EUVIMEDEUVIMED entdecken
Crossref

Machine Learning and Sensor-Based Approaches for Driver Fatigue and Drowsiness Detection: A Mini Review

Kavin Ramakrishnan, Velmurugan S

Journal of Computer Allied Intelligence · 28.02.2026 · 4 (1) · 1-10

DOI: 10.69996/jcai.2026001 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Journal of Computer Allied Intelligence, online erschienen am 28.02.2026. Verfasst von Kavin Ramakrishnan und 1 weiteren Autor:innen.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Driver sleepiness is a significant contributing factor to road traffic accidents worldwide, resulting in substantial morbidity, mortality, and socioeconomic burden. Fatigue impairs critical driving functions, including reaction time, sustained attention, situational awareness, and decision-making capacity, thereby increasing the likelihood of errors in high-risk scenarios. Consequently, the development of reliable and accurate fatigue detection systems is essential for improving road safety. Traditional mitigation strategies—such as self-reported fatigue assessment, scheduled rest breaks, and public awareness campaigns—have demonstrated limited effectiveness due to variability in individual compliance and the subjective nature of fatigue measurement. In response, technology-driven approaches have gained increasing attention. Continuous driver monitoring through wearable sensors, multimodal sensing platforms, and advancements in artificial intelligence (AI) provides a promising pathway toward real-time fatigue detection and intervention. Current detection methodologies can be broadly categorized into behavioral, physiological, and vehicle-based approaches. Behavioral indicators—including eye closure, blink rate, facial expressions, and head movements—are typically derived from computer vision techniques and offer non-intrusive monitoring. Physiological measures, such as heart rate variability, electroencephalography (EEG), and electrocardiography (ECG), provide insights into neurophysiological and autonomic changes associated with drowsiness, although they may introduce practical and usability constraints. Vehicle-based metrics, including steering patterns, lane-keeping performance, and pedal input dynamics, offer indirect yet scalable indicators of driver impairment. Recent research has emphasized multimodal frameworks that integrate heterogeneous data sources to enhance detection accuracy under real-world driving conditions. By leveraging advanced AI and machine learning algorithms, these systems can model complex, nonlinear relationships between fatigue-related signals. Despite promising progress, challenges related to sensor reliability, user acceptance, data privacy, and real-time deployment remain significant. Nevertheless, ongoing advancements in intelligent monitoring technologies represent a safetyoriented approach with substantial potential to mitigate fatigue-related driving risks and improve transportation safety outcomes.

Zitieren

APA

Ramakrishnan, K., & S, V. (2026). Machine Learning and Sensor-Based Approaches for Driver Fatigue and Drowsiness Detection: A Mini Review. Journal of Computer Allied Intelligence, 4(1), 1-10. https://doi.org/10.69996/jcai.2026001

Vancouver

Ramakrishnan K, S V. Machine Learning and Sensor-Based Approaches for Driver Fatigue and Drowsiness Detection: A Mini Review. Journal of Computer Allied Intelligence. 2026;4(1):1-10. doi:10.69996/jcai.2026001

BibTeX
@article{Ramakrishnan2026,
  author = {Ramakrishnan, Kavin and S, Velmurugan},
  title = {Machine Learning and Sensor-Based Approaches for Driver Fatigue and Drowsiness Detection: A Mini Review},
  journal = {Journal of Computer Allied Intelligence},
  year = {2026},
  volume = {4},
  number = {1},
  pages = {1--10},
  doi = {10.69996/jcai.2026001}
}
Kostenlos starten →