift through them on their own. This paper presents a methodology for identifying bus stops that are not meeting performance standards for schedule adherence and the factors that cause inadequate performance. The methodology is designed to be automated and therefore can be applied efficiently to AVL–APC data for an entire transit network. Use of this proposed method will enable transit agencies to identify service quality issues and their root causes more efficiently.