Developing a bus service reliability evaluation and visualization framework using archived AVL / APC data

Transit service reliability is important for both transit agencies and passengers. Transit agencies are interested in ways of measuring and evaluating their service quality, identifying reliability problems, understanding causes and effects of unreliable services, and proposing strategies to improve overall service quality. However, the vast amount of automatic vehicle location (AVL) and automatic passenger counter (APC) data have not been fully utilized in many transit agencies. Therefore, the objective of this study is to develop a framework that can translate massive AVL and APC data into valuable reliability measures that can help transit agencies to evaluate and visualize their service quality and identify reliability problems in the route level. Both aggregated performance measures and detailed operational records are calculated and visualized based on the archived AVL and APC data. Aggregated performance measures are useful in identifying recurrent reliability problems and evaluating performance, such as schedule deviation, headway variation, and bus bunching incidents; while detailed operational records are helpful to understand specific problems, examples include time-space diagram and animation of bus movements. Results show the proposed examples can help identify reliability problems.