Driver drowsiness detection based on non-intrusive metrics considering individual specifics

ObjectivesDrowsy driving is a serious highway safety problem. If drivers could be warned before they became too drowsy to drive safely, some drowsiness-related crashes could be prevented. The presentation of timely warnings, however, depends on reliable detection. To date, the effectiveness of drowsiness detection methods has been limited by their failure to consider individual differences. The present study sought to develop a drowsiness detection model that accommodates the varying individual effects of drowsiness on driving performance.MethodsNineteen driving behavior variables and four eye feature variables were measured as participants drove a fixed road course in a high fidelity motion-based driving simulator after having worked an 8-h night shift. During the test, participants were asked to report their drowsiness level using the Karolinska Sleepiness Scale at the midpoint of each of the six rounds through the road course. A multilevel ordered logit (MOL) model, an ordered logit model, and an artificial neural network model were used to determine drowsiness.ResultsThe MOL had the highest drowsiness detection accuracy, which shows that consideration of individual differences improves the models’ ability to detect drowsiness. According to the results, percentage of eyelid closure, average pupil diameter, standard deviation of lateral position and steering wheel reversals was the most important of the 23 variables.ConclusionThe consideration of individual differences on a drowsiness detection model would increase the accuracy of the model's detection accuracy.     

Identifying cognitive distraction using steering wheel reversal rates

The influence of driver distraction on driving performance is not yet well understood, but it can have detrimental effects on road safety. In this study, we examined the effects of visual and non-visual distractions during driving, using a high-fidelity driving simulator. The visual task was presented either at an offset angle on an in-vehicle screen, or on the back of a moving lead vehicle. Similar to results from previous studies in this area, non-visual (cognitive) distraction resulted in improved lane keeping performance and increased gaze concentration towards the centre of the road, compared to baseline driving, and further examination of the steering control metrics indicated an increase in steering wheel reversal rates, steering wheel acceleration, and steering entropy. We show, for the first time, that when the visual task is presented centrally, drivers’ lane deviation reduces (similar to non-visual distraction), whilst measures of steering control, overall, indicated more steering activity, compared to baseline. When using a visual task that required the diversion of gaze to an in-vehicle display, but without a manual element, lane keeping performance was similar to baseline driving. Steering wheel reversal rates were found to adequately tease apart the effects of non-visual distraction (increase of 0.5° reversals) and visual distraction with offset gaze direction (increase of 2.5° reversals). These findings are discussed in terms of steering control during different types of in-vehicle distraction, and the possible role of manual interference by distracting secondary tasks.     

Monotony of road environment and driver fatigue: a simulator study

Studies have shown that drowsiness and hypovigilance frequently occur during highway driving and that they may have serious implications in terms of accident causation. This paper focuses on the task induced factors that are involved in the development of these phenomena. A driving simulator study was conducted in order to evaluate the impact of the monotony of roadside visual stimulation using a steering wheel movement (SWM) analysis procedure. Fifty-six male subjects each drove during two different 40-min periods. In one case, roadside visual stimuli were essentially repetitive and monotonous, while in the other one, the environment contained disparate visual elements aiming to disrupt monotony without changing road geometry. Subject's driving performance was compared across these conditions in order to determine whether disruptions of monotony can have a positive effect and help alleviate driver fatigue. Results reveal an early time-on-task effect on driving performance for both driving periods and more frequent large SWM when driving in the more monotonous road environment, which implies greater fatigue and vigilance decrements. Implications in terms of environmental countermeasures for driver fatigue are discussed.     

