Effect of Vehicular Lanes on Pedestrian Gap Acceptance Behaviour

Walking is a key element of balanced transportation system that has often been ignored when planning any transportation facility. Pedestrians select appropriate gaps in vehicular stream depending on their demographic, vehicle and pedestrian behavioural characteristics and roadway geometry. If, pedestrians are crossing a typical urban two lane undivided road, then they may search more near gaps in traffic stream instead of far gaps. Whereas, pedestrians need to search vehicular gaps in each lane with respect to direction of travel (curb or median) while crossing more number of vehicular lanes (six lane divided road). Such pedestrian behavioural aspects need to be studied to understand the pedestrian gap acceptance behaviour while crossing more number of vehicular lanes. With this in mind, the objective of this study is to model pedestrian gap selection behaviour while crossing six lane divided road. The pedestrian gap acceptance behaviour has also been modeled to find out the pedestrian critical gap during road crossing by using multiple linear regression technique with effect of pedestrian behavioural characteristics. Probability of pedestrian gap acceptance has also been modeled considering pedestrian decision making process based on the discrete choice theory. To develop these models, field survey has been conducted at urban six lane divided un-controlled mid-block location at Worli in Mumbai, India. The real video data was used to set up pedestrian behavioural characteristics as explanatory variables for the model. The most important explanatory variables such as vehicular gap size, movement of pedestrian from the curb or median, rolling gap (pedestrian rolling over available small gaps), type of gap, pedestrian speed change condition and pedestrian waiting time have been included in the developed model. Further model transferability is checked with other location and it has been observed that the pedestrian gap acceptance behaviour models developed in this study represents quite well the pedestrian behaviour at urban six lane divided uncontrolled mid-block location. The inference of these models will be useful to assess existing facilities and suggest suitable corrective measures to improve pedestrian safety. (C) 2013 The Authors. Published by Elsevier Ltd.