Does heart rate variability monitoring accurately detect driver fatigue?
HRV-based driver fatigue detection systems show highly variable accuracy (44% to 100%) and require further standardized, real-world testing before deployment.
Driver fatigue detection systems have potential to improve road safety by preventing crashes and saving lives. Conventional driver monitoring systems based on driving performance and facial features may be challenged by the application of automated driving systems. This limitation could potentially be overcome by monitoring systems based on physiological measurements. Heart rate variability (HRV) is a physiological marker of interest for detecting driver fatigue that can be measured during real life driving. This systematic review investigates the relationship between HRV measures and driver fatigue, as well as the performance of HRV based fatigue detection systems. With the applied eligibility criteria, 18 articles were identified in this review. Inconsistent results can be found within the studies that investigated differences of HRV measures between alert and fatigued drivers. For studies that developed HRV based fatigue detection systems, the detection performance showed a large variation, where the detection accuracy ranged from 44% to 100%. The inconsistency and variation of the results can be caused by differences in several key aspects in the study designs. Progress in this field is needed to determine the relationship between HRV and different fatigue causal factors and its connection to driver performance. To be deployed, HRV-based fatigue detection systems need to be thoroughly tested in real life conditions with good coverage of relevant driving scenarios and a sufficient number of participants.
Lu et al. (Thu,) studied this question.