Lane heading difference: An innovative model for drowsy driving detection using retrospective analysis around curves

Driving while sleepy is a serious contributor to automobile accidents. Previous research has shown that drowsy drivers produce systematic errors (variability) in vehicle behavior which are detectable using vehicle monitoring technology. The current study developed a new methodological approach using a vehicle heading difference metric to detect drowsy driving more effectively than other more commonly used methods. Twenty participants completed a driving scenario as well as several measures of fatigue in five testing sessions across a night of sleep deprivation. Each simulated highway driving session lasted 20 min, and was analyzed for lateral lane position variability and vehicle heading difference variability with two statistical methods. Fatigue measures monitored reaction time, attention, and oculomotor movement. The results showed that examining lane heading difference using the absolute value of the raw data detected driving variability better across the night than other statistical models. The results from the fatigue measures indicated an increase in reaction time and response lapses, as well as a decrease in oculomotor reactivity across the night. These results suggest that in fatigued drivers the statistical model using the absolute value of lane heading could be an improved metric for drowsy driving detection that could accurately detect detriments in driving ability at lower levels of fatigue.     

Evaluation of experience-based fatigue countermeasures

Background and objective

Strategies adopted by drivers in order to cope with fatigue and falling asleep at the wheel include a variety of activities that may invigorate the body and/or the mind. The objectives of the current study were to examine the effectiveness of an energy drink and a non-traditional manual-dexterity/mastication activity as fatigue countermeasures.

Method

Twenty subjects participated in this driving simulator study. Each driving session lasted 2 h and each driver drove under three conditions: after consumption of an energy drink, while engaged in a self-paced manual-dexterity/mastication secondary task (shelling and eating sunflower seeds), and in a control condition with neither. Fatigue effects were assessed on three dimensions: subjective evaluations (using the Swedish Occupational Fatigue Inventory), a physiological indicator (heart rate variability), and driving performance measures (speed, steering, and lane deviations).

Results

The subjective and physiological measures showed a significant effect of both treatments in counteracting the effects of fatigue when compared to the control condition. The results of the driving performance measures indicated that the energy drink was effective in counteracting fatigue, while the secondary task was as effective as the energy drink in counteracting fatigue on measures that did not rely on hand movements.

Conclusions

Drinking an energy drink prior to the driving task has a significant, positive effect in counteracting fatigue, though it may have long-term negative rebound effects. The manual-dexterity/mastication secondary task can temporarily counteract the subjective and physiological effects of fatigue while driving, but can interfere with vehicle handling.     

Real-time driver drowsiness feedback improves driver alertness and self-reported driving performance

Driver drowsiness has been implicated as a major causal factor in road accidents. Tools that allow remote monitoring and management of driver fatigue are used in the mining and road transport industries. Increasing drivers’ own awareness of their drowsiness levels using such tools may also reduce risk of accidents. The study examined the effects of real-time blink-velocity-derived drowsiness feedback on driver performance and levels of alertness in a military setting. A sample of 15 Army Reserve personnel (1 female) aged 21–59 (M = 41.3, SD = 11.1) volunteered to being monitored by an infra-red oculography-based Optalert Alertness Monitoring System (OAMS) while they performed their regular driving tasks, including on-duty tasks and commuting to and from duty, for a continuous period of 4–8 weeks. For approximately half that period, blink-velocity-derived Johns Drowsiness Scale (JDS) scores were fed back to the driver in a counterbalanced repeated-measures design, resulting in a total of 419 driving periods under “feedback” and 385 periods under “no-feedback” condition. Overall, the provision of real-time feedback resulted in reduced drowsiness (lower JDS scores) and improved alertness and driving performance ratings. The effect was small and varied across the 24-h circadian cycle but it remained robust after controlling for time of day and driving task duration. Both the number of JDS peaks counted for each trip and their duration declined in the presence of drowsiness feedback, indicating a dynamic pattern that is consistent with a genuine, entropy-reducing feedback mechanism (as distinct from random re-alerting) behind the observed effect. Its mechanisms and practical utility have yet to be fully explored. Direct examination of the alternative, random re-alerting explanation of this feedback effect is an important step for future research.     

Feasibility of force platform based roadside drowsiness screening – A pilot study

Previous research on driver drowsiness detection has focused on developing in-car systems that continuously monitor the driver while driving and warn him/her when drowsiness compromises safety. In occupational settings a simple test of postural control has showed sensitivity to work shift induced fatigue in drivers. Whether the test is feasible for surveillance purposes in roadside settings is unknown. The present research sought to evaluate the feasibility of using a force platform test of postural control as a breathalyzer-like drowsiness-test at the roadside.     

