Identifying the Causes of Drivers’ Hazardous States Using Driver Characteristics, Vehicle Kinematics, and Physiological Measurements

Darzi, Ali; Gaweesh, Sherif M.; Ahmed, Mohamed M.; Novak, Domen · 2018 · DOAJ

DOI: 10.3389/fnins.2018.00568

archive: archived pipeline: cataloged verified

Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)

Summary

This study addresses the critical gap in automated driver monitoring systems: while existing technologies can detect hazardous driver states (HDS) such as distraction or fatigue, they rarely identify the specific cause. Identifying the cause is essential for intelligent vehicles to tailor interventions appropriately. The researchers aimed to determine if specific causes of HDS—sleep deprivation, adverse weather, high traffic density, and cell phone use—could be automatically classified using a combination of driver characteristics, vehicle kinematics, and physiological measurements. The experimental design involved 21 healthy participants who completed four 45-minute simulated driving sessions. Two sessions induced mild sleep deprivation (participants slept less than 6 hours), while the other two represented alert states. Within each session, drivers navigated eight scenarios varying by weather (sunny vs. snowy), traffic density (low vs. high), and cell phone usage (present vs. absent). Data collection included three feature sets: driver characteristics (personality, stress, and mood via questionnaires), eight vehicle kinematics measures (e.g., throttle force, lane position, velocity), and four physiological signals (respiration, electrocardiogram, skin conductance, and body temperature). The researchers tested classification accuracy for each HDS cause using individual feature sets and all possible combinations thereof. The results demonstrated high classification accuracies for identifying the causes of hazardous states. Sleep deprivation was classified with the highest accuracy at 98.8%, followed by traffic density at 91.4%, cell phone use at 82.3%, and weather conditions at 71.5%. The analysis revealed that vehicle kinematics were the most effective features for classifying environmental factors like weather and traffic density. In contrast, physiological measures and driver characteristics were more useful for identifying intrinsic states such as sleep deprivation and cell phone distraction. Combining all three feature sets generally improved performance, though a secondary classification scheme that provided information about other concurrent HDS causes did not significantly increase accuracy. The significance of this work lies in its potential to enhance intelligent vehicle intervention systems. By accurately identifying not just the presence of a hazardous state but its specific cause, automated systems can provide targeted responses, such as suggesting a rest break for fatigue or disabling phone functions for distraction. The study confirms that a multimodal approach, integrating kinematic, physiological, and characteristic data, is superior to single-source monitoring. This framework supports the development of more nuanced and effective safety technologies that can adapt to the diverse factors contributing to driving impairment.

Provenance

The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed.

Summary generated by qwen3.6-27b-prismaquant on 2026-06-25; verification: verified.

Topics

Ranked by relevance to this paper. Hover a topic for its definition.

• drowsiness
• drowsiness detection algorithms
• sleep deprivation
• situational awareness
• stress driving
• distraction detection algorithms

Information type

What kind of knowledge this paper contributes, grouped by family — independent of topic (what it is about) and method (how it was studied).

• Empirical Findings: physiological data
• Theoretical Contribution: theory or model, conceptual framework