Why they take the risk to perform a direct left turn at intersections: A data-driven framework for cyclist violation...

Area of Science:

• Traffic Safety
• Transportation Engineering
• Behavioral Analysis

Background:

• Bicycle crashes at intersections are a major traffic safety concern.
• Direct left turns (DLTs) by cyclists, especially when mixed with vehicles, increase risk due to lack of exposure data.
• Understanding the mechanisms behind risky cycling behaviors like DLTs is crucial for prevention.

Purpose of the Study:

• To develop a framework for detecting direct left turn (DLT) events using bike-sharing data.
• To investigate the contributing factors influencing cyclists' DLT behavior.
• To inform strategies for reducing DLT rates and enhancing cyclist safety.

Main Methods:

• Proposed a direct left turn (DLT) detection framework utilizing bike-sharing trajectories.
• Employed a random parameters logit model with heterogeneity in means and variances (RPLHMV) to analyze DLT behavior.
• Statistical analysis of DLT cases to identify influencing variables.

Main Results:

• DLTs are more frequent on weekdays during peak commuting hours with high demand.
• Law-abiding cyclists are influenced by external factors, while risk-takers are driven by habits.
• Key factors include event time, passing time, cycling speed, waiting time, passing space, and preference for DLTs.

Conclusions:

• The study provides a novel method for DLT detection and analysis using bike-sharing data.
• Identified distinct behavioral patterns and influencing factors for DLT violations.
• Findings can guide the development of targeted interventions to improve intersection safety for cyclists.