2013 Session: 724

• Hot-Spot Identification: Categorical Binary Model Approach
  Abstract: This paper presents an alternative methodology for hot-spot identification based on a probabilistic model. In this methodology, the ranking criterion for hot-spot identification conveys the probability of a site being a hot-spot or a non-hot spot. A binary choice model was used to link the outcome to a set of factors that characterize the risk of the sites under analysis based on our use of two categories (0/1) for the dependent variable. The proposed methodology consists of two main steps. First, a threshold value for the number of accidents is set to distinguish hot spots from safe sites (category 1 or 0, respectively). Based on this classification, a binary model is applied that allows the construction of an ordered site list using the probability of a site being a hot-spot. The second step involves the choice of a selection strategy. The selection strategy can target a fixed number of sites with the greatest probability or, alternatively, all sites exceeding a specific probability, such as 0.5. A demonstration of the proposed methodology is provided using simulated data. For the simulation design, urban intersection data from Porto, Portugal, covering a five-year period were used. The results of the binary model showed a good fit. To evaluate and compare the probabilistic method with other commonly used methods, measures were used to test the performance of each method in terms of its power to detect the “true” hot spots. The test results indicate that the proposed method is superior to two commonly used methods. The gains of using this method are related to the simplicity of its application, while critical issues such as prior distribution effect assumptions and the regression-to-the-mean phenomenon are overcome. Further, the proposed model provides a realistic and intuitive perspective and supports easy practical application.
  Authors: Ferreira, Sara Pinho; Couto, António Fidalgo
  Authors: Ferreira, Sara Pinho; Couto, António Fidalgo
  Year: 2013
  Document Type: Paper
  Subject: Data and Information Technology; Safety and Human Factors
  Session: 724
  Paper Number: 13-0095
• Application of Stochastic Gradient Boosting Technique to Enhance Reliability of Real-Time Risk Assessment Using Automatic Vehicle Identification and Remote Traffic Microwave Sensor Data
  Abstract: This study proposes a recent promising machine learning technique to enhance the reliability of real-time risk assessment on freeways. Stochastic Gradient Boosting (SGB) is utilized to identify hazardous conditions based on traffic data collected from multiple detection systems; automatic vehicle identification (AVI) and remote traffic microwave sensors (RTMS), real-time weather stations and roadway geometry. SGB’s key strengths lie in its capability to fit complex nonlinear relationships, handling different types of predictors and accommodating missing values with no need for prior transformation of the predictor variables or elimination of outliers, which is the case of real-time applications. Boosting multiple simple trees together overcomes the drawback of single tree models of poor prediction accuracy and provides fast and superior predictive performance. In this paper, three models were calibrated; full model that is augmenting all available data and another two models to explicitly compare between the prediction performance of traffic data that are collected from different sources (AVI and RTMS) at the same location. The results from the preliminary analysis as well as the calibrated models indicate that crash prediction from AVI is comparably equivalent to RTMS data. Moreover, the full model achieved superior classification accuracy identifying about 89% of crash cases in the