Driving safety evaluation for hazardous materials vehicle drivers based on visual characteristics

OBJECTIVE: To develop a driving safety evaluation model for hazardous materials vehicle drivers based on visual characteristics. METHODS: Twenty-three professional hazardous materials vehicle drivers were recruited to participate in driving simulation experiments under normal, cognitively distracted, and operationally distracted conditions. Based on the Big Five personality questionnaire, K-means clustering was applied to classify the drivers into three types: balanced, calm, and impulsive. Significance analyses were conducted to identify differences in visual characteristics among driver types across driving states. An evaluation model was subsequently constructed by combining the Analytic Hierarchy Process (AHP)-Entropy Weight Method with fuzzy comprehensive evaluation, and the model was validated with case studies. RESULTS: Under distracted driving, significant differences were observed in single fixation duration, cumulative fixation duration, fixation frequency, single saccade duration, blink frequency, and pupil diameter compared to normal driving. Pupil diameter emerged as a universal sensitive indicator across all personality types, while the response patterns of visual characteristics to distraction type varied heterogeneously by personality. The model evaluations for the three types-balanced, calm, and impulsive-were highly consistent with the drivers' personality traits and visual behavior changes. CONCLUSION: This study integrates personality traits with visual characteristic parameters, filling a gap in the visual-based safety evaluation of hazardous materials vehicle drivers. The proposed AHP-Entropy Weight fuzzy comprehensive evaluation model provides a scientific and reliable method for the safety management of hazardous materials road transportation and for personalized driver safety early warning systems.