Abstract

ABSTRACT
Justifying and prioritizing roadway lighting: A Case Study of Quebec Highways
Mahnoush Heydari

Roadway lighting is an effective countermeasure capable of reducing night-time motorized collisions under the right circumstances. Its initial viability can be learnt through collision modification factors showing beneficial effects of roadway lighting on local roads. However, this requires time-series of data from several years before and after the implementation of lighting. There is a need to estimate collision modification factors of lighting from locally observed cross sectional data from as little as one year. There is also a lack of a practical method capable of replacing the complicated warrant system to support decisions of whether or not to illuminate roads. Such method should be able to identify and prioritize segments that will benefit the most from being illuminated. This research presents a method to estimate collision modification factors with as little as one year of data. In addition, this research presents a practical method that identifies and prioritizes candidate road segments for being illuminated. A case study of Quebec's highways found that lighting is an effective countermeasure and that expected benefits approximate 60% reduction in night time collisions. It was found that segment size plays an important role and that Bayesian data fusion can be used to abstract from segment size to estimate a generic collision modification factor. It was found that safety performance functions for desired land use and sites type can be used in combination with the observed number of collisions to classify those sites expected to observe benefits from being illuminated.