Preventing drowsiness accidents by an alertness maintenance device

The reported experiment investigated in an advanced driving simulator whether drivers' alertness can be maintained in drowsiness-inducing conditions by a special game-like system, a 'gamebox'. Drowsiness was assessed by self-rating and eye-closures. Mental effort was assessed by a subjective workload rating scale and by a physiological measure (the 0.1 Hz component of heart rate). Driving quality and safety were assessed by steering movements, time-to-line crossings, and by the occurrence of safety-related driving errors--solid line crossings, driving off-road incidents and accidents. When driving with the gamebox, drivers reported a lower degree of drowsiness and fewer instances of sleep episodes as compared to a control condition. Driving with the device resulted in fewer incidents and accidents, and these occurred later in the session. The quality of vehicle control deteriorated progressively over the course of the session, but less so in the gamebox condition. Hence, the results corroborate the notion that mental activity counteracts drowsiness in prolonged driving.     

Research on Driver Fatigue Early Warning Method Based on ARM

Fatigue driving is one of the important reasons of road traffic accidents,fatigue driving is refers to the driver in a long time continuous driving or physical fatigue condition,and then come into being physiological and psychological function disorder. In order to overcome the limitation of single sensor in the fatigue test,aimed at the requirements of monitoring on the fatigue driving,this article designed an driver fatigue monitor system based ARM926EJ-S as a controller,it is used to determine the driver's fatigue and reduce the traffic acc ident. On the basis of fully considering the source correlation and complementary,it adopts the method of multi-source information fusion; by monitoring the pulse,heart rate,temperature of the human body,steering wheel grip strength to realized the fatigue level. The system of graphical interface adopts UCGUI. Finally,testing the main function modules of early warning system,the feasibility of the proposed early warning system is verified fusion.     

Determination of pre-impact occupant postures and analysis of consequences on injury outcome. Part I: a driving simulator study

This paper considers pre-impact vehicle maneuvers and analyzes the resulting driver motion from their comfort seating position. Part I of this work consists of analyzing the driver behavior during a crash. The study is conducted using the LAMIH driving simulator and involves 76 participants. The emergency situation is created by a truck emerging from behind a tractor on the opposite side of the road and tearing along the participant. The driver positioning throughout the simulation is recorded via five video cameras allowing view of the front scene, the driver face, feet and pedals, hands on the steering wheel and global lateral view. Data related to braking force, seat pressure, muscular activity for major groups of muscles and actions on the steering wheel are also collected.The typical response to this type of emergency event is to brace rearward into the seat and to straighten the arms against the steering wheel, or, to swerve to attempt to avoid the impacting vehicle. While turning the steering wheel, the forearm can be directly positioned on the airbag module at time of crash which represents a potential injurious situation.These positions are used in Part II to determine scenario of positions for numerical simulation of a frontal collision.     

Driver state examination—Treading new paths

A large proportion of crashes in road driving can be attributed to driver fatigue. Several types of fatigue are discussed, comprising sleep-related fatigue, active task-related fatigue (as a consequence of workload in demanding driving situations) as well as passive task-related fatigue (as related to monotonous driving situations). The present study investigated actual states of fatigue in a monotonous driving situation, using EEG measures and a long-lasting driving simulation experiment, in which drivers had to keep the vehicle on track by compensating crosswind of different strength. Performance data and electrophysiological correlates of mental fatigue (EEG Alpha and Theta power, Inter Trial Coherence (ITC), and auditory event-related potentials to short sound stimuli) were analyzed. Driving errors and driving lane variability increased with time on task and with increasing crosswind. The posterior Alpha and Theta power also increased with time on task, but decreased with stronger crosswind. The P3a to sound stimuli decreased with time on task when the crosswind was weak, but remained stable when the crosswind was strong. The analysis of ITC revealed less frontal Alpha and Theta band synchronization with time on task, but no effect of crosswind. The results suggest that Alpha power in monotonous driving situations reflects boredom or attentional withdrawal due to monotony rather than the decline of processing abilities as a consequence of high mental effort. A more valid indicator of declining mental resources with increasing time on task seems to be provided by brain oscillatory synchronization measures and event-related activity.     

