Novice driver crashes: The relation between putative causal factors, countermeasures, real world implementations, and policy - A case study in simple, scalable solutions.

Novice drivers are at a greatly inflated risk of crashing. This led in the 20th century to numerous attempts to develop training programs that could reduce their crash risk. Yet, none proved effective. Novice drivers were largely considered careless, not clueless. This article is a case study in the United States of how a better understanding of the causes of novice driver crashes led to training countermeasures targeting teen driving behaviors with known associations with crashes. These effects on behaviors were large enough and long-lasting enough to convince insurance companies to develop training programs that they offered around the country to teen drivers. The success of the training programs at reducing the frequency of behaviors linked to crashes also led to several large-scale evaluations of the effect of the training programs on actual crashes. A reduction in crashes was observed. The cumulative effect has now led to state driver licensing agencies considering as a matter of policy both to include items testing the behaviors linked to crashes on licensing exams and to require training on safety critical behaviors. The effort has been ongoing for over a quarter century and is continuing. The case study highlights the critical elements that made it possible to move from a paradigm shift in the understanding of crash causes to the development and evaluation of crash countermeasures, to the implementation of those crash countermeasures, and to subsequent policy changes at the state and federal level. Key among these elements is the development of simple, scalable solutions.

Booster Dose of Attention Training Program for Young Novice Drivers: A Longitudinal Driving Simulator Evaluation Study.

OBJECTIVE: We examined the effectiveness of a second exposure to ACCEL, a novel driving training program, on latent hazard anticipation (HA) performance several months after their first exposure. BACKGROUND: Past research has demonstrated that PC-based driver training programs can improve latent HA performance in young novice drivers, but these improvements are below the ceiling level. METHOD: Twenty-five participants were randomly assigned to either the Placebo group, the ACCEL-1 group, or the ACCEL-2 group. Following the completion of the assigned training program, participants drove a series of eighteen scenarios incorporating latent hazards in a high-fidelity driving simulator with their eyes tracked. Participants returned two to six months following the first session and completed either the placebo program (ACCEL-1 and Placebo groups), or a second dose of training program (ACCEL-2 group), again followed by simulated evaluation drives. RESULTS: The ACCEL-2 group showed improved HA performance compared to the ACCEL-1 and Placebo groups in the second evaluation. CONCLUSION: ACCEL enhances young novice drivers' latent HA performance. The effectiveness of ACCEL is retained up to 6 months, and a second dose further improves HA performance. APPLICATION: Policy makers should consider requiring such training before the completion of graduate driver license programs. Young novice drivers that do not show successful latent HA performance could be required to complete additional training before being allowed to drive without restrictions.

Using SHRP2 NDS data to examine infrastructure and other factors contributing to older driver crashes during left turns at signalized intersections.

Drivers age 65 and over have higher rates of crashes and crash-related fatalities than other adult drivers and are especially over-represented in crashes during left turns at intersections. This research investigated the use of SHRP2 Naturalistic Driving Study (NDS) data to assess infrastructure and other factors contributing to left turn crashes at signalized intersections, and how to improve older driver safety during such turns. NDS data for trips involving signalized intersections and crash or near-crash events were obtained for two driver age groups: drivers age 65 and over (older drivers) and a sample of drivers age 30-49, along with NDS pre-screening and questionnaire data. Video scoring of all trips was performed to collect additional information on intersection and trip conditions. To identify the most influential factors of crash risk during left turns at signalized intersections, machine learning and regression models were used. The results found that in the obtained NDS dataset, there was a relatively small volume of crashes during left turns at signalized intersections. Further, model results found the statistically significant variables of crash risk for older drivers were associated more with health and cognitive factors rather than the infrastructure or design of the intersections. The results suggest that a study using only SHRP2 NDS data will not lead to definitive findings or recommendations for infrastructure changes to increase safety for older drivers at signalized intersections and during left turns. Moreover, the findings of this study indicates the need to consider other data sources and data collection methods to address this critical literature gap in older driver safety.

Real-time prediction and avoidance of secondary crashes under unexpected traffic congestion.

