Driving behavior at midblock crosswalks with Rectangular Rapid Flashing Beacons: Hidden Markov model approach using...

Area of Science:

• Traffic Safety
• Human Factors in Transportation
• Road Safety Engineering

Background:

• Pedestrian fatalities are rising, particularly on urban roads away from intersections.
• Rectangular Rapid Flashing Beacons (RRFBs) are used to enhance pedestrian visibility and driver awareness at midblock crosswalks.
• Existing research often focuses on binary driver yielding behavior, neglecting dynamic deceleration patterns.

Purpose of the Study:

• To analyze driver deceleration behavior at crosswalks using Hidden Markov Models (HMMs).
• To investigate the impact of RRFB activation and pedestrian presence on driver braking patterns.
• To inform the design of advanced driver-assistance systems for vehicle-pedestrian crash avoidance.

Main Methods:

• Utilized naturalistic driving data from the Safety Pilot Model Deployment (SPMD) program.
• Applied four HMMs to analyze braking and jerk data, considering RRFB status and pedestrian presence.
• Converted time-based data to distance-based using a moving window approach for analysis.

Main Results:

• Confirmed high driver compliance with activated RRFBs when pedestrians were present.
• Observed that 20% of drivers slowed to under 20 mph even without pedestrians at activated RRFBs.
• Found drivers initiated braking earlier (180m) and ceased braking sooner (70m) at activated RRFBs without pedestrians, but braked longer and more firmly with pedestrians present.

Conclusions:

• Activated RRFBs significantly influence driver deceleration, enhancing safety at midblock crosswalks.
• Driver behavior at crosswalks is modulated by RRFB signals and pedestrian presence, indicating potential for technological intervention.
• Understanding dynamic braking patterns is crucial for developing effective countermeasures against vehicle-pedestrian crashes.