Time for sleep science to wake up to drowsy driver monitoring

Driver monitoring systems (DMS) that can detect drowsiness and distraction in real-time are now included in the European new car assessment program and are already widely used in the road transport sector [1]. Rosekind et al. make a critical call for sleep scientists and clinicians to collaborate with the transport technology sector and contribute to guiding DMS development, evaluation, and use [2]. Drowsiness remains a key cause of road crashes. The Safe System approach to road safety acknowledges that humans are fallible. While it is important to minimize human error, it is not feasible to eliminate sleep loss, circadian and sleep disorders impacts on drowsiness and driver behavior. Vehicle innovations and road design have implemented numerous strategies that have reduced crash frequency and the risk of mortality and injury from all causes including drowsiness [3]. Rosekind et al. highlight the development of advanced driver assistance systems (e.g. lane departure warnings), automated driving systems that take over vehicle operations and DMS, which are all likely to be beneficial in reducing drowsy driving crash risk [2]. DMS specifically targets monitoring drivers for drowsiness and inattention to the roadway, or distraction. They offer great potential as an objective method to evaluate causes for driver drowsiness, the impact of preventive strategies and assess driving risk and treatment response in sleep disorders patients.

Technology designed to continuously detect and alert to drowsiness in real-time is established in the transport sector with rapidly expanding inclusion in passenger vehicles [4]. Most technologies are camera-based, and scan the driver's face to objectively detect signs of drowsiness such as eye closure [4]. Wearable technologies can detect drowsiness through EEG analysis, heart rate variability, or eye blink parameters [4]. Many technologies have demonstrated a high level of accuracy for detecting drowsiness in the laboratory, but there are few field-based studies [4, 5].

DMS are intended to provide a life-saving last line of defense by waking the driver during a falling asleep incident, or simply draw attention to a state of drowsiness the driver was unaware of, enabling the use of drowsiness countermeasures. While taking a break, nap, or caffeine usage can reduce drowsiness and crash risk [6, 7], it is unclear whether DMS alerts prompt drivers to utilize these measures. Anecdotally drivers have reported that DMS feedback improves their awareness of their own drowsiness state [8]. Limited studies have shown that implementing DMS with in-cabin alerts for drowsiness can reduce drowsiness events by over 60% and improve driving performance in occupational settings [9, 10]. The impact of in-cabin alerts appears to be enhanced by communicating the occurrence of driver drowsiness events to transport company supervisors in occupational settings. Further work is required to determine whether drowsiness in DMS use translates to reduced crashes and injuries.

At present, there are no clear standards for how individual drivers or companies should respond to individual or recurrent episodes of drowsiness. An important first step is to develop normative levels of DMS detected drowsiness, so that drivers or companies with drivers outside of this range can address potential underlying issues. Analysis of DMS data within companies can be used to assess the impact of different aspects of driver schedules on driver drowsiness, such as time of day, shift and break duration, and sequential shifts [11]. This could enable companies to identify the circumstances under which drowsiness is regularly occurring and adjust schedule features and implement countermeasures. Susceptibility to the impacts of sleep loss is variable between individuals [12]. Technology offers the potential to identify when individuals are more frequently impacted by regular drowsiness during transport operations, whether due to genetic susceptibility, sleep disorders, or other reasons. Supporting this concept, real-world driver monitoring studies show that most drowsiness events occur in a small proportion of drivers [13]. The European Union has included assessment of fatigue and distraction technology in safety ratings of new motor vehicles from 2023. Inevitably as this technology is broadly implemented within the passenger vehicle sector it will identify individuals with sleep disorders and regular drowsiness while driving. At present the sector is ill-prepared regards how to advise and manage drivers identified through what will essentially become a widespread screening process for driver drowsiness.

There are some limitations of drowsiness DMS. One hurdle to overcome is driver acceptance. While some DMS are fitted with anti-tampering equipment, they ultimately rely on driver adherence to respond to alerts. Some drivers have expressed concern with the discomfort of wearables or the privacy of camera-based technologies [8]. Hence, bringing drivers along on the journey of technology implementation is essential. Related to this is the perception of DMS accuracy. A naturalistic study of a prototype drowsiness detection device found that almost half of drivers occasionally ignored warnings, particularly when they felt they were inaccurate. Hence, evaluation of and continuous improvements to detection accuracy will be critical to DMS acceptance and subsequent efficacy. It is not yet clear whether drivers habituate to alerts over time and this must be evaluated in longitudinal studies so that long-term efficacy of DMS systems is understood.

The sleep disorders community has not yet embraced vehicle technology in clinical management of sleep disorders patients. Questionnaires to assess drowsiness in sleep disorder patients are modestly related to driving impairment and crash risk and are open to response bias [14]. Laboratory objective measures of sleepiness (maintenance of wakefulness test) are also moderately related to crash risk [15]. They assess a single time point, and are time-consuming and expensive, and hence are not regularly used to monitor patients. Preliminary studies suggest that drowsiness monitoring technology can distinguish which sleep apnea patients have regular episodes of drowsiness while driving [16]. There is much to do to determine if this provides a good indication of crash risk and develop normal criteria for different types of technology. As DMS become more widely available in passenger vehicles they will likely provide a great opportunity to assess fitness to drive and response to treatment in sleep disorders patients.

As highlighted by Rosekind et al., there is a need to develop and regulate robust methodologies for assessment of the accuracy and effectiveness of fatigue monitoring technologies. Standards exist for evaluating the impact of drugs on driving impairment, using instrumented vehicle technology to assess variation in driver steering, reflected in variability in lateral vehicle lane position [17]. Sleep loss, circadian rhythm, and sleep disorders' effects on drivers result in similar impacts on variation in vehicle lane position [18, 19]. This offers a pathway to develop standards for DMS, and determining their ability to detect and monitor drowsiness-related driving impairment. A high degree of accuracy has been demonstrated for detecting drowsiness-related out-of-lane excursions in some technologies during controlled instrumented vehicle studies [20], but accuracy in the field environment remains to be evaluated.

As uptake of DMS that can identify driver drowsiness becomes widespread sleep scientists and clinicians need to engage in the development of the science, regulation, and use of these novel technologies. They offer promise in better understanding drowsiness-related crashes, reducing crash risk in occupational settings and passenger vehicles and guiding management of sleep disorders patients.

Disclosure Statements

Financial disclosure: No funding was received for this work. The authors are employed through the Institute for Breathing and Sleep and Austin Health. MH and JC have previously received research grant support from the Co-operative Research Center for Alertness, Safety, and Productivity. Nonfinancial disclosure: MH and JC have previously received equipment support for research projects from Optalert and Seeing Machines.