According to the Federal Highway Administration, nonrecurring congestion contributes to nearly half of the overall congestion. Temporal disruptions impact the effective use of the complete roadway, due to speed reduction and rubbernecking resulting from primary incidents that in turn provoke secondary incidents. There is an additional reduction of discharge flow caused by secondary incident that significantly increases total delay. Therefore, it is important to sequentially predict the probability of secondary incidents and develop appropriate countermeasures to reduce the associated risk. Advanced computing techniques were used to easily understand and reliably predict secondary incident occurrences that have low sample mean and a small sample size. The likelihood of a secondary incident was sequentially predicted from the point of incident response to the eventual road clearance. The quality of predictions improved with the availability of additional information. The prediction performance of the principled Bayesian learning approach to neural networks (bnn) was compared to the Stochastic Gradient Boosted Decision Trees (gbdt). A pedagogical rule extraction approach, trepan, which extracts comprehensible rules from the neural networks, improved the ability to understand secondary incidents in a simplified manner. With an acceptable accuracy, gbdt is a useful tool that presents the relative importance of the predictor variables. Unexpected traffic congestion incurred by an incident is a dominant causative factor for the occurrence of secondary incidents at different stages of incident clearance. This symbolic description represents a series of decisions that may assist emergency operators by improving their decision-making capabilities. Analyzing causes and effects of traffic incidents helps traffic operators develop incident-specific strategic plans for prompt emergency response and clearance. Application of the model in connected vehicle environments will help drivers receive proactive corrective feedback before a crash. The proposed methodology can be used to alert drivers about potential highway conditions and may increase the drivers' awareness of potential events when no rerouting is possible, optimal or otherwise.

Effectiveness of visual warnings on young drivers hazard anticipation and hazard mitigation abilities.

Previous studies have demonstrated that young drivers fail both to scan for and mitigate latent hazards mostly due to their cluelessness. This study aims to investigate whether these skills could be improved by providing young drivers with alerts in advance of the upcoming threat using a driving simulator experiment. In particular, the warning was presented on the head-up displays (HUD) either 2 s, 3 s or 4 s in advance of a latent threat. The hazard anticipation, hazard mitigation and attention maintenance performance of forty-eight young drivers aged 18-25 was evaluated across eight unique scenarios either in the presence or in the absence of latent threat alerts displayed on a HUD. There were four groups overall: one control group (no alert) and three experimental groups (2 s alert, 3 s alert and 4 s alert). The analysis of the hazard anticipation data showed that all three experimental groups with HUD warnings (2 s, 3 s, 4 s) significantly increased the likelihood that drivers would glance towards latent pedestrian and vehicle hazards when compared to the control group. The hazard mitigation analysis showed that in situations involving a pedestrian threat, HUD alerts provided 3 or 4 s in advance of a potential threat led drivers to travel significantly slower than the control group or the 2 s group. No significant effect of a HUD alert on drivers' speed was found when the latent hazard was a vehicle. An analysis of eye behaviors showed that only 7 out of 597 glances at the HUD were longer than 2 s safety-threshold, indicating that the warnings do not seem to distract the driver.

Effective cues for accelerating young drivers' time to transfer control following a period of conditional automation.

OBJECTIVE: During conditional automated driving, a transition from the automated driving suite to manual control requires the driver to take over control at a moment's notice. Thus, it is critical that a driver be made situationally aware as quickly as possible in those conditions where he or she may not be paying full attention. Recent research suggests that specific cues about upcoming hazards (e.g., "crosswalk ahead") can increase the drivers' situation awareness during these safety-critical take-over situations when compared with a general cue ("take over control"). The current study examines whether this increased situation awareness which occurs as a result of more specific cues translates into improved hazard mitigation performance within the same limited time window. METHOD: Fifty-seven drivers were randomly assigned to one of five between-subjects conditions (one control condition and four experimental auditory cue conditions) that varied in the specificity of information provided about an upcoming hazard. The four experimental conditions included a period of conditional automated driving where the driver was engaged in a driving-irrelevant task and looked away from the forward roadway prior to a take-over request. Drivers in the fifth condition had no cue and drove manually throughout. The same six simulator scenarios were used in all five conditions to evaluate how well the driver mitigated a hazard. The average velocity, standard deviation of velocity, and average absolute acceleration were recorded along with the glance behaviors of drivers. RESULTS: In general, during the 4s prior to a latent hazard (following the alerting cues in the automated driving conditions), the more likely a driver was to glance towards a latent hazard, the more likely the driver was to reduce his or her speed. Moreover, analyses focusing solely on hazard mitigation behavior revealed patterns that mirrored the glance behavior results. Specifically, drivers that were presented with cues that described the environments in which hazards were likely to occur were more likely to demonstrate vehicle behaviors that were consistent with speed reductions (lower velocity, higher speed variability, and higher absolute acceleration) than were drivers who were presented general cues or cues about the identity of the upcoming hazards. CONCLUSION: Even in as little as 4s prior to a potential hazard, cues that inform the driver of the environment in which the hazard is likely to occur increase the likelihood that the driver mitigates the crash compared with drivers who are provided general information or threat identity information.

