Predicting driver drowsiness using vehicle measures: Recent insights and future challenges

Abstract

Introduction Driver drowsiness is a significant contributing factor to road crashes. One approach to tackling this issue is to develop technological countermeasures for detecting driver drowsiness, so that a driver can be warned before a crash occurs. Method The goal of this review is to assess, given the current state of knowledge, whether vehicle measures can be used to reliably predict drowsiness in real time. Results Several behavioral experiments have shown that drowsiness can have a serious impact on driving performance in controlled, experimental settings. However, most of those studies have investigated simple functions of performance (such as standard deviation of lane position) and results are often reported as averages across drivers, and across time. Conclusions Further research is necessary to examine more complex functions, as well as individual differences between drivers. Impact on Industry A successful countermeasure for predicting driver drowsiness will probably require the setting of multiple criteria, and the use of multiple measures.