Driving performance impairments due to hypovigilance on monotonous roads

Drivers’ ability to react to unpredictable events deteriorates when exposed to highly predictable and uneventful driving tasks. Highway design reduces the driving task mainly to a lane-keeping manoeuvre. Such a task is monotonous, providing little stimulation and this contributes to crashes due to inattention. Research has shown that driver’s hypovigilance can be assessed with EEG measurements and that driving performance is impaired during prolonged monotonous driving tasks. This paper aims to show that two dimensions of monotony – namely road design and road side variability – decrease vigilance and impair driving performance. This is the first study correlating hypovigilance and driver performance in varied monotonous conditions, particularly on a short time scale (a few seconds). We induced vigilance decrement as assessed with an EEG during a monotonous driving simulator experiment. Road monotony was varied through both road design and road side variability. The driver’s decrease in vigilance occurred due to both road design and road scenery monotony and almost independently of the driver’s sensation seeking level. Such impairment was also correlated to observable measurements from the driver, the car and the environment. During periods of hypovigilance, the driving performance impairment affected lane positioning, time to lane crossing, blink frequency, heart rate variability and non-specific electrodermal response rates. This work lays the foundation for the development of an in-vehicle device preventing hypovigilance crashes on monotonous roads.     

Driver behavior following an automatic steering intervention

The study investigated driver behavior toward an automatic steering intervention of a collision mitigation system. Forty participants were tested in a driving simulator and confronted with an inevitable collision. They performed a naïve drive and afterwards a repeated exposure in which they were told to hold the steering wheel loosely. In a third drive they experienced a false alarm situation. Data on driving behavior, i.e. steering and braking behavior as well as subjective data was assessed in the scenarios. Results showed that most participants held on to the steering wheel strongly or counter-steered during the system intervention during the first encounter. Moreover, subjective data collected after the first drive showed that the majority of drivers was not aware of the system intervention. Data from the repeated drive in which participants were instructed to hold the steering wheel loosely, led to significantly more participants holding the steering wheel loosely and thus complying with the instruction. This study seems to imply that without knowledge and information of the system about an upcoming intervention, the most prevalent driving behavior is a strong reaction with the steering wheel similar to an automatic steering reflex which decreases the system's effectiveness. Results of the second drive show some potential for countermeasures, such as informing drivers shortly before a system intervention in order to prevent inhibiting reactions.     

Driving simulator sickness: Impact on driving performance,influence of blood alcohol concentration,and effect of repeated simulator exposures

Simulator sickness is a major obstacle to the use of driving simulators for research, training and driver assessment purposes. The purpose of the present study was to investigate the possible influence of simulator sickness on driving performance measures such as standard deviation of lateral position (SDLP), and the effect of alcohol or repeated simulator exposure on the degree of simulator sickness. Twenty healthy male volunteers underwent three simulated driving trials of 1 h’s duration with a curvy rural road scenario, and rated their degree of simulator sickness after each trial. Subjects drove sober and with blood alcohol concentrations (BAC) of approx. 0.5 g/L and 0.9 g/L in a randomized order. Simulator sickness score (SSS) did not influence the primary outcome measure SDLP. Higher SSS significantly predicted lower average speed and frequency of steering wheel reversals. These effects seemed to be mitigated by alcohol. Higher BAC significantly predicted lower SSS, suggesting that alcohol inebriation alleviates simulator sickness. The negative relation between the number of previous exposures to the simulator and SSS was not statistically significant, but is consistent with habituation to the sickness-inducing effects, as shown in other studies. Overall, the results suggest no influence of simulator sickness on SDLP or several other driving performance measures. However, simulator sickness seems to cause test subjects to drive more carefully, with lower average speed and fewer steering wheel reversals, hampering the interpretation of these outcomes as measures of driving impairment and safety. BAC and repeated simulator exposures may act as confounding variables by influencing the degree of simulator sickness in experimental studies.     