The use of a driving simulator to determine how time pressures impact driver aggressiveness.

Speeding greatly attributes to traffic safety with approximately a third of fatal crashes in the United States being speeding-related. Previous research has identified being late as a primary cause of speeding. In this driving simulator study, a virtual drive was constructed to evaluate how time pressures, or hurried driving, affected driver speed choice and driver behavior. In particular, acceleration profiles, gap acceptance, willingness to pass, and dilemma zone behavior were used, in addition to speed, as measures to evaluate whether being late increased risky and aggressive driving behaviors. Thirty-six drivers were recruited with an equal male/female split and a broad distribution of ages. Financial incentives and completion time goals calibrated from a control group were used to generate a Hurried and Very Hurried experimental group. As compared to the control group, Very Hurried drivers selected higher speeds, accelerated faster after red lights, accepted smaller gaps on left turns, were more likely to pass a slow vehicle, and were more likely to run a yellow light in a dilemma zone situation. These trends were statistically significant and were also evident with the Hurried group but a larger sample would be needed to show statistical significance. The findings from this study provide evidence that hurried drivers select higher speeds and exhibit riskier driving behaviors. These conclusive results have possible implications in areas such as transportation funding and commercial motor vehicle safety.

Large reductions are possible in older driver crashes at intersections.

Among all crash types, the largest percentage of older driver fatalities occur at intersections. Many explanations have been offered for older drivers' increased risks of crashing at intersections; however, only recently was it determined that older drivers were much less likely to glance for latent threats after entering an intersection than middle-aged drivers. In response, training programmes were designed to increase the frequency of such glances. The programmes have proven effective, doubling the frequency of these glances for up to a period of two years post-training. The programmes take only an hour to administer and are not directly targeted at remediating any of the underlying declines in cognitive, visual or motor function that can explain the decrease in the frequency of glances for threat vehicles among older drivers. The first question we addressed was, what are the basic declines that can explain the decrease in glances for threat vehicles? The second question we addressed was, how did the training programme achieve the results it did without directly addressing these declines? We hypothesise that drivers are learning to decouple hand, foot and head movements in the training programmes and that this serialisation of behaviour essentially sidesteps the major declines in cognitive, visual and motor functions. We provide evidence that the assumptions of the decoupling hypothesis about the capabilities of older drivers when the movements are decoupled, are consistent with the evidence from existing experiments. More research is needed to evaluate this hypothesis.

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.

TEXTING WHILE DRIVING: EVALUATION OF GLANCE DISTRIBUTIONS FOR FREQUENT/INFREQUENT TEXTERS AND KEYPAD/TOUCHPAD TEXTERS.

The threat that cell-phones pose to driving has been a well researched topic. There are fewer studies of the threat that texting creates for drivers, but the risks are obvious and the few existing studies confirm this. What is not obvious is whether frequent texters will expose themselves to the same risks as infrequent texters. This is important to know because many texters, especially teens who text frequently, may consider themselves immune to the dangers of texting while driving. As such, a comparison of frequent and infrequent texters was undertaken on a driving simulator. It is also not immediately clear what effects the different types of interfaces have on driving performance while text messaging. The interfaces under evaluation included keypad or "qwerty" phones (e.g., Blackberries) and touchpad phones (iPhone). It was found that the frequent and infrequent texters were equally likely to glance at least once for more than 2s inside the vehicle while sending a text message. It was also found that touchpad texters had a larger number of glances above the 2s threshold than keypad users, though this difference was not significant. The implications of this for future public policy are discussed.

