Event-Based Modeling of Driver Yielding Behavior to Pedestrians at Two-Lane Roundabout Approaches.

Area of Science:

• Traffic Engineering
• Human Factors in Transportation
• Road Safety

Background:

• Drivers often fail to yield to pedestrians at roundabouts, unlike other intersection types.
• Understanding factors influencing yielding behavior at roundabouts is crucial for pedestrian safety.

Purpose of the Study:

• To identify factors affecting driver yielding to pedestrians at two-lane roundabouts.
• To model the likelihood of driver yielding based on identified contributing factors.

Main Methods:

• Logistic regression modeling applied to 1150 pedestrian crossings at six roundabouts.
• Analysis of factors including pedestrian visibility (white cane), vehicle lane position, approach speed, and roundabout leg (entry vs. exit).

Main Results:

• Yielding is more likely at roundabout entry legs than exit legs.
• Drivers yield more to pedestrians with white canes and less at higher speeds.
• Far-lane drivers and those turning right at exit legs show lower yielding probabilities.

Conclusions:

• Driver yielding behavior at roundabouts is influenced by specific factors, with speed and pedestrian visibility being key.
• Pedestrian waiting position (curb vs. street) did not significantly impact yielding rates.
• Findings can inform traffic engineering practices and microsimulation models for improved roundabout safety.