驾驶疲劳的心电和肌电信号分析

目的探讨疲劳情境下驾驶员心电信号和肌电信号的有效特征。方法被试为12名年轻驾驶员,标准刺激为96张不同驾驶情景图片,要求被试连续2.5 h对图片进行按键反应。结果随着疲劳程度的增强,心率变异性呈整体上升趋势,平均功率频率呈下降趋势。结论心率、表面肌电值可以作为测量驾驶员疲劳程度的有效指标,人体的疲劳程度可能不是连续的增加,而是疲劳和清醒交替不断出现。     

Driving performance impairments due to hypovigilance on monotonous roads

Drivers' ability to react to unpredictable events deteriorates when exposed to highly predictable and uneventful driving tasks. Highway design reduces the driving task mainly to a lane-keeping manoeuvre. Such a task is monotonous, providing little stimulation and this contributes to crashes due to inattention. Research has shown that driver's hypovigilance can be assessed with EEG measurements and that driving performance is impaired during prolonged monotonous driving tasks. This paper aims to show that two dimensions of monotony - namely road design and road side variability - decrease vigilance and impair driving performance. This is the first study correlating hypovigilance and driver performance in varied monotonous conditions, particularly on a short time scale (a few seconds). We induced vigilance decrement as assessed with an EEG during a monotonous driving simulator experiment. Road monotony was varied through both road design and road side variability. The driver's decrease in vigilance occurred due to both road design and road scenery monotony and almost independently of the driver's sensation seeking level. Such impairment was also correlated to observable measurements from the driver, the car and the environment. During periods of hypovigilance, the driving performance impairment affected lane positioning, time to lane crossing, blink frequency, heart rate variability and non-specific electrodermal response rates. This work lays the foundation for the development of an in-vehicle device preventing hypovigilance crashes on monotonous roads.     

Reliability of simulator driving tool for evaluation of sleepiness,fatigue and driving performance

Study objective

To compare the impact of extended wakefulness (i.e., sleepiness) and prolonged driving (i.e., fatigue) at the wheel in simulated versus real-life driving conditions.

Design

Participants drove on an INRETS-MSIS SIM2 simulator in a research laboratory or an open French highway during 3 nocturnal driving sessions. A dose–response design of duration of nocturnal driving was used: a 2 h short driving session (3–5 AM), a 4 h intermediate driving session (1–5 AM) and an 8 h long driving session (9 PM–5 AM).

Participants

Two groups of healthy male drivers (20 for simulated driving and 14 drivers for real driving; mean age ± SD = 22.3 ± 1.6 years), free of sleep disorders.

Measurements

Number of inappropriate line crossings, self-rated fatigue and sleepiness were recorded in the last hour of driving sessions to control the effects of prior waking time and time of day.

Results

Compared to the daytime reference session, both simulated and real driving performance were affected by a short nocturnal driving session (P < .05 and P < .001, respectively). Extension of nocturnal driving duration affected simulated performance nonlinearly and more severely than that of real driving (P < .001).Compared to the daytime reference session, short nocturnal simulated and real driving sessions increased self-perceived fatigue and sleepiness. Real and simulated driving conditions had an identical impact on fatigue and sleepiness during extended periods of nocturnal driving.

Conclusions

In healthy subjects, the INRETS-MSIS SIM2 simulator appropriately measures driving impairment in terms of inappropriate line crossings related to extended wakefulness but has limitations to measure the impact of extended driving on drivers’ performance.     