Long Term Effects of Hazard Anticipation Training on Novice Drivers Measured on the Open Road.

(a) The purpose of this study was to determine whether novice drivers that were trained to anticipate hazards did so better than novice drivers who were not so trained immediately after training and up to one year after training occurred. (b) Novice drivers who had held their restricted license for about one month were randomly assigned to a PC-based hazard anticipation training program (RAPT) or a placebo (control) training program. The programs took about one hour to complete. The effects of training were assessed in a field drive by using patterns of eye movements to assess whether drivers anticipated a potential unseen hazard. (c) The effects of training persisted over time. In the field test immediately after training, the RAPT group anticipated the hazards 65.8% of the time whereas; the control group anticipated them only 47.3% of the time. Six or more months later, the groups were brought back for a second field test and the effects of training did not diminish; the RAPT group anticipated the hazards 61.9% of the time compared to 37.7% for the control group.

Do Advance Yield Markings Increase Safe Driver Behaviors at Unsignalized, Marked Midblock Crosswalks? Driving Simulator Study.

In the United States, 78% of pedestrian crashes occur at noninter-section crossings. As a result, unsignalized, marked midblock crosswalks are prime targets for remediation. Many of these crashes occur under sight-limited conditions in which the view of critical information by the driver or pedestrian is obstructed by a vehicle stopped in an adjacent travel or parking lane on the near side of the crosswalk. Study of such a situation on the open road is much too risky, but study of the situation in a driving simulator is not. This paper describes the development of scenarios with sight limitations to compare potential vehicle-pedestrian conflicts on a driving simulator under conditions with two different types of pavement markings. Under the first condition, advance yield markings and symbol signs (prompts) that indicated "yield here to pedestrians" were used to warn drivers of pedestrians at marked, midblock crosswalks. Under the second condition, standard crosswalk treatments and prompts were used to warn drivers of these hazards. Actual crashes as well as the drivers' point of gaze were measured to determine if the drivers approaching a marked midblock crosswalk looked for pedestrians in the crosswalk more frequently and sooner in high-risk scenarios when advance yield markings and prompts were present than when standard markings and prompts were used. Fewer crashes were found to occur with advance yield markings. Drivers were also found to look for pedestrians much more frequently and much sooner with advance yield markings. The advantages and limitations of the use of driving simulation to study problems such as these are discussed.

Dynamic Cellular Adhesion Mediated by Copolymeric Nanofilm Substrates.

Amphiphilic block copolymers are finding increased potential in biological and medical research due to their innate alternating hydrophilic and hydrophilic blocks/segments which can be used to package therapeutics, or coat a broad array of biological interfaces. Some studies are already directed towards utilizing these copolymers' ability to form micelles or vesicles to develop novel methods of drug delivery to prevent inflammation or pro-cancer activity. Our study, however, aims to investigate the more fundamental cell-block copolymer interaction for use in protective nanofilms to prevent bio-fouling of non-tissue based implantable devices. Block copolymers could potentially fill the demand for biologically inert, highly functionalizable biomaterials desirable for this type of application. Two such polymers used in our study include PMOXA-PDMS-PMOXA triblock copolymer and PEO/PMMA diblock copolymer. Each block copolymer possesses hydrophilic and hydrophobic blocks that enable it to mimic the cell lipid membrane. So far we have shown that triblock copolymer is capable of inhibiting the accumulation of murine macrophages onto glass substrates. Preliminary evidence has suggested that the triblock copolymer has anti-adsorptive as well as non-inflammatory capabilities during short incubation periods (7 days) in vitro. While the diblock copolymer displays minimal anti-adsorptive activities, nanofilms comprised of a mixture of the two copolymers were able to significantly reduce macrophage accumulation onto glass substrates. The disparate behavior seen by macrophages on the different materials may be due to specific inherent properties such as preference for hydrophobic vs. hydrophilic surfaces and/or rough vs. smooth nano-textures. Furthermore, the specific end groups of the two polymers may exhibit varying capacities to resisting non-specific protein adsorption. Continued investigation outlining the physical and chemical properties desirable for an anti-adsorptive nano-film coating will serve as a basis upon which to design durable implant-tissue interfaces that can react to various external stimuli.