Search

The knot of endogeneity and unobserved heterogeneity are causes of rendering parameter estimates inconsistent in real-time crash prediction. This study intends to alleviate the potential sources of these issues in detecting critical crashes, involving fatal or incapacitating injuries, versus non-critical crashes through a 402-mile Interstate-80 in Wyoming. Among different types of endogeneity, the problem of errors-in-variables and simultaneity was respectively mitigated by conflating disaggregated real-time traffic observations with aggregated environmental conditions and removing secondary crashes from the dataset. The possibility of omitted variables and unobserved heterogeneity were dealt by using random intercepts in hierarchical modeling under Bayesian inference. Three models were calibrated. Model-1 treated all predictors as fixed parameters. Model-2 and Model-3, respectively, considered weather and road surface conditions as random intercepts. Model-2 outperformed the others where the Intraclass Correlation Coefficients confirmed that the crash dataset is more nested within weather conditions. Results indicated that critical crashes require more interaction between vehicles, and shaping backward shockwave reduces their risk on Interstate-80 with a comparatively less traffic volume. Furthermore, considering different categories of weather and road surface conditions, critical crashes are more likely to occur on dry pavement and cloudy conditions compared to the wet surface and rainy days.

Providing drivers with real-time weather information and driving assistance during adverse weather, including fog, is crucial for safe driving. The primary focus of this study was to develop an affordable in-vehicle fog detection method, which will provide accurate trajectory-level weather information in real-time. The study used the SHRP2 Naturalistic Driving Study (NDS) video data and utilized several promising Deep Learning techniques, including Deep Neural Network (DNN), Recurrent Neural Network (RNN), Long Short-Term Memory (LSTM), and Convolutional Neural Network (CNN). Python programming on the TensorFlow Machine Learning library has been used for training the Deep Learning models. The analysis was done on a dataset consisted of three weather conditions, including clear, distant fog and near fog. During the training process, two optimizers, including Adam and Gradient Descent, have been used. While the overall prediction accuracy of the DNN, RNN, LSTM, and CNN using the Gradient Descent optimizer were found to be around 85 %, 77 %, 84 %, and 97 %, respectively; much improved overall prediction accuracy of 88 %, 91 %, 93 %, and 98 % for the DNN, RNN, LSTM, and CNN, respectively, were observed considering the Adam optimizer. The proposed fog detection method requires only a single video camera to detect weather conditions, and therefore, can be an inexpensive option to be fitted in maintenance vehicles to collect trajectory-level weather information in real-time for expanding as well as updating weather-based Variable Speed Limit (VSL) systems and Advanced Traveler Information Systems (ATIS)., (Copyright © 2020 Elsevier Ltd. All rights reserved.)

The presence of fog has a significant adverse impact on driving. Reduced visibility due to fog obscures the driving environment and greatly affects driver behavior and performance. Lane-keeping ability is a lateral driver behavior that can be very crucial in run-off-road crashes under reduced visibility conditions. A number of data mining techniques have been adopted in previous studies to examine driver behavior including lane-keeping ability. This study adopted an association rules mining method, a promising data mining technique, to investigate driver lane-keeping ability in foggy weather conditions using big trajectory-level SHRP2 Naturalistic Driving Study (NDS) datasets. A total of 124 trips in fog with their corresponding 248 trips in clear weather (i.e., 2 clear trips: 1 foggy weather trip) were considered for the study. The results indicated that affected visibility was associated with poor lane-keeping performance in several rules. Furthermore, additional factors including male drivers, a higher number of lanes, the presence of horizontal curves, etc. were found to be significant factors for having a higher proportion of poor lane-keeping performance. Moreover, drivers with more miles driven last year were found to have better lane-keeping performance. The findings of this study could help transportation practitioners to select effective countermeasures for mitigating run-off-road crashes under limited visibility conditions., (Copyright © 2019 Elsevier Ltd. All rights reserved.)

Introduction: Driving in foggy weather conditions has been recognized as a major safety concern for many years. Driver behavior and performance can be negatively affected by foggy weather conditions due to the low visibility in fog. A number of previous studies focused on driver performance and behavior in simulated environments. However, very few studies have examined the impact of foggy weather conditions on specific driver behavior in naturalistic settings., Method: This study utilized the second Strategic Highway Research Program (SHRP2) Naturalistic Driving Study (NDS) dataset to evaluate driver lane-keeping behavior in clear and foggy weather conditions. Preliminary descriptive analysis was conducted and a lane-keeping model was developed using the ordered logistic regression approach to achieve the study goals., Results: This study found that individual variables such as visibility, traffic conditions, lane change, driver marital status, and geometric characteristics, as well as some interaction terms (i.e., weather and gender, surface condition and driving experience, speed limit and mileage last year) significantly affect lane-keeping ability. An important finding of this study illustrated that affected visibility caused by foggy weather conditions decreases lane-keeping ability significantly. More specifically, drivers in affected visibility conditions showed 1.37 times higher Standard Deviation of Lane Position (SDLP) in comparison with drivers who were driving in unaffected visibility conditions., Conclusions: These results provide a better understanding of driver lane-keeping behavior and driver perception of foggy weather conditions. Moreover, the results might be used to improve Lane Departure Warning (LDW) systems algorithm by allowing them to account for the effects of fog on visibility. Practical Applications: These results provide a better understanding of driver lane-keeping behavior and driver perception of foggy weather conditions. Moreover, the results might be used to improve Lane Departure Warning (LDW) systems algorithm by allowing them to account for the effects of fog on visibility., (Copyright © 2018 Elsevier Ltd and National Safety Council. All rights reserved.)

