The view from the road: the contribution of on-road glance-monitoring technologies to understanding driver behavior

R2DRV: study protocol for longitudinal assessment of driving after mild TBI in young drivers

BACKGROUND

Mild traumatic brain injury (mTBI) and traffic-related injuries are two major public health problems disproportionately affecting young people. Young drivers, whose driving skills are still developing, are particularly vulnerable to impaired driving due to brain injuries. Despite this, there is a paucity of research on how mTBI impacts driving and when it is safe to return to drive after an mTBI. This paper describes the protocol of the study, R2DRV, Longitudinal Assessment of Driving After Mild TBI in Young Drivers, which examines the trajectory of simulated driving performance and self-reported driving behaviors from acutely post-injury to symptom resolution among young drivers with mTBI compared to matched healthy drivers. Additionally, this study investigates the associations of acute post-injury neurocognitive function and cognitive load with driving among young drivers with and without mTBI.

METHODS

A total of 200 young drivers (ages 16 to 24) are enrolled from two study sites, including 100 (50 per site) with a physician-confirmed isolated mTBI, along with 100 (50 per site) healthy drivers without a history of TBI matched 1:1 for age, sex, driving experience, and athlete status. The study assesses primary driving outcomes using two approaches: (1) high-fidelity driving simulators to evaluate driving performance across four experimental study conditions at multiple time points (within 96 h of injury and weekly until symptom resolution or 8 weeks post-injury); (2) daily self-report surveys on real-world driving behaviors completed by all participants.

DISCUSSION

This study will fill critical knowledge gaps by longitudinally assessing driving performance and behaviors in young drivers with mTBI, as compared to matched healthy drivers, from acutely post-injury to symptom resolution. The research strategy enables evaluating how increased cognitive load may exacerbate the effects of mTBI on driving, and how post-mTBI neurocognitive deficits may impact the driving ability of young drivers. Findings will be shared through scientific conferences, peer-reviewed journals, and media outreach to care providers and the public.

Designing a Driver’s Hazard Perception Test Based on the Neural Brain Images Analysis (fMRI)

Background: Studies show that weakness in hazard perception is a major cause of traffic accidents, leading to high consequences. Objectives: This study aimed to design a valid and reliable driver’s Hazard Perception Test (HPT) based on neural imaging, reaction time, and miss rate in two groups of experienced and inexperienced drivers. Methods: Different roads, including urban, intercity, and rural, were filmed from drivers’ visual angles to examine the real road conditions. All videos were screened according to some quality factors. Then, hazard onset was determined for screened videos. The validity of the test was performed in three steps. Miss rates and reaction times to hazardous situations were measured. In the second step, 35 selected videos were broadcasted to 16 experienced and 16 novice drivers on a functional magnetic resonance imaging (fMRI). Finally, using 18 videos with statistically significant differences in neuro-cerebral neuronal activity, miss rate and reaction time were picked up for driver’s HPT. Results: The mean differences in reaction time, miss rate, and active neurons in the task of perceiving hazards in two groups of drivers were equal to 1.58 seconds, 29.55%, and 5248 neurons, respectively. There was a significant correlation between active neurons and miss rate (r = 0.556, P < 0.001). Eventually, the 18-videos of the valid test became HPT software. Conclusions: Application of this valid test is suggested for assessing the hazard perception of drivers, particularly those who are responsible for transporting staff and goods in the studied country.

Single feature spatio-temporal architecture for EEG Based cognitive load assessment

The study of electroencephalography (EEG) data for cognitive load analysis plays an important role in identification of stress-inducing tasks. This can be useful in applications such as optimal work allocation, increasing efficiency in the workplace and ensuring safety in difficult work environments. In order for such systems to be realistically deployable, easy acquisition and processing of the data on a wearable device is imperative. Current techniques primarily perform offline processing to analyse a multi-channel EEG to make a post facto assessment. This work focusses on building a new deep learning architecture that performs a single feature based spatio-temporal analysis of EEG data. This is achieved by creating a brain topographic map based on a single feature followed by spatio-temporal analysis using the developed network architecture. Data from two cognitive load experiments on the Physionet EEGMAT dataset were used to validate the performance. The network achieves an accuracy of 98.3% which is better than similar state-of-the-art approaches. Moreover, the proposed approach facilitates analysis of the spatial propagation of a signal, which is not possible through conventional EEG signal representations.

