David Levinson, Associate Professor, Civil Engineering
The widespread implementation of automated vehicle location (AVL) systems and automatic passenger counters (APC) in the transit industry has opened new venues in transit operations and system monitoring. Metro Transit, the primary transit agency in the Twin Cities, Minnesota region, has been testing various intelligent transportation systems (ITS) since 1999. In 2005, it fully implemented an AVL system and partially implemented an APC system, but in the following years, there was little effort to employ such data to evaluate different aspects of performance. This research capitalized on the availability of such data to better assess performance issues of one particular route in the Metro Transit system. The researchers used the archived data from the location systems of buses running on an example cross-town route to conduct a microscopic analysis of performance and reliability issues. They generated a series of analytical models to predict run time, schedule adherence, and reliability of the transit route at two scales: the time point segment and the route level. The methodology included multiple approaches to display ITS data within a GIS environment to allow visual identification of problem areas along routes. The methodology also used statistical models generated at the time point segment and bus route level of analysis to demonstrate ways to identify reliability issues and what causes them. The analytical models showed that while headways are being maintained, schedule revisions are needed in order to improve run time, and suggested that many scheduled stops along this route are underutilized and so should be consolidated.