The effect of reduction of visibility on crash occurrence has recently been a major concern. Although visibility detection systems can help to mitigate the increased hazard of limited-visibility, such systems are not widely implemented and many locations with no systems are experiencing considerable number of fatal crashes due to reduction in visibility caused by fog and inclement weather. On the other hand, airports' weather stations continuously monitor all climate parameters in real-time, and the gathered data may be utilized to mitigate the increased risk for the adjacent roadways. This study aims to examine the viability of using airport weather information in real-time road crash risk assessment in locations with recurrent fog problems. Bayesian logistic regression was utilized to link six years (2005-2010) of historical crash data to real-time weather information collected from eight airports in the State of Florida, roadway characteristics and aggregate traffic parameters. The results from this research indicate that real-time weather data collected from adjacent airports are good predictors to assess increased risk on highways., (Copyright © 2014 Elsevier Ltd. All rights reserved.)

• Developing a realistic microsimulation model requires adjusting driving behavior parameters. • This study updated lane change models in various weather using trajectory-level SHRP2 NDS dataset. • Representative key lane change parameters were extracted from an automated algorithm. • The demonstrated methodology could be used to develop weather-specific microsimulation models. Introduction: Adverse weather has a considerable negative impact on safety and mobility of transportation networks. Microsimulation models are one of the potential tools that could be used to evaluate the safety and operational impacts of adverse weather. The development of a realistic microsimulation model requires the adjustment of driving behavior parameters with disaggregate trajectory-level data. This study presented a novel approach to update and adjust lane change model parameters for the development of realistic microsimulation models in different weather conditions by leveraging the trajectory-level data from SHRP2 Naturalistic Driving Study (NDS). Method: Representative key lane change parameters in various weather conditions were extracted from an automatic identification algorithm. These lane change parameters were used to develop microsimulation models in VISSIM in an attempt to assess the safety and operational impacts of adverse weather on a freeway weaving segment. Results: The evaluation of safety impacts of adverse weather with regard to three Surrogate Measures of Safety (SMoS) namely Time-to-Collision (TTC), Post Encroachment Time (PET), and Deceleration Rate to Avoid Collision (DRAC) suggested that extreme adverse weather (including heavy rain, heavy snow, and heavy fog) produced a higher total number of simulated conflicts compared to clear weather. The operational analysis results revealed that adjusted parameters in most of the adverse weather produced lower average speeds with higher total travel times and total delays than clear weather. Conclusions: The outcomes of safety and operational assessments for the adjusted parameters showed that the development of microsimulation models should be based on weather-specific, rather than default parameters. Practical Applications: The methodology presented in this study could be adopted by transportation agencies to develop weather-specific microsimulation models. Moreover, the demonstrated approach could be used to evaluate different Connected Vehicle (CV) applications related to lane change in terms of safety and operations in microsimulation platforms. [ABSTRACT FROM AUTHOR]

The primary objective of this study was to develop a trajectory-level weather detection system capable of providing real-time weather information at the road surface level using only a single video camera. Two texture-based features, including histogram of oriented gradient (HOG) and local binary pattern (LBP), were extracted from images and used as classification parameters to train the weather detection models using several machine learning classifiers, such as gradient boosting (GB), random forest (RF), and support vector machine (SVM). In addition, a unique multilevel model, based on a hierarchical structure, was also proposed to increase detection accuracy. Evaluation results revealed that the multilevel model provided an overall accuracy of 89.2%, which is 3.2%, 7.5%, and 7.9% higher compared to the SVM, RF, and GB model, respectively, using the HOG features. Considering the LBP features, the multilevel model also produced the best performance with an overall accuracy of 91%, which is 1.6%, 8.6%, and 9% higher compared to the SVM, RF, and GB models, respectively. A sensitivity analysis using the proposed multilevel model revealed that the classification accuracy improved with the increasing number of HOG and LBP features at the expense of more computational powers. The proposed weather detection method is cost-efficient and can be made widely available mainly due to the recent booming of smartphone cameras and can be used to expand and update the current weather-based variable speed limit (VSL) systems in a connected vehicle (CV) environment. [ABSTRACT FROM AUTHOR]

