Research Papers (2009 – 2013)

Abstract

The construction of a bicycle path is expected to improve safety of non-motorized road-users by guiding the turning movements of vehicles. This study applies stochastic models to estimate the critical gap or lag acceptance of vehicles and pedestrians, before and after the construction of bicycle path at the signalized intersection. A case study of Boulevard de Maisonneuve Ouest and Rue Mackay in Sir George Williams Campus of Concordia University is presented. A Back propagation Neural Network model (BPN) and a Binomial logit model (BNL) are applied to determine the contribution of the characteristics of road users and traffic condition to the vehicle’s left-turning decision and pedestrian’s crossing decision. The study reveals that critical gap acceptance by vehicles was increased, while critical lag acceptance by pedestrians was decreased after the construction of bicycle path. The BPN model identifies that bicycle attributes have significant influence on the road user’s decision after the construction of a bicycle path. The BNL model identifies the following bicycle attributes: the presence at conflict zone, the distance from the conflict point and speed, as the significant estimators of critical gap for left turning vehicles after the construction of a bicycle path. The BNL model also estimates the significant impact of bicycle attributes to the pedestrian lag acceptance decision after the construction of bicycle path. This study concludes that; the construction of bicycle path, not only enforce the vehicle to be cautious during left-turning, but also encourages the pedestrians to cross the road with decreased lag duration.

Shohel Reza Amin, Ciprian Alecsandru, Luis Amador-Jimenez