Effects of built environment on bicycle wrong Way riding behavior: A data-driven approach

Area of Science:

• Urban planning
• Transportation engineering
• Traffic safety

Background:

• Bicycle wrong way riding (WWR) poses significant traffic safety risks.
• Lack of exposure data hinders understanding of WWR and its relation to the built environment (BE).

Purpose of the Study:

• To develop a framework for detecting WWR using bike-sharing data.
• To analyze the impact of BE factors on WWR frequency.
• To inform targeted bicycle safety improvement strategies.

Main Methods:

• Utilized bike-sharing trajectory data from Chengdu, China.
• Employed a Negative Binomial-based Additive Decision Tree model.
• Investigated relationships between BE factors and WWR occurrences.

Main Results:

• WWR distribution is independent of daily time periods.
• Road length is a more significant factor than road level or direction.
• Key BE factors influencing WWR include proximity to companies, transit stops, residences, and catering facilities.
• Educational facilities emerge as influential during non-peak hours.
• Non-linear effects of BE variables on WWR frequencies were observed.

Conclusions:

• Built environment characteristics significantly influence bicycle wrong way riding.
• Geographically specific policies are necessary for effective bicycle safety improvements.
• Understanding non-linear BE impacts is crucial for WWR mitigation.