How does driver fatigue monitor system design affect carsharing drivers? An approach to the quantification of driver mental stress and visual attention

In the sharing transportation market, the mobility service quality and user experience are significantly influenced by the design level of in-vehicle electronic components. The acousto-optic information from the driver fatigue monitor system (DFMS), commonly installed in cars rented from carsharing platforms, causes frequent driver distraction and deteriorates the customer service. The current study proposed to quantify the mental stress of drivers when noticing DFMS, to determine the reasonability of driver visual attention allocation in unfamiliar shared cars. Firstly, innovatively using support vector machine (SVM) to reveal the extent to which the subject's heart rate variability (HRV) time-domain features, heart rate (HR) and eye-tracking data could predict the DFMS configuration type being observed by the subject. Through training, the finally determined model achieved relatively good classification accuracy of 86.957%. Then by means of entropy weight method, the importance of HRV, HR and eye movement indicators was sorted. The results indicated that in terms of the interpretation of psychological load in visual interaction with DFMS, HRV indicators were more important. Finally, comparisons were made between different DFMS configurations through two-way analysis of variance. The novel perspective applied in the study exposed that drivers were inclined to present lower psychological load in the case that a voice existed in the warning prompt. Furthermore, monitoring cameras installed independently on the instrument desk showed more advantages in reducing the subject's pressure than those integrated on the steering wheel. The data analytics in this evidence-based study provides a reference basis for the optimization of human–computer interaction in emerging carsharing services, and helps to make future mobility more user-friendly and societally beneficial.