The Michigan Dimensional Accident Surveilance Model is being developed by the Michigan Department of State Highways and Transportation. The model aims to objectively analyze the entire roadway system (not just locations with the worst accident histories), to select candidate locations for upgrading that are the most sensitive to correction, and to choose sets of the most likely cost-effective corrective measures. This paper introduces the procedures; reports on progress, accomplishments, and shortcomings to date; and stimulates related interest and development elsewhere. The model may be visualized as grouping all roadway segments with identical predetermined physical and accident characteristics into one cell of a multicell array. Subsequent statistical analysis, cost estimating, and accident predictions assess probable impacts on transforming all sites from one cell type to another. Data sources are several master tapes of accident reports, road features inventory files, and a traffic volume file. The identification of segments having a statistically significant numer of accidents and the determination of logical countermeasures work well. The prediction of expected accidents for each corrective action at present lacks precision and requires additional work. /Author/

Media Info

  • Media Type: Print
  • Features: Figures;
  • Pagination: pp 16-19
  • Monograph Title: Accident analysis, ride quality, driver education, and behaviour research
  • Serial:

Subject/Index Terms

Filing Info

  • Accession Number: 00194269
  • Record Type: Publication
  • Files: TRIS, TRB
  • Created Date: Jun 13 1979 12:00AM