Modeling drivers’ reaction when being tailgated: A Random Forests Method

Background: Tailgating is a common aggressive driving behavior that has been identified as one of the leading causes of rear-end crashes. Previous studies have explored the behavior of tailgating drivers and have reported effective solutions to decrease the amount or prevalence of tailgating. This paper tries to fill the research gap by focusing on understanding highway tailgating scenarios and examining the leading vehicles’ reaction using existing naturalistic driving data. Method: A total of 1,255 tailgating events were identified by using the one-second time headway threshold criterion. Four types of reactions from the leading vehicles were identified, including changing lanes, slowing down, speeding up, and making no response. A Random Forests algorithm was employed in this study to predict the leading vehicle’s reaction based on corresponding factors including driver, vehicle, and environmental variables. Results: The analysis of the tailgating scenarios and associated factors showed that male drivers were more frequently involved in tailgating events than female drivers and that tailgating was more prevalent under sunny weather and in daytime conditions. Changing lanes was the most prevalent reaction from the leading vehicle during tailgating, which accounted for more than half of the total events. The results of Random Forests showed that mean time headway, duration of tailgating, and minimum time headway were three main factors, which had the greatest impact on the leading vehicle drivers’ reaction. It was found that in 95% of the events, leading vehicles would change lanes when being tailgated for two minutes or longer. Practical Applications: Results of this study can help to better understand the behavior and decision making of drivers. This understanding can be used in designing countermeasures or assistance systems to reduce tailgating behavior and related negative safety consequences.     

Effects of environment, vehicle and driver characteristics on risky driving behavior at work zones

This study aims to analyze the effects of environment, vehicle and driver characteristics on the risky driving behavior at work zones. A decision tree is developed using the classification and regression tree (CART) algorithm to graphically display the relationship between the risky driving behavior and its influencing factors. This approach could avoid the inherent problems occurred in the conventional logistic regression models and further improve the model prediction accuracy. Based on the Michigan M-94/I-94/I-94BL/I-94BR highway work zone driving behavior data, the decision tree comprising 33 leaf nodes is built. Bad weather, poor road and light conditions, partial/no access control, no traffic control devices, turning left/right and driving in an old vehicle are found to be associated with the risky driving behavior at work zones. The middle-aged drivers, who are going straight ahead in their vehicles with medium service time and equipped with an airbag system, are more likely to take risky behavior at lower work zone speed limits. Further, the middle-aged male drivers engage in risky driving behavior more frequently than the middle-aged female drivers. The number of lanes exhibits opposing effects on risky behavior under different traveling conditions. More specifically, the risky driving behavior is associated with the single-lane road under bad light or weather conditions while drivers are more likely to engage in risky behavior on the multi-lane road under good light conditions.     

Is road rage increasing? Results of a repeated survey

PROBLEM: We report on trends in road rage victimization and perpetration based on population survey data. METHOD: Based on repeated cross-sectional telephone surveys of Ontario adults between July 2001 and December 2003, logistic regression analyses examined differences between years in road rage victimization and perpetration in the previous year controlling for demographic characteristics. RESULTS: The prevalence of any road rage victimization in the previous year decreased significantly from 47.5% in 2001 to 40.6% in 2003, while prevalence of any road rage perpetration remained stable (31.0% to 33.6%). Logistic regression analyses revealed that the odds of experiencing any road rage victimization was 33% higher in 2001 and 30% higher in 2002, than in 2003. DISCUSSION: Survey data provide a valuable perspective on road rage trends, but efforts to track road rage incidents is also needed. SUMMARY: In Ontario, the proportion of adults experiencing any road rage victimization decreased from 2001 to 2003 while the proportion reporting any road rage perpetration remained stable. IMPACT ON INDUSTRY: None.     

