Using Municipal Vehicles as Sensor Platforms to Monitor the Health and Performance of the Traffic Control System

In this study, sets of video imagery are obtained from transit buses in regular service on The Ohio State University campus and processed into time-of-day traffic volume estimates on major campus roadways. The traffic volume estimates are obtained from vehicles identified in the video imagery using an approach previously developed by the investigators to translate vehicles identified from sensors mounted on mobile platforms into traditional traffic volume estimates. Hourly volumes estimated from the processed video imagery are compared to hourly volumes obtained from concurrently collected road tube data. The results are seen to be encouraging and agree with the underlying hypothesis that volumes obtained from bus-based imagery are similar to short-duration traffic counts, which will be noisy for any one observation but which can lead to good estimates of time-of-day traffic volumes when aggregated over the multiple observations that can be obtained from a transit bus. When longer time-of-day periods are considered or when the volume estimates are used to produce estimates of vehicle distance travelled across a set of roadway segments, performance of the estimates obtained from the video imagery improves further. In addition to showing the promise of using already available video imagery obtained at low cost from transit buses in regular service, this study included important educational and outreach components. Specifically, the empirical validation study used to assess the performance of the proposed approach was designed in the context of a term project for a transportation data acquisition and analysis class.

Language

  • English

Media Info

  • Media Type: Digital/other
  • Edition: Final Report
  • Features: Appendices; Figures; Photos; References; Tables;
  • Pagination: 32p

Subject/Index Terms

Filing Info

  • Accession Number: 01760023
  • Record Type: Publication
  • Contract Numbers: 69A3551747111
  • Files: UTC, TRIS, ATRI, USDOT
  • Created Date: Nov 23 2020 11:50AM