Enhancing driving safety with customized music: A simulator study using entropy weight and Bayesian Networks.

Area of Science:

• Human-Computer Interaction
• Cognitive Psychology
• Traffic Safety

Background:

• Music listening is common while driving, influencing driver arousal and performance.
• Optimal music complexity for driving varies by scenario and individual.
• This study examines music's impact on drivers with varying characteristics during sudden events.

Purpose of the Study:

• To investigate how different driving characteristics influence responses to sudden events under various music conditions.
• To determine the effects of music tempo and lyrics on driving performance across driver skill levels.

Main Methods:

• Eight driving-related traits were combined into 'driving characteristics'.
• Twenty-eight drivers were grouped (qualified, good, excellent) and performed tasks with music (tempo, lyrics).
• Driving performance was scored using the Entropy Weight Method; Bayesian Networks analyzed collision risks.

Main Results:

• Music without lyrics (0.73±0.02) yielded better performance than music with lyrics (0.68±0.02).
• The excellent driver group (0.79±0.03) significantly outperformed good (0.66±0.02) and qualified (0.66±0.02) groups.
• An interaction showed slow-tempo music with lyrics was better than fast-tempo; music's impact on speed and collision varied by driving characteristics.

Conclusions:

• Superior drivers handle complex music better, but individual strategies cause performance variations.
• Drivers should choose music aligning with their capabilities for safer driving.
• Findings can inform personalized music recommendations for enhanced vehicle safety and driver response.