Driver Vision Based Perception-Response Time Prediction and Assistance Model on Mountain Highway Curve

Area of Science:

• Human-computer interaction
• Road safety engineering
• Cognitive psychology

Background:

• Current driving assistance systems lack human-like responsiveness.
• Predicting driver perception-response time (PRT) is crucial for enhancing safety on complex roads.
• Mountain highway curves present unique challenges for drivers and require specialized attention.

Purpose of the Study:

• To develop a predictive model for driver PRT on mountain highway curves.
• To design an assistance model that positively influences driver PRT.
• To identify key factors affecting driver PRT in this environment.

Main Methods:

• Field tests collected real-time driving data and driver vision information.
• A driver-vision lane model was developed to quantify curve elements within the driver's visual field.
• A multinomial log-linear model predicted PRT using environmental, visual, and vehicle data.

Main Results:

• The driver-vision lane model and specific visual elements significantly influenced PRT.
• The developed assistance model demonstrated a positive impact on driver PRT.
• Visual geometry, guidance, and information integrality of curves were identified as critical factors.

Conclusions:

• Driver-vision interaction is a key determinant of PRT on mountain curves.
• Optimizing visual design and information delivery in curves enhances driver performance.
• Human-centered design principles are vital for advanced driving assistance systems.