Citation

Aron M, Seidowsky R, Cohen S. Transp. Res. C Emerg. Technol. 2013; 28: 168-180.

Copyright

(Copyright © 2013, Elsevier Publishing)

DOI

10.1016/j.trc.2010.12.006

PMID

unavailable

Abstract

Motorway operators deploy dynamic lane management operations to reduce recurrent congestions. The aim of this paper is to contribute to the safety assessment of such an operation which has been introduced on a French urban motorway weaving section. It consists in the opening of an auxiliary lane partly taken on the hard shoulder. The assessment is based on the observational before-after method in road safety, which compares the empirical number of accidents with a prediction of the number of those which would have occurred, had no change happened. A comparison site is used to identify the trend on the number of accidents between the before and after periods, independently of the effect of the operation. A variant of this method, which takes into account the traffic regimes, is developed here. This variant removes some bias, which may exist when the traffic patterns on the comparison site and on the assessed site are different. We highlight here how critical is the choice of the comparison site; how significant is the correlation between the risks of accidents by type and the traffic regimes, because that correlation, if changing over time, induces a bias; how meaningful is the provided analytic expression of the bias. Although inconclusive - due to the small number of recorded accidents - concerning the final safety impact of the operation, the main questions are discussed - reduction in traffic density leading to a reduction in the number of accidents, migration of traffic towards the opening hours, migration of density and accidents downstream. This contributes to a better understanding of the evolution of traffic safety and of the safety impact of the lane management operation.