An Automated Process of Identifying High-Risk Roads for Speed Management Intervention
Submission Date: November 10, 2016 Journal
Infrastructure Risk Rating (IRR) is a significant input to the speed management framework, set to be introduced as part of NZ Transport Agency’s Speed Management Guide. It is a road assessment methodology designed to assess risk based on infrastructure elements and interactions with surrounding land use, independent of crash history. The road safety risk is assessed by coding each road and roadside feature; such as land use, road stereotype and alignment; that feeds into the IRR model so that a risk rating can be determined. The methodology was originally developed as a manual coding exercise using street view imagery. However, this approach is neither economic nor time efficient when applied across a large network as is the requirement of the speed management framework.
This paper presents a geospatial process to automate the calculation of IRR. The process utilises various national and regional geospatial datasets to extract road features needed to calculate IRR. A comparison of the automated process outputs with manually coded IRR data of the same network resulted in a matching rate of almost 90 percent, hereby confirming the validity of the automated process. Aside from demonstrating the true potential of transport related data, this innovative approach will enable road controlling authorities to efficiently identify parts of their network where speed management intervention is most likely to reduce road trauma.