FRODDO project new scientific paper explores gaze dynamics to reveal age-related physiological patterns across driving events
The latest FRODDO project research has been published, exploring how eye movements and pupil responses can contribute to a better understanding of driver behaviour and inform future driver monitoring technologies. As driver assistance systems become increasingly sophisticated, understanding human behaviour and the way drivers interact with changing road conditions remains a key challenge.
The research investigates how physiological signals can be analysed to identify driving events, capture patterns of visual attention and examine potential differences between younger and older drivers.
Twenty-seven participants completed a simulated driving scenario involving nine events while their physiological signals were recorded. The research examined different characteristics of visual behaviour, including the distribution of gaze, horizontal eye movement patterns and changes in pupil responses. Rather than relying exclusively on conventional measures, the researchers combined complementary analytical approaches to capture both the characteristics and the temporal evolution of these signals. By investigating how these signals evolve throughout different driving events, the research seeks to identify patterns that can help characterise driver behaviour. This approach highlights the potential of physiological data as a complementary source of information for understanding how drivers perceive and respond to specific situations.
Beyond analysing visual behaviour, the research explores how machine learning can be used to distinguish between driving events and how the resulting predictions can be interpreted. Understanding which characteristics of gaze behaviour contribute to a model’s decisions is an important consideration for the development of transparent and reliable driver monitoring technologies. Explainability techniques can help researchers better understand how models operate and assess their potential limitations. The study also investigates temporal patterns in physiological signals, offering a complementary perspective on how visual responses evolve over the course of a driving event.
Drivers do not all perceive and respond to their surroundings in the same way. Age and other individual characteristics can influence visual exploration and physiological responses, making these differences relevant to the design and evaluation of driver monitoring systems. By examining younger and older participants, the study explores how gaze and pupil patterns may vary across age groups. This perspective highlights the importance of considering individual variability when developing technologies intended to support drivers in different situations.
Read the full publication here
