Identifying factors associated with roadside work zone collisions using machine learning techniques.

Identifying factors that are associated with the probability of roadside work zone collisions enables decision makers to better assess and control the risk of scheduling a particular maintenance or construction activity by modifying the characteristics of the operation. This can be achieved by studying the effect of work zone properties on the risk of roadside work zone collisions. Much of the existing work in this area is based on data in the police traffic collision reports, which do not include data on the characteristics of the work zone itself. This paper develops a comprehensive data set of 42 features describing time, location, work zone characteristics, traffic volume, and road properties. Using recent machine learning techniques such as extreme gradient boosting classifiers on this extensive set of features allows for more accurate analysis to identify factors that affect the risk of work zone collisions or indicate higher than baseline chances of a roadside crash. Our statistical analysis reveals 10 important features and shows that four of these features are significantly associated with higher probabilities of roadside work zone collisions.

Speeding in highway work zone: An Evaluation of methods of speed control.

Highway workers frequently work in close proximity of live traffic in highway work zones, traffic accidents therefore have devastating effects on worker safety. In order to reduce the potential for such accidents, methods involving use of advisory signs and police presence have been used to mitigate accident risks and improve safety for highway workers. This research evaluates the magnitude of the speeding problem in highway work zones and the effects of four levels of police presence on improving work zone safety. Speed data were collected in six different work zone locations in northern and southern California and used to determine the magnitude and nature of speeding problem in highway work zones. In addition data were collected over 11 test-days in four work zones with four levels of police presence: radar speed display with police decal and lighting, passive use of a police vehicle with radar speed display, passive use of a police vehicle without radar speed display, and active police speed enforcement near work zones. This paper analyzes this data using statistical methods to evaluate the effectiveness of these different methods of speed control on the safety of the work zone. Four Measures of Effectiveness (MOE) were used in this evaluation consisting of average speed reduction, speed variance, 85th percentile speed, and proportion of high speed vehicles. The results indicate that all levels of police presence provided statistically significant improvements in one or more of the MOEs.

Factors influencing injury severity to highway workers in work zone intrusion accidents.

OBJECTIVE: Highway workers in work zones are in close proximity to traveling vehicles, exposing them to injury risks when vehicles intrude into the work zone. The purpose of this research was to perform an analysis of injuries endured by highway workers due to intrusion accidents and to identify factors that would have a significant effect on injury severity. METHODS: Ten years of California work zone injury data were collected and analyzed. The data were first used to determine trends in work zone injury. Statistical models were also created to evaluate variables influencing injury severity. Statistical models included multiple correspondence analysis, Cox proportional hazard regression, logistic regression, and Poisson regression. RESULTS: Statistical analysis of California injury data identified the 4 variables of accident/work zone location, work zone duration, time of day, and type of activity performed by the worker as having the most significant impact on injury severity. The results show that locations such as those on freeways/highways and stationary lane closures result in more severe injuries than work zones on city streets. Short-term stationary and short-duration work zones had increased odds of nonminor injuries compared to mobile work zones. For the time of day, the results indicate that the odds of more serious injuries are higher during nonpeak hours than during peak rush hours. Finally, workers on foot have greater odds of experiencing a more severe injury versus workers inside vehicles. CONCLUSION: This research has shown that considering the effects of work zone location, duration, time of day, and worker activity can have the most significant impact on risk of injury to workers. Understanding these factors can provide a basis for planning and design of work zones to improve worker safety.

A biomechanical evaluation of whiplash using a multi-body dynamic model.

This paper presents a biomechanical evaluation of whiplash injury potential during the initial extension motion of the head in a rear-end collision. A four-segment dynamic model is developed in the sagittal plane for the analysis. The model response is validated using the existing experimental data and is shown to simulate the "S-shape" kinematics of the cervical spine and the resulting dynamics observed in human and cadaver experiments. The model is then used to evaluate the effects of parameters such as collision severity, head/headrest separation, and the initial head orientation in the sagittal plane on the "S-shape" kinematics of the cervical spine and the resulting neck loads. It is shown, for example, that the cervical spine forms an "S-shape" for a range of change in speeds and that at lower and higher speeds changes the spine does not form the "S-shape." Furthermore, it is shown that the "S-shape" formation also depends on the head to headrest separation distance.