Modeling lane-transgressing behavior of e-bike riders on road sections with marked bike lanes: A survival analysis approach

Abstract

Objective: Due to limited road space, e-bike riders are likely to cross the division mark on a road section and ride on the motor lane, which becomes a major cause of traffic accidents involving e-bikes. The objective of this study is to qualitatively and quantitatively investigate the relationship between lane-transgressing behavior and factors including traffic condition and road geometric design.

Methods: A survival analysis approach is proposed to model e-bikers’ lane-transgressing behavior, including the Kaplan-Meier curve and the Cox proportional hazards model. The Kaplan-Meyer non-parametric regression is utilized to analyze the influence of each factor on the lane-transgressing probability. The Cox proportional hazards model is applied to thoroughly evaluate multidimensional influencing factors effects.

Results: The speed of e-bikes, the width of the non-motorized lane, the density of non-motorized vehicles, and the traffic condition of the adjacent motorized lane have significant influences on the lane-transgressing behavior. The lane-transgressing risk drops from 83.6% to 28.8% when the speed decreases from 35 km/h to 20 km/h. With wider non-motorized lanes or lower non-motorized vehicle density, the risk can be reduced by over 60%. Besides, when traffic condition of motorized vehicles changes from high turbulence condition to low turbulence condition, the lane-transgressing risk can be increased by 144.5%.

Conclusions: E-bikers behave differently under the effects of various factors. In the perspective of the safety and the design, these effects need to be considered. The width of the non-motorized lane is recommended to be no less than 260 cm. The physically separated facilities, such as guardrail, are highly recommended to avoid the interference between the motorized vehicles and bicycles. This study provides a reference for a guidance for the management of the mixed traffic flow with e-bikes.

Keywords:

• Lane-transgressing behavior
• electric bicycle
• survival analysis
• Kaplan-Meier non-parametric regression
• Cox proportional hazards model
• safety

Data availability

The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.

Additional information

Funding

This study was jointly supported by the National R&D Plan of China (2017YFC0803906), the National Natural Science Foundation of China (Grant No. 51408049).