This paper provides valuable insights into lane-changing characteristics based on driver behavior in fog and clear weather conditions using the Naturalistic Driving Study (NDS) data collected by the second Strategic Highway Research Program (SHRP2). While a few studies focused on lane-changing behavior based on driver type, the impact of adverse weather conditions was not addressed. This study tested several hypotheses to identify significant differences in number of lane-changing events per mile and lane-changing durations in fog and clear weather under different traffic regimes. Additionally, different distributions of lane-changing durations were fitted. Using K-means cluster analysis technique, drivers were classified into two categories: conservative and aggressive. It was found that in heavy fog the mean lane-changing durations were significantly higher than clear weather under mixed-flow conditions. Moreover, the cluster analysis results revealed that conservative drivers had longer lane-changing durations in heavy fog conditions compared to clear weather. The results of this study could be used in microsimulation model calibration and validation related to lane change in reduced visibility due to fog and various traffic conditions. [ABSTRACT FROM AUTHOR]

Connected vehicle (CV) technology aims to improve drivers' situational awareness through audible and visual warnings displayed on a human–machine interface (HMI), thus reducing crashes caused by human error. This paper developed a driving simulator test bed to assess the readability and usefulness of the Wyoming CV applications. A total number of 26 professional drivers were recruited to participate in a driving-simulator study. Prior to driving the simulator, the participants were trained on both the concept of CV technology and the developed CV applications as well as the operation of the driving simulator. Three driving simulation scenarios were designed. For each scenario, participants drove two times: one with the HMI turned on and another one with the HMI turned off. After driving the simulator, a comprehensive revealed-preference survey was employed to collect the participants' perceptions of CV technology and Wyoming CV applications. Results show that the Wyoming CV applications were most favored under poor-visibility driving conditions. Among the Wyoming CV applications, forward collision warning and rerouting applications were experienced as the most useful. Approximately 89% of the participants stated that the Wyoming CV applications provided them with improved road condition information and increased their experienced safety while driving; 65% of the participants stated the CV applications and the HMI did not introduce distraction from the primary task of driving. Finally, this paper concludes that the design of CV HMI needs to balance a trade-off between the readability of the warnings and drivers' capability to safely recognize and timely respond to the received warnings. [ABSTRACT FROM AUTHOR]

It is known that adverse weather conditions can affect driver performance due to reduction in visibility and slippery surface conditions. Lane keeping is one of the main factors that might be affected by weather conditions. Most of the previous studies on lane keeping have investigated driver lane-keeping performance from driver inattention perspective. In addition, the majority of previous lane-keeping studies have been conducted in controlled environments such as driving simulators. Therefore, there is a lack of studies that investigate driver lane-keeping ability considering adverse weather conditions in naturalistic settings. In this study, the relationship between weather conditions and driver lane-keeping performance was investigated using the SHRP2 naturalistic driving data for 141 drivers between 19 and 89 years of age. Moreover, a threshold was introduced to differentiate lane keeping and lane changing in naturalistic driving data. Two lane-keeping models were developed using the logistic regression and multivariate adaptive regression splines (MARS) to better understand factors affecting driver lane-keeping ability considering adverse weather conditions. The results revealed that heavy rain can significantly increase the standard deviation of lane position (SDLP), which is a very widely used method for analyzing lane-keeping ability. It was also found that traffic conditions, driver age and experience, and posted speed limits have significant effects on driver lane-keeping ability. An interesting finding of this study is that drivers have a better lane-keeping ability in roadways with higher posted speed limits. The results from this study might provide better insights into understanding the complex effect of adverse weather conditions on driver behavior. [ABSTRACT FROM AUTHOR]

• New methodologies were developed to add weather component to trajectory-level data. • Spatial and temporal thresholds were utilized to identify weather-related trips. • Frameworks were proposed to analyze trajectory-level naturalistic driving data. • Data reduction procedures were developed to manage the high resolution data. • A tool was developed to synchronize the time series, video footage, and radar data. Adverse weather conditions play a considerable role in the safety and efficiency of the transportation network. Many studies have aimed to quantify the impact that different weather conditions have on transportation safety and mobility; however, most studies have evaluated the network capacity, average speed, and other macroscopic measures without capturing specific driving characteristics. In order to understand specific driving behavior and performance characteristics that exist during different environmental conditions, high resolution vehicle data and video footage are required. The SHRP2 sponsored the generation of a large Naturalistic Driving Study (NDS) database – which provides vehicle time series data, front and rear video, driver video, external sensor readings, and driver surveys – and the Roadway Information Database (RID) – which is a complementary database with geospatial data for commonly driven roads in the NDS and other ancillary data sources, including annual traffic, roadway geometry, accident reports, weather conditions, and 511 alerts. The purpose of this study is to leverage these SHRP2 databases and weather data from the National Climatic Data Center (NCDC) to extract trips that occur during adverse weather conditions. The extraction of weather-related trips from a NDS is unprecedented, and this study presents three complementary methodologies used in parallel to acquire relevant trips from the SHRP2 NDS database. A semi-automated data reduction procedure was developed to process the raw trip files into a format that further analysis and modeling could be completed. This novel approach to NDS trip acquisition and reduction could be extended to other naturalistic driving studies worldwide. [ABSTRACT FROM AUTHOR]