The influence of daily sleep patterns of commercial truck drivers on driving performance

Fatigued and drowsy driving has been found to be a major cause of truck crashes. Lack of sleep is the number one cause of fatigue and drowsiness. However, there are limited data on the sleep patterns (sleep duration, sleep percentage in the duration of non-work period, and the time when sleep occurred) of truck drivers in non-work periods and the impact on driving performance. This paper examined sleep patterns of 96 commercial truck drivers during non-work periods and evaluated the influence these sleep patterns had on truck driving performance. Data were from the Naturalistic Truck Driving Study. Each driver participated in the study for approximately four weeks. A shift was defined as a non-work period followed by a work period. A total of 1397 shifts were identified. Four distinct sleep patterns were identified based on sleep duration, sleep start/end point in a non-work period, and the percentage of sleep with reference to the duration of non-work period. Driving performance was measured by safety-critical events, which included crashes, near-crashes, crash-relevant conflicts, and unintentional lane deviations. Negative binomial regression was used to evaluate the association between the sleep patterns and driving performance, adjusted for driver demographic information. The results showed that the sleep pattern with the highest safety-critical event rate was associated with shorter sleep, sleep in the early stage of a non-work period, and less sleep between 1 a.m. and 5 a.m. This study also found that male drivers, with fewer years of commercial vehicle driving experience and higher body mass index, were associated with deteriorated driving performance and increased driving risk. The results of this study could inform hours-of-service policy-making and benefit safety management in the trucking industry.     

Development of models for on-board detection of driver impairment

Two of the leading causes of automobile accidents are driver impairment due to alcohol and drowsiness. Apparently, a relatively large percentage of these accidents occur because drivers are unaware of the degree to which they are impaired. The purpose of this research was to develop models, utilizing changes in driver behavior, which could detect driver impairment due to alcohol, drowsiness, or the combination of alcohol and drowsiness, and which could be practically implemented in an automobile. A computer-controlled automobile simulator was used to simulate a nighttime highway driving scenario for six drivers who participated in each of four conditions: a control condition, an alcohol condition, a sleep-deprived condition, and a combined alcohol and sleep-deprived condition. The results indicated that a useful on-board drowsiness detection device is possible and practical for highway driving. The results also showed that on-board alcohol impairment detection may be possible at levels below the legal driving limit in most states (BAC 0.1%).     

The effect of a low and high resource intervention on older drivers’ knowledge,behaviors and risky driving

This study aimed to compare an in-class Seniors on the MOVE (Mature Operators Vehicular Education) interactive multi-session driving curriculum with a self-guided MOVE curriculum for older adults. Using a two group randomized design, we sought to determine if there are between-group differences in older drivers’ knowledge and safety behaviors among participants. Forty-four participants with an average age of 79 years (SD = 7.1) were randomly assigned to the original MOVE program (SOM-A) or a lower resource (SOM-B) self-guided intervention. SOM-A is a four session program designed to improve older drivers safety knowledge and better understand skills for safer driving. SOM-B is a self-guided program with one required in-class session and one optional session. Subsequent to completion of both curricula, participants were offered CarFit, a comprehensive check of how well a senior driver and their vehicle work together. Baseline, post-intervention and 6-month follow up questionnaires were completed by participants. We found significant differences (p = .01) in the mean driving safety knowledge scores when comparing participants in SOM-A (3.7, SD 2.0) to those in SOM-B (0.87, SD 2.6). With regard to behavioral outcomes, we focused on always wearing a seatbelt, talking with a health care provider about driving ability, and sitting 10–12 inches from the steering wheel. The vast majority of participants reported always wearing their seat belts (SOM-A 100%, SOM-B 92%, p = 1.0), and very few reported talking with their doctors (SOM-A Baseline – 0%, Follow up 1 – 0%, p = n/a). Mean behavior change scores for participants sitting 10–12 inches from the steering wheel were significantly more likely among SOM-A (mean = .65, SD = .5) participants than those in SOM-B (mean = .29, SD = .5, p = .01) at first follow-up. Taken together, these findings suggest that the more intensive program is more effective and that driving safety programs focused on behaviors to self evaluate driving abilities continue to be needed to help older drivers remain safer on the road as they age. The involvement of health care providers in such efforts may be an untapped potential.