Risk of crashes among self-employed truck drivers: Prevalence evaluation using fatigue data and machine learning...

Area of Science:

• Occupational Health
• Transportation Safety
• Data Science

Background:

• Transportation industry shifts towards outsourced labor, impacting self-employed truck drivers.
• Extended working hours contribute to driver fatigue and elevate crash risks.
• Prevalence of substance use (smoking, alcohol, drugs) among surveyed drivers is high.

Purpose of the Study:

• Investigate factors contributing to fatigue and impairment in truck driving.
• Develop a machine learning (ML) model to predict traffic crash risk for truck drivers.
• Enhance safety and well-being for truck drivers and the public.

Main Methods:

• Administered a comprehensive questionnaire to 363 self-employed truck drivers in São Paulo, Brazil.
• Collected data on sociodemographics, health, sleep patterns, and working conditions.
• Utilized eight machine learning algorithms to predict crash likelihood.

Main Results:

• Drivers reported driving ~14.6 hours before fatigue and sleeping ~5.9 hours in 24 hours.
• Significant impact of loading/unloading wait times on working and rest hours.
• ML models achieved prediction accuracy rates between 78% and 85% for truck driver crashes.

Conclusions:

• Validated the development of accurate ML-derived models for predicting truck driver crash risk.
• Findings support policy development for improved truck driver safety and public health.
• Highlight the need to address working conditions and fatigue in the trucking industry.