Abstract Manager

Post-implementation evaluations in road safety often struggle for one of two reasons: the multi-year timeframe needed to collect a robust set of crash data, and the number of sites required to draw reliable conclusions.

This project aims to overcome these two challenges in what is the largest known post-implementation evaluation using video analytics. The project involved first using video analytics to diagnose road safety issues at 50 intersections, then implementation of low cost risk mitigation countermeasures, and finally, using video analytics again to measure risk reduction. The 50 intersections were distributed across five cities with 10 intersections each in Toronto, Vancouver, Calgary, Halifax, and Hamilton.

In total, 119 targeted safety improvements were implemented, representing an average of just over 2 implementations per intersection. The average attained risk reduction was 85%. The most common implementations included pavement markings, left-turn traffic calming, signage improvements, speed zone signs, and right turn radius reductions. Signal related countermeasure implementations included leading pedestrian intervals and protected left turns. Some areas in Vancouver which saw a marked improvement in bicycle safety also saw a huge increase in cycling volumes.

Some treatments do not have crash-based CMFs, which means that the only way to know their effectiveness is to conduct post-implementation evaluations. This national project led by Miovision shows one way to successfully conduct rapid post-implementation evaluations at scale to better plan road safety work in the future.