Cost Effective Strategies for Estimating Statewide AADT

Annual Average Daily Traffic (AADT) is defined as the average daily measure of the total volume of vehicles on a roadway segment over a year. AADT is one of the most important traffic measures used in any transportation engineering project (e.g., highway investment decision making, transportation planning, highway maintenance, air quality compliance study, traffic safety analysis, and travel demand modeling). Thus, the accuracy of AADT estimation is critical for any transportation problems that use AADT as an input parameter. This report covers the various aspects of AADT data collection and estimation. The researchers reviewed existing legislation for AADT data collection, including Highway Performance Monitoring System and Highway Safety Improvement Program. A survey, regarding current data collection procedures and needs, was distributed to both U.S. States and Canadian Provinces as well as a second version for South Carolina cities and counties. A review was also conducted on each available technology for data collection from simple but effective pneumatic tubes to high tech video systems. From this information, the research team developed recommendations for data collection improvements in South Carolina. Lastly, this project developed and evaluated three AADT estimation models and determined that a Machine Learning model, Support Vector Regression, had the most cost-effective and accurate AADT prediction. Based on this finding, the research team developed a computer software, called estimAADTion, which can be used by the South Carolina Department of Transportation (SCDOT) to predict AADT based on 24-hour short-term counts.

• Record URL:
  https://www.scdot.scltap.org/wp-content/uploads/2019/04/SPR-717-Final-Report.pdf
• Corporate Authors:

  Clemson University

  Department of Civil Engineering
  Clemson, SC  United States  29634

  University of South Carolina, Columbia

  Department of Civil and Environmental Engineering, 300 South Main Street
  Columbia, SC  United States  29208

  South Carolina Department of Transportation

  Office of Materials and Research
  1406 Shop Road
  Columbia, SC  United States  29201

  Federal Highway Administration

  1200 New Jersey Avenue, SE
  Washington, DC  United States  20590
• Authors:
  • Chowdhury, Mashrur
  • Huynh, Nathan
  • 0000-0002-4605-5651
  • Khan, Sakib Mahmud
  • Khan, Md Zadid
  • Brunk, Katherine
  • Torkjazi, Mohammad
  • 0000-0002-2445-6024
  • Islam, Sababa
  • Keehan, McKenzie
  • Shiri, Samaneh
• Publication Date: 2019-2

Language

• English

Media Info

• Media Type: Digital/other
• Edition: Final Report
• Features: Appendices; Figures; Photos; References; Tables;
• Pagination: 148p

Subject/Index Terms

• TRT Terms: Accuracy; Annual average daily traffic; Cost effectiveness; Data collection; Estimating; Machine learning; Software; States; Surveys
• Uncontrolled Terms: Support vector regression
• Geographic Terms: Canadian Provinces; South Carolina
• Subject Areas: Data and Information Technology; Highways; Planning and Forecasting;

Filing Info

• Accession Number: 01702883
• Record Type: Publication
• Report/Paper Numbers: FHWA-SC-18-10
• Contract Numbers: SPR No. 717
• Files: TRIS, ATRI, USDOT, STATEDOT
• Created Date: Apr 24 2019 4:44PM