Putting Policy Into Practice: School-Level Compliance With and Implementation of State Concussion Laws

CONTEXT

Each year, approximately 2 million US children 18 years or younger sustain a concussion, a type of mild traumatic brain injury (TBI). Concussions can have detrimental effects on physical, cognitive, emotional, or sleep health.

POLICY

Between 2009 and 2014, all 50 US states and Washington, District of Columbia, enacted state concussion laws aimed to increase awareness about concussion and reduce the prevalence and severity of this injury. Most state laws include the following core tenets: (1) immediate removal from play after an actual or suspected concussion; (2) medical clearance before an athlete can return to play (RTP); and (3) concussion education for athletes, parents, and coaches.

IMPLEMENTATION

State concussion laws allow for substantial interpretation at the school level, resulting in considerable variation in the content of school written concussion policies and the level of implementation of state law requirements at the school level.

EVALUATION

We assessed the degree of high school written concussion policy compliance with the respective state law and examined the relationship between concussion policy compliance and school-level implementation of concussion laws. Seventy-one school officials completed a semistructured telephone interview and submitted their school's written concussion policy. Of the 71 policies analyzed, most complied with the removal-from-play, RTP, and concussion education tenets (90.1%, 97.2%, and 76.1%, respectively). The majority of participants reported that their school implemented the removal-from-play (91.5%), RTP (93.0%), and concussion education (80.6%) tenets well or very well. No significant relationships were found between researcher-rated school policy compliance and school-reported implementation of state law requirements at the school level.

DISCUSSION

Our findings suggest that most participating schools complied with their state concussion law and implemented law requirements well or very well. Future studies should identify facilitators and barriers to the implementation of state concussion laws at the school level.

Driving impairments and duration of distractions: Assessing crash risk by harnessing microscopic naturalistic driving data

Distracted and impaired driving is a key contributing factor in crashes, leading to about 35% of all transportation-related deaths in recent years. Along these lines, cognitive issues like inattentiveness can further increase the chances of crash involvement. Despite its prevalence and importance, little is known about how the duration of these distractions is associated with critical events, such as crashes or near-crashes. With new sensors and increasing computational resources, it is possible to monitor drivers, vehicle performance, and roadway features to extract useful information, e.g., eyes off the road, indicating distraction and inattention. Using high-resolution microscopic SHRP2 naturalistic driving data, this study conducts in-depth analysis of both impairments and distractions. The data has more than 2 million seconds of observations in 7394 baselines (no event), 1228 near-crashes, and 617 crashes. The event data was processed and linked with driver behavior and roadway factors. The intervals of distracted driving during the period of observation (15 seconds) were extracted; next, rigorous fixed and random parameter logistic regression models of crash/near-crash risk were estimated. The results reveal that alcohol and drug impairment is associated with a substantial increase in crash/near-crash event involvement of 34%, and the highest correlations with crash risk include duration of distraction through dialing on a cellphone, texting while driving, and reaching for an object. Using detailed pre-crash data from instrumented vehicles, the study contributes by quantifying crash risk vis-à-vis detailed driving impairment and information on secondary task involvement, and discusses the implications of the results.

The Role of Eye Tracking Technology in Assessing Older Driver Safety

A growing body of literature is focused on the use of eye tracking (ET) technology to understand the association between objective visual parameters and higher order brain processes such as cognition. One of the settings where this principle has found practical utility is in the area of driving safety.

METHODS

We reviewed the literature to identify the changes in ET parameters with older adults and neurodegenerative disease.

RESULTS

This narrative review provides a brief overview of oculomotor system anatomy and physiology, defines common eye movements and tracking variables that are typically studied, explains the most common methods of eye tracking measurements during driving in simulation and in naturalistic settings, and examines the association of impairment in ET parameters with advanced age and neurodegenerative disease.

