Analysis of Historical Travel Time Data

http://dx.doi.org/10.13023/KTC.RR.2015.122015PDFTech ReportKTC-15-12/SPR12-444-1FTraffic volumeTraffic congestionTravel timeTime-based performance measuresPrivate sector speed dataLink-referenced ATP dataMAP-21United StatesUniversity of Kentucky Transportation CenterChen, MeiZhang, XuGreen, EricUniversity of Kentucky Transportation CenterUniversity of Kentucky. College of EngineeringKentucky Transportation CabinetUnited States. Federal Highway AdministrationUS Transportation CollectionTravel speed is a critical piece of information for many applications. The Moving Ahead for Progress in the 21st Century Act (MAP-21) states that calculating travel speeds illuminates the performance of the nation\u2019s highway system. New technologies make the collection of speed data more straightforward than ever. Private vendors collect and sell this data, and KTC purchased speed data for 2010-2013 from NAVTEQ (now HERE). The main objectives of the research were to: Evaluate private sector speed data and its use in generating travel time based performance measures, Create a mechanism to integrate this speed data with networks maintained by KYTC and Kentucky Metropolitan Planning Organizations (MPOs) to facilitate congestion management and travel model improvement, Generate performance measures, including travel time index, planning time index, buffer index, annual hours of delay, and percentage travel under congested conditions This study assessed private sector speed data, its potential as a robust data source, and its limitations. Evaluation of the data indicated that link-referenced ATP data offered the best value for a wide range of applications. It offered details on speed distribution and provided critical insights into the dynamics of congestion and the variability of travel times. Among the three types of data, the link-referenced ATP data should be the first choice when future purchases of private sector speed data are made. These data provided critical support to develop the performance measures required by MAP-21. These data need to be linked with traffic volumes to generate the full range of performance measures. Data remain sparse for roads with lower functional classifications, especially collectors and local streets. When sample size is a concern, the research team recommends that data from other sources (such as Bluetooth, radar, and others) be used to supplement the private sector speed data. A range of congestion and reliability performance measures have been generated from these data and were provided to KYTC and MPO stakeholders in the form of geodatabases. Other applications can benefit from these data, including: the calibration and validation of simulation models, travel demand models, and air quality analyses.75