Driver distraction by digital billboards? Structural equation modeling based on naturalistic driving study data: A case study of Iran

Introduction: Digital billboards (DBs) are a competing factor for attracting drivers’ attention; evidence shows that DBs may cause crashes and vehicle conflicts because they catch drivers’ attention. Because of the complexity of a system that includes road conditions, driver features, and environmental factors, it is simply not possible to identify relationships between these factors. Thus, the present study was conducted to provide a well-organized procedure to analyze the effects of DBs on drivers’ behavior and measure factors responsible for drivers’ distraction in Babol, Iran, as a case study. Method: Corresponding data were collected through a Naturalistic Driving Study (NDS) of 78 participants when facing DBs (1,326 samples). These data were analyzed by applying structural equation modeling (SEM) to concurrently recognize relationships between endogenous and exogenous variables. Human, environmental, and road factors were determined as exogenous latent variables in a model to evaluate their influences on drivers’ distraction as an endogenous variable. Results: The results showed that road, environmental, and human factors reciprocally interact with drivers’ distraction, although the estimated coefficient of human factors was more of a factor than that of the other groups. Furthermore, younger drivers, beginner drivers, and male drivers (as human factors); night and unclear weather like a rainy day (as environmental factors); and installing DBs at complicated traffic positions like near-intersections (as road factors) were determined to be the main factors that increase the possibility of drivers’ distraction. Finally, model assessment was suggested using the goodness-of-fit indices.     

Identifying an appropriate driving behaviour scale for the occupational driving context: The DBQ vs. the ODBQ

Self reported driving behaviour in the occupational driving context has typically been measured through scales adapted from the general driving population (i.e., the Manchester Driver Behaviour Questionnaire, (DBQ), Reason et al., 1990). However, research suggests that occupational driving is influenced by unique factors operating within the workplace environment, and thus, a behavioural scale should reflect those behaviours prevalent and unique within the driving context. To overcome this limitation, Newnam et al. (2011) developed the Occupational Driver Behaviour Questionnaire ((ODBQ), Newnam et al., 2011) which utilises a relevant theoretical model to assess the impact of the broader workplace context on driving behaviour. Although the theoretical argument has been established, research is yet to examine whether the ODBQ or the DBQ is a more sensitive measure of the workplace context. As such, this paper identifies selected organisational factors (i.e., safety climate and role overload) as predictors of the DBQ and the ODBQ and compares the relative predictive value in both models. In undertaking this task, 248 occupational drivers were recruited from a community-oriented nursing population. As predicted, hierarchical regression analyses revealed that the organisational factors accounted for a significantly greater proportion of variance in the ODBQ than the DBQ. These findings offer a number of practical and theoretical applications for occupational driving practice and future research.     

Market basket analysis of safety at active highway-railroad grade crossings

IntroductionData from the Federal Railroad Administration show high numbers of incidents at the approximately 210,446 highway-railroad grade crossings across the United States. One cause for this unsettling trend is the problem of drivers stopping within the dynamic envelope zone (DEZ) of the train while in queue. This research seeks to study DEZ stopping behavior at highway-railroad grade crossings by evaluating regulatory signage and further analyze variables that may affect this behavior. Method: A comparative safety analysis is undertaken to evaluate the effectiveness of the standard “Do Not Stop on Tracks” (R8-8) sign by using percentage change calculations and chi-squared tests. The study then conducts a market basket analysis (MBA) to extrapolate on these results and to identify underlying factors associated with observed driver behavior using variables influenced by visibility, perception, and maneuverability. Results: Rather low reductions in safety violations resulted from the R8-8 installation, including a 2.2% reduction in DEZ stopping behavior and only a slight 3.7% increase in compliance. The results of the MBA identified associations that affirmed assumptions about driver behavior, while other associations were not as direct but altogether helped broaden the understanding of interactions at grade crossings. Conclusions: This study concluded that the R8-8 had a positive but minimal effect on driver behavior at the grade crossings. The MBA successfully demonstrated its value by providing further insight on the safety analysis and by increasing the number of variables that can be analyzed simultaneously. The methodology offers the scientific community an innovative approach to analyzing driver behavior. Practical Applications: The results identified important variables for developing preventive measures, which will ultimately help reduce safety violations at grade crossings. The MBA can provide practical insight for railroad safety professionals and transportation engineers when determining problematic intersections and can be used to improve the education on grade crossing interactions.     

Effect prediction of time-gaps for adaptive cruise control (ACC) and in-vehicle tasks on bus driver performance

Researchers have devoted a great deal of attention to the effects of driver assist systems on driver performance. This article describes a modeling approach to simulate the effects of time-gaps for adaptive cruise control (ACC) and manual in-vehicle tasks on bus-driver performance. A concept model was built with the knowledge of modularization, parameterization, and parallel processing. By running the model, the predictions for the effects of five levels of time-gaps and two types of in-vehicle tasks were collected in three measures: (1) mean gap, (2) minimum gap, and (3) collision rate. The model performed well in prediction, especially when driving with in-vehicle tasks. Predictions from the model were validated by the experiment with a verified fixed-base bus-driving simulator, used in the authors’ previous studies. Throughout the modeling approach, this research provides a theoretical and accurate way to assess effects of time-gaps and vehicle-equipped interfaces. In follow-up research, the authors will apply this approach to evaluate other driving assist systems (e.g. collision warning systems and navigation systems) to create a customized software kit.     