CONCLUSION

ET technology is becoming less expensive, more portable, easier to use, and readily applicable in a variety of clinical settings. Older adults and especially those with neurodegenerative disease may have impairments in visual search parameters, placing them at risk for motor vehicle crashes. Advanced driver assessment systems are becoming more ubiquitous in newer cars and may significantly reduce crashes related to impaired visual search, distraction, and/or fatigue.

The uulmMAC Database-A Multimodal Affective Corpus for Affective Computing in Human-Computer Interaction

In this paper, we present a multimodal dataset for affective computing research acquired in a human-computer interaction (HCI) setting. An experimental mobile and interactive scenario was designed and implemented based on a gamified generic paradigm for the induction of dialog-based HCI relevant emotional and cognitive load states. It consists of six experimental sequences, inducing Interest, Overload, Normal, Easy, Underload, and Frustration. Each sequence is followed by subjective feedbacks to validate the induction, a respiration baseline to level off the physiological reactions, and a summary of results. Further, prior to the experiment, three questionnaires related to emotion regulation (ERQ), emotional control (TEIQue-SF), and personality traits (TIPI) were collected from each subject to evaluate the stability of the induction paradigm. Based on this HCI scenario, the University of Ulm Multimodal Affective Corpus (uulmMAC), consisting of two homogenous samples of 60 participants and 100 recording sessions was generated. We recorded 16 sensor modalities including 4 × video, 3 × audio, and 7 × biophysiological, depth, and pose streams. Further, additional labels and annotations were also collected. After recording, all data were post-processed and checked for technical and signal quality, resulting in the final uulmMAC dataset of 57 subjects and 95 recording sessions. The evaluation of the reported subjective feedbacks shows significant differences between the sequences, well consistent with the induced states, and the analysis of the questionnaires shows stable results. In summary, our uulmMAC database is a valuable contribution for the field of affective computing and multimodal data analysis: Acquired in a mobile interactive scenario close to real HCI, it consists of a large number of subjects and allows transtemporal investigations. Validated via subjective feedbacks and checked for quality issues, it can be used for affective computing and machine learning applications.

A Survey to Understand Emotional Situations on the Road and What They Mean for Affective Automotive UIs

In this paper, we present the results of an online survey (N = 170) on emotional situations on the road. In particular, we asked potential early adopters to remember a situation where they felt either an intense positive or negative emotion while driving. Our research is motivated by imminent disruptions in the automotive sector due to automated driving and the accompanying switch to selling driving experiences over horsepower. This creates a need to focus on the driver’s emotion when designing in-car interfaces. As a result of our research, we present a set of propositions for affective car interfaces based on real-life experiences. With our work we aim to support the design of affective car interfaces and give designers a foundation to build upon. We find respondents often connect positive emotions with enjoying their independence, while negative experiences are associated mostly with traffic behavior. Participants who experienced negative situations wished for better information management and a higher degree of automation. Drivers with positive emotions generally wanted to experience the situation more genuinely, for example, by switching to a “back-to-basic” mode. We explore these statements and discuss recommendations for the design of affective interfaces in future cars.

Glass half-full: On-road glance metrics differentiate crashes from near-crashes in the 100-Car data

BACKGROUND

Much of the driver distraction and inattention work to date has focused on concerns over drivers removing their eyes from the forward roadway to perform non-driving-related tasks, and its demonstrable link to safety consequences when these glances are timed at inopportune moments. This extensive literature has established, through the analyses of glance from naturalistic datasets, a clear relationship between eyes-off-road, lead vehicle closing kinematics, and near-crash/crash involvement.

OBJECTIVE

This paper looks at the role of driver expectation in influencing drivers' decisions about when and for how long to remove their eyes from the forward roadway in an analysis that consider the combined role of on- and off-road glances.

METHOD

Using glance data collected in the 100-Car Naturalistic Driving Study (NDS), near-crashes were examined separately from crashes to examine how momentary differences in glance allocation over the 25-s prior to a precipitating event can differentiate between these two distinct outcomes. Individual glance metrics of mean single glance duration (MSGD), total glance time (TGT), and glance count for off-road and on-road glance locations were analyzed. Output from the AttenD algorithm (Kircher and Ahlström, 2009) was also analyzed as a hybrid measure; in threading together on- and off-road glances over time, its output produces a pattern of glance behavior meaningful for examining attentional effects.

