Utilizing Traffic Signal Pedestrian Push-Button Data for Pedestrian Planning and Safety Analysis

Transportation planning, traffic monitoring, and traffic safety analysis require detailed information about pedestrian volumes, but such data are usually lacking. Fortunately, recent research has demonstrated the accuracy of pedestrian volumes estimated from push-button data contained within high-resolution traffic signal controller log data. Such data are available continuously for many locations. This project takes advantage of these novel pedestrian traffic signal data to advance pedestrian traffic monitoring and improve pedestrian traffic safety by applying them as estimates of volume and exposure, often alongside advanced machine learning techniques. Through a series of five studies, the authors identify temporal patterns in pedestrian activity; study the accuracy of pedestrian volume estimation methods over time; use machine learning methods to improve the quality and completeness of pedestrian time-series data; analyze crashes to identify a "safety in numbers" effect for pedestrians; and apply a new deep learning model to better understand factors affecting pedestrian crash severity. Altogether, this work leverages novel pedestrian traffic signal data to further research and efforts in pedestrian traffic monitoring and safety.

Language

  • English

Media Info

  • Media Type: Digital/other
  • Edition: Final Report
  • Features: Figures; Maps; References; Tables;
  • Pagination: 106p

Subject/Index Terms

Filing Info

  • Accession Number: 01923705
  • Record Type: Publication
  • Report/Paper Numbers: MPC-24-525
  • Contract Numbers: MPC-622
  • Files: UTC, NTL, TRIS, USDOT, STATEDOT
  • Created Date: Jul 8 2024 9:08AM