Physiological signal analysis for fatigue level of experienced and inexperienced drivers

Objective: We studied the changes in driving fatigue levels of experienced and inexperienced drivers at 3 periods of the day: 9:00 a.m.–12:00 p.m., 12:00 p.m.–2:00 p.m., and 4:00 p.m.–6:00 p.m.

Methods: Thirty drivers were involved in 120-min real-car driving, and sleepiness ratings (Stanford Sleepiness Scale, SSS; Hoddes et al. 1973 Hoddes E, Zarcone V, Smythe H, Phillips R, Dement WC. Quantification of sleepiness: a new approach. Psychophysiology. 1973;10:431–436.[Crossref], [PubMed], [Web of Science ®] , [Google Scholar]), electroencephalogram (EEG) signals, and heart rates (HRs) were recorded. Together with principal component analysis, the relationship between EEG signals and HR was explored and used to determine a comprehensive indicator of driving fatigue. Then the comprehensive indicator was assessed via paired t test.

Results: Experienced and inexperienced drivers behaved significantly differently in terms of subjective fatigue during preliminary trials. At the beginning of trials and after termination, subjective fatigue level was aggravated with prolonged continuous driving. Moreover, we discussed the changing rules of EEG signals and HR and found that with prolonged time, the ratios of δ and β waves significantly declined, whereas that of the θ wave significantly rose. The ratio of (α + θ)/β significantly rose both before trials and after termination, but HR dropped significantly. However, one-factor analysis of variance shows that driving experience significantly affects the θ wave, (α + θ)/β ratio, and HR.

Conclusions: We found that in a monotonous road environment, fatigue symptoms occurred in inexperienced drivers and experienced drivers after about 60 and 80 min of continuous driving, respectively. Therefore, as for drivers with different experiences, restriction on continuous driving time would avoid fatigued driving and thereby eliminate traffic accidents. We find that the comprehensive indicator changes significantly with fatigue level. The integration of different indicators improves the recognition accuracy of different driving fatigue levels.     

A challenge to the assumed generalizability of prediction and countermeasure for risky driving: different factors predict different risky driving behaviors

In road safety, it may be debated whether all risky behaviors are sufficiently similar to be explained by similar factors. The often assumed generalizability of the factors that influence risky driving behaviors has been inadequately tested. Study 1 (N=116) examined the role of demographic, personality and attitudinal factors in the prediction of a range of risky driving behaviors, for young drivers. Results illustrated that different driving behaviors were predicted by different factors (e.g., speeding was predicted by authority--rebellion, while drink driving was predicted by sensation seeking and optimism bias). Study 2 (N=127) examined the generalizability of these results to the general driving population. Study 1 results did not generalize. Predictive factors remained behavior-specific, but different predictor-behavior relationships were observed in the community sample. Overall, results suggest that future research and practice should focus on a multi-factor framework for specific risky driving behaviors, rather than assuming generalizability across behaviors and driving populations.     

A spectral power analysis of driving behavior changes during the transition from nondistraction to distraction

Objective: This article investigated and compared frequency domain and time domain characteristics of drivers' behaviors before and after the start of distracted driving.

Method: Data from an existing naturalistic driving study were used. Fast Fourier transform (FFT) was applied for the frequency domain analysis to explore drivers' behavior pattern changes between nondistracted (prestarting of visual–manual task) and distracted (poststarting of visual–manual task) driving periods. Average relative spectral power in a low frequency range (0–0.5 Hz) and the standard deviation in a 10-s time window of vehicle control variables (i.e., lane offset, yaw rate, and acceleration) were calculated and further compared. Sensitivity analyses were also applied to examine the reliability of the time and frequency domain analyses.