RESULTS

Individual glance metrics calculated at the epoch-level and binned by 10-s units of time across the available epoch lengths revealed that drivers in near-crashes have significantly longer on-road glances, and look less frequently between on- and off- road locations in the moments preceding a precipitating event as compared to crashes. During on-road glances, drivers in near-crashes were found to more frequently sample peripheral regions of the roadway than drivers in crashes. Output from the AttenD algorithm affirmed the cumulative net benefit of longer on-road glances and of improved attention management between on- and off-road locations.

CONCLUSION

The finding of longer on-road glances differentiating between safety-critical outcomes in the 100-Car NDS data underscores the importance of attention management in how drivers look both on and off the road. It is in the pattern of glances to and from the forward roadway that drivers obtained critical information necessary to inform their expectation of hazard potential to avoid a crash.

APPLICATION

This work may have important implications for attention management in the context of the increasing prevalence of in-vehicle demands as well as of vehicle automation.

Evaluation of the effectiveness of a multi-skill program for training younger drivers on higher cognitive skills

Training programs exist that prove effective at teaching novice drivers to anticipate latent hazards (RAPT), mitigate hazards (ACT) and maintain attention (FOCAL). The current study (a) measures the effectiveness of a novel integrated training program (SAFE-T) that takes only a third as long to complete compared to the three individual training programs and (b) determines if integrating the training of all the three higher cognitive skills would yield results comparable to the existing programs. Three groups were evaluated: SAFE-T, RAPT and Placebo. The results show that the drivers in the SAFE-T-trained group were more likely to anticipate hazards, quicker and more effective at responding to hazards, and more likely to maintain glance durations under a critical threshold of 2 s as compared to drivers in the Placebo-trained group who received a control program that does not actively train on any of the three cognitive skills. Moreover, the results show that the drivers in the SAFE-T trained group were just as likely to anticipate hazards as the drivers in the RAPT trained group. Finally, when compared with prior studies, the drivers in the SAFE-T trained group showed similar effects of attention maintenance training.

Assessing Cognitive Distraction in the Automobile

OBJECTIVE

The objective was to establish a systematic framework for measuring and understanding cognitive distraction in the automobile.

BACKGROUND

Driver distraction from secondary in-vehicle activities is increasingly recognized as a significant source of injuries and fatalities on the roadway.

METHOD

Across three studies, participants completed eight in-vehicle tasks commonly performed by the driver of an automobile. Primary, secondary, subjective, and physiological measures were collected and integrated into a cognitive distraction scale.

RESULTS

In-vehicle activities, such as listening to the radio or an audio book, were associated with a low level of cognitive workload; the conversation activities of talking to a passenger in the vehicle or conversing with a friend on a handheld or hands-free cell phone were associated with a moderate level of cognitive workload; and using a speech-to-text interfaced e-mail system involved a high level of cognitive workload.

CONCLUSION

The research established that there are significant impairments to driving that stem from the diversion of attention from the task of operating a motor vehicle and that the impairments to driving are directly related to the cognitive workload of these in-vehicle activities. Moreover, the adoption of voice-based systems in the vehicle may have unintended consequences that adversely affect traffic safety.

APPLICATION

These findings can be used to help inform scientifically based policies on driver distraction, particularly as they relate to cognitive distraction stemming from the diversion of attention to other concurrent activities in the vehicle.

On-road experiment to assess drivers' detection of roadside targets as a function of headlight system, target placement, and target reflectance

Adaptive headlights swivel with steering input to keep the beams on the roadway as drivers negotiate curves. To assess the effects of this feature on driver's visual performance, a field experiment was conducted at night on a rural, unlit, and unlined two-lane road during which 20 adult participant drivers searched a set of 60 targets. High- (n=30) and low- (n=30) reflectance targets were evenly distributed on straight road sections and on the inside or outside of curves. Participants completed three target detection trials: once with adaptive high-intensity discharge (HID) headlights, once with fixed HID headlights, and once with fixed halogen headlights. Results indicated the adaptive HID headlights helped drivers detect targets that were most difficult to see (low reflectance) at the points in curves found by other researchers to be most crucial for successful navigation (inside apex). For targets placed on straight stretches of road or on the outside of curves, the adaptive feature provided no significant improvement in target detection. However, the pattern of results indicate that HID lamps whether fixed or adaptive improved target detection somewhat, suggesting that part of the real world crash reduction measured for this adaptive system (Highway Loss Data Institute (HLDI), 2012a) may be due to the differences in the light source (HID vs. halogen). Depending on the scenario, the estimated benefits to driver response time associated with the tested adaptive (swiveling HID) headlights ranged from 200 to 380ms compared with the fixed headlight systems tested.

