Changes in physiological indices and deceleration behaviour as functions of braking demands and driver's physiological cluster. (April 2019)

• Record Type: Journal Article
• Authors: Musicant, Oren; Laufer, Ilan; Botzer, Assaf
• Is Part Of: Transportation research. Volume 62(2019)
• Page Start: 406
• Page End: 415
• Publication Date: 2019-04
• Subjects: Braking intensity -- Physiological indices -- Field test -- Cluster analysis; Automobile drivers -- Psychology -- Periodicals; Automobile driving -- Psychological aspects -- Periodicals; Transportation -- Psychological aspects -- Periodicals; 629.283019
• DOI: 10.1016/j.trf.2019.01.013
• Languages: English
• ISSNs: 1369-8478

Abstract

Highlights: Drivers can be clustered according to their physiological responses. Clusters mainly differed in heart rate and heart rate variability. Drivers may have different physiological responses but still perform the same way. Inclusion of clusters to prediction models improve prediction accuracy. Abstract: Physiological indices (e.g. heart rate and skin conductance) hold potential as monitoring tools for driver behaviour and mental workload. However, their effectiveness in this respect is presently limited by the large variability among individuals' physiological responses. The purpose of the present study was to examine whether drivers could be categorised into clusters according to their physiological responses to braking demands and whether the cluster type is related to deceleration intensity (in g units). Twenty-five drivers (males, aged 24–39) participated in a field experiment. We manipulated the braking demands using road signs to communicate the speed before braking ('50' or '60') and the target speed for braking ('30' or 'stop'). In an additional session, we asked drivers to brake intensively. We applied an advanced version of the k-means cluster analysis to the physiological data and adopted a two-physiological-cluster solution. Next, we calibrated an RM-ANOVA to estimate the effect of the physiological cluster and the braking demands on both the physiological indices and braking intensity. The results indicated that drivers could be assigned to one of the two clusters, mainly according to their heart rate and heart rate variability indices and less according to their skin conductance levels. However, the physiological cluster did not exert a significant effect on the actual deceleration intensity for any of the braking demands. These results indicate that different physiological responses to identical braking demands may result in similar performance outcomes. The practical implications of our observations might relate to the integration of physiological indices in driver support systems.