Big Transportation Data Analytics

The world of Big Data is moving fast. There are several private sector Big Data vendors using data from vehicle probes, and from Wi-Fi, cellular, and Bluetooth data traces to synthesize transportation information such as vehicle speeds and origin-destination travel flows. In addition, Utah Department of Transportation (UDOT)'s Traffic Operations Center recently just purchased HERE traffic probe data to assist traffic management and road network performance assessment. UDOT also possesses Big Data in the form of Performance Measurement System (PeMS), that has been archived since 2008. This historic data set should be mined to learn new things about the direct and proximate causes of traffic volume change and travel reliability. There is also an urging need to develop analytical approaches that leverage existing historic data to reveal methods for estimating traffic, with the potential to reduce the burden of the annual traffic count program. UDOT maintains an annual traffic count program which requires the acquisition of hundreds of short-duration counts each year. This traffic count effort represents a significant cost to UDOT, while also exposing UDOT staff to the dangers inherent to being exposed to traffic. The proposed research will seek to determine whether statistical modeling and/or machine learning methods might be employed to partially or fully supplant the short-duration traffic count program and, in doing so, reduce the effort, cost, and staff exposure of UDOT's traffic count program. The primary objective of this research is to determine whether statistical modeling and/or machine learning can be applied to for estimating/predicting traffic conditions (e.g. vehicle miles of travel (VMT) and reliability) when it is conflated with other available data sets, such as demographics, income, highway capacity, and highway improvement projects. The secondary objective is to develop analytical methods that can be used by UDOT in the future to estimate traffic volumes and reliability for short-duration traffic count sites, based on other available data, and quantitative relationships. A range of analytical methods will be attempted, from well-established methods (e.g. traffic engineering techniques) to more sophisticated methods, such as advanced statistical modeling and machine learning.


  • English


  • Status: Completed
  • Funding: $90000
  • Contract Numbers:


  • Sponsor Organizations:

    Office of the Assistant Secretary for Research and Technology

    University Transportation Centers Program
    Department of Transportation
    Washington, DC  United States  20590
  • Managing Organizations:

    Mountain-Plains Consortium

    North Dakota State University
    Fargo, ND  United States  58108
  • Project Managers:

    Tolliver, Denver

  • Performing Organizations:

    University of Utah, Salt Lake City

    College of Engineering, Department of Civil Engineering
    Salt Lake City, UT  United States  84112-0561
  • Principal Investigators:

    Liu, Xiaoyue Cathy

  • Start Date: 20171115
  • Expected Completion Date: 20220731
  • Actual Completion Date: 20210322
  • USDOT Program: University Transportation Centers Program
  • Source Data: MPC-543

Subject/Index Terms

Filing Info

  • Accession Number: 01651461
  • Record Type: Research project
  • Source Agency: Mountain-Plains Consortium
  • Contract Numbers: 69A3551747108
  • Files: UTC, RIP
  • Created Date: Nov 24 2017 12:52PM