Back to Resources

Research Report 483 Crash prediction models for signalised intersections: signal phasing and geometry

Published: | Category: Safety, security and public health , Research programme , Research & reports | Audience: General

In most cities and towns, the majority of crash black-spots occur at major intersections. Given this, crash reduction studies often focus on the major signalised intersections. However, there is limited information that links the phasing configuration, degree of saturation and overall cycle time to crashes.

While a number of analysis tools are available for assessing the efficiency of intersections, there are very few tools that can assist engineers in assessing the safety effects of intersection upgrades and new intersections.

Data from 238 signalised intersection sites in Auckland, Wellington, Christchurch, Hamilton, Dunedin and Melbourne were used to develop crash prediction models for key crash-causing movements at traffic signals. Separate models were built for peak periods and for motor vehicles and pedestrians.

The key crash types that were analysed were right-angle, right-turning, lost-control and rear-end type crashes.

Keywords: traffic safety, crash prediction models, signalised intersections, pedestrian and cycle crashes, New Zealand data collection

Publication details

  • Author:
  • Published: June 2012
  • Reference: 483
  • ISBN/ISSN: ISBN 978-0-478- 39442-9 (electronic)