From Attention to Distraction: Driver State Analysis From an Engineering Perspective – A Review

This paper reviews methods for defining, assessing, measuring, and testing driver distraction. It aims to exceed existing literature by integrating insights from interdisciplinary research fields, including engineering, psychology, medicine, and computer science. Using a systematic literature review approach, this review critically examines past and present studies, establishing a common ground and lexicon for the analysis. Distinct from prior reviews, which focused on segmented facets of distraction, this work comprehensively explores the inherent limits of attention. The core contribution of this paper lies in its holistic approach to assessing and measuring distraction. The assessment supplies fundamental distraction characteristics (e.g., category, risk) to guide the measuring process. Vehicular, behavioural, physiological, and subjective measurements types are considered. They are analysed, identifying their common trends across distraction categories and focusing on hybrid measurements (i.e., a combination of measurement types). Such hybrid measurements appear to offer a complete understanding of the distraction phenomenon. Furthermore, the paper discusses driver-specific features that affect measurements outcomes, such as age, experience, and cultural and geographical factors. This leads to innovative insights necessary for developing future tailored distraction mitigation strategies. Finally, the paper illuminates methodological similarities and differences between driving simulator tests and real driving field tests. This review advances the research field by providing a multidisciplinary foundation to define, assess, measure and test driver distraction. Future research paths should leverage the identified measurement trends across distraction categories to develop targeted driver monitoring systems.