Development of Data-Processing Framework for Transit Performance Analysis

A methodological data-processing framework is developed to process a massive amount of transit data, including vehicle location, passenger count, and electronic fare transactions. The developed data analysis methodology can allow a number of applications, such as transit route performance measurement, to support decision making for transit planning and operation. The data analysis methodology is demonstrated by the use of 1 month of archived transit data obtained from Metro Transit in the Twin Cities of Minnesota. A route-based transit performance analysis at time point level is discussed to evaluate route running time and schedule adherence. An application interface was developed to analyze bus adherence at time points and link travel time performance. The data-processing framework has the capability to support studies on transfer activities and transit rider's origin and destination inference. The developed data analysis methodology has demonstrated its capability to analyze transit performance and to support further research on other intelligent transit applications.