Modeling dilemma zone at urban signalized intersections using crowdsourced trajectory data.

Area of Science:

• Traffic Safety Engineering
• Driver Behavior Analysis
• Transportation Systems

Background:

• The stop/go dilemma at yellow traffic signals poses significant risks for rear-end and red-light running collisions.
• Existing methods for defining the dilemma zone, particularly Type II, have limitations, while Type I requires high-quality trajectory data often difficult to obtain.
• Accurate modeling and evaluation of the Type I dilemma zone dynamics represent a critical research gap.

Purpose of the Study:

• To accurately quantify the Type I dilemma zone using a large dataset of crowdsourced vehicle trajectory data.
• To address limitations in existing dilemma zone modeling and evaluation methods.
• To analyze driver behavior within the dilemma zone to assess collision risks and inform safety strategies.

Main Methods:

• Utilized a large sample of crowdsourced vehicle trajectory data for dilemma zone quantification.
• Implemented quantile regression to integrate driver-vehicle attributes into stopping and clearing distance calculations.
• Analyzed driver behavior in the approach area to identify potential collision risks.

Main Results:

• The Type I dilemma zone is consistently formed at very high vehicle approach speeds across 15 intersection approaches.
• The proposed method achieved low root mean squared errors of 14.8 ft and 25.1 ft for zone boundary estimation.
• Demonstrated the superiority of the proposed method over existing dilemma zone quantification techniques.

Conclusions:

• Crowdsourced data and quantile regression provide an accurate method for quantifying the Type I dilemma zone.
• Understanding dilemma zone dynamics is crucial for developing effective intersection safety and signal timing strategies.
• This research offers an empirical foundation for enhancing traffic safety and reducing intersection crashes.