Computer vision and driver distraction: developing a behaviour-flagging protocol for naturalistic driving data

Naturalistic driving studies (NDS) allow researchers to discreetly observe everyday, real-world driving to better understand the risk factors that contribute to hazardous situations. In particular, NDS designs provide high ecological validity in the study of driver distraction. With increasing dataset sizes, current best practice of manually reviewing videos to classify the occurrence of driving behaviours, including those that are indicative of distraction, is becoming increasingly impractical. Current statistical solutions underutilise available data and create further epistemic problems. Similarly, technical solutions such as eye-tracking often require dedicated hardware that is not readily accessible or feasible to use. A computer vision solution based on open-source software was developed and tested to improve the accuracy and speed of processing NDS video data for the purpose of quantifying the occurrence of driver distraction. Using classifier cascades, manually-reviewed video data from a previously published NDS was reanalysed and used as a benchmark of current best practice for performance comparison. Two software coding systems were developed - one based on hierarchical clustering (HC), and one based on gender differences (MF). Compared to manual video coding, HC achieved 86 percent concordance, 55 percent reduction in processing time, and classified an additional 69 percent of target behaviour not previously identified through manual review. MF achieved 67 percent concordance, a 75 percent reduction in processing time, and classified an additional 35 percent of target behaviour not identified through manual review. The findings highlight the improvements in processing speed and correctly classifying target behaviours achievable through the use of custom developed computer vision solutions. Suggestions for improved system performance and wider implementation are discussed.

Adolescence, attention allocation, and driving safety

Motor vehicle crashes are the leading source of morbidity and mortality in adolescents in the United States and the developed world. Inadequate allocation of attention to the driving task and to driving hazards are important sources of adolescent crashes. We review major explanations for these attention failures with particular focus on the roles that brain immaturity and lack of driving experience play in causing attention problems. The review suggests that the potential for overcoming inexperience and immaturity with training to improve attention to both the driving task and hazards is substantial. Nevertheless, there are large individual differences in both attentional abilities and risky driving tendencies that pose challenges to novice driver policies. Research that can provide evidence-based direction for such policies is urgently needed.

Landscape heritage objects' effect on driving: a combined driving simulator and questionnaire study

According to the literature, landscape (panoramas, heritage objects e.g. landmarks) affects people in various ways. Data are primarily developed by asking people (interviews, photo sessions, focus groups) about their preferences, but to a lesser degree by measuring how the body reacts to such objects. Personal experience while driving a car through a landscape is even more rare. In this paper we study how different types of objects in the landscape affect drivers during their drive. A high-fidelity moving-base driving simulator was used to measure choice of speed and lateral position in combination with stress (heart rate measure) and eye tracking. The data were supplemented with questionnaires. Eighteen test drivers (8 men and 10 women) with a mean age of 37 were recruited. The test drivers were exposed to different new and old types of landscape objects such as 19th century church, wind turbine, 17th century milestone and bus stop, placed at different distances from the road driven. The findings are in some respect contradictory, but it was concluded that that 33% of the test drivers felt stressed during the drive. All test drivers said that they had felt calm at times during the drive but the reason for this was only to a minor degree connected with old and modern objects. The open landscape was experienced as conducive to acceleration. Most objects were, to a small degree, experienced (subjective data) as having a speed-reducing effect, much in line with the simulator data (objective data). Objects close to the road affected the drivers' choice of' lateral position. No significant differences could be observed concerning the test drivers' gaze between old or modern objects, but a significant difference was observed between the test drivers' gaze between road stretches with faraway objects and stretches without objects. No meaningful, significant differences were found for the drivers' stress levels as measured by heart rate.