Results: Results of the mixed model analyses from the time and frequency domain analyses all showed significant degradation in lateral control performance after engaging in visual–manual tasks while driving. Results of the sensitivity analyses suggested that the frequency domain analysis was less sensitive to the frequency bandwidth, whereas the time domain analysis was more sensitive to the time intervals selected for variation calculations. Different time interval selections can result in significantly different standard deviation values, whereas average spectral power analysis on yaw rate in both low and high frequency bandwidths showed consistent results, that higher variation values were observed during distracted driving when compared to nondistracted driving.

Conclusions: This study suggests that driver state detection needs to consider the behavior changes during the prestarting periods, instead of only focusing on periods with physical presence of distraction, such as cell phone use. Lateral control measures can be a better indicator of distraction detection than longitudinal controls. In addition, frequency domain analyses proved to be a more robust and consistent method in assessing driving performance compared to time domain analyses.     

Graduated driver licensing and safer driving

Graduated Driver Licensing (GDL) inserts between the leaner permit and full licensure an intermediate or "provisional" license that allows novices to drive unsupervised but subject to provisions intended to reduce the risks that accompany entry into highway traffic. Introduction of GDL has been followed by lowered accident rates, resulting from both limiting exposure of novices to unsafe situations and by helping them to deal with them more safely. Sources of safer driving include extended learning, early intervention, contingent advancement, and multistage instruction. To extend the learning process, most GDL systems lengthen the duration of the learner phase and require a specified level of adult-supervised driving. Results indicate that extended learning can reduce accidents substantially if well structured and highly controlled. Early intervention with novice traffic violators have shown both a general deterrent effect upon novice violators facing suspension and a specific effect upon those who have experienced it. Making advancement to full licensure contingent upon a violation-free record when driving on the provisional license has also evidenced a reduction in accidents and violations during that phase of licensure. Multistage instruction attempts development of advanced skills only after novices have had a chance to master more basic skills. Although this element of GDL has yet to be evaluated, research indicates crash reduction is possible in situations where it does not increase exposure to risk. While the various elements of GDL have demonstrated potential benefit in enhancing the safety of novice drivers, considerable improvement in the nature and enforcement of GDL requirements is needed to realize that potential.     

Difficulties in emotion regulation and risky driving among Lithuanian drivers

Background: Risky driving is a common cause of traffic accidents and injuries. However, there is no clear evidence of how difficulties in emotion regulation contribute to risky driving behavior, particularly in small post-Soviet countries. The present study aimed to investigate the relationship between difficulties in emotion regulation and self-reported risky driving behavior in a sample of Lithuanian drivers.

Methods: A total of 246 nonprofessional Lithuanian drivers participated in a cross-sectional survey. Difficulties in emotion regulation were assessed using the Difficulties in Emotion Regulation Scale (DERS; Gratz and Roemer 2004), and risky driving behavior was assessed using the Manchester Driver Behaviour Questionnaire (DBQ; Lajunen et al. 2004).

Results: Males scored higher than females in aggressive violations and ordinary violations. Females scored higher for the nonacceptance of emotional responses, whereas males had more difficulties with emotional awareness than females. More difficulties in emotion regulation were positively correlated with driving errors, lapses, aggressive violations, and ordinary violations for both males and females. Structural equation modeling showed that difficulties in emotion regulation explained aggressive and ordinary violations more clearly than lapses and errors. When controlling for interactions among the distinct regulation difficulties, difficulties with impulse control and difficulties engaging in goal-directed behavior predicted risky driving. Furthermore, nonacceptance of emotional responses and limited access to emotion regulation strategies were related to less violations and more driving errors.

Conclusion: Emotion regulation difficulties were associated with the self-reported risky driving behaviors of Lithuanian drivers. This provides useful hints for improving driver training programs in order to prevent traffic injuries.     

认知分心时驾驶人应激反应特征分析

为预防由认知分心影响驾驶人应激反应能力引发的交通事故,开展试验,在驾驶模拟器平台上模拟城市、乡镇、山区和高速公路等应激场景,测试有无分心任务2种条件下驾驶人的应激反应,采集驾驶人操作数据和车辆行驶数据,并利用Facelab5型眼动仪记录驾驶人眼动数据;选取相关表征指标并分析试验数据。结果表明:相比于正常驾驶,认知分心驾驶导致驾驶人视觉搜索范围变窄;无论正常驾驶还是认知分心驾驶,扫视行为均以中、低幅度为主;认知分心导致驾驶人对加速踏板的控制能力减弱;与正常驾驶相比,认知分心时驾驶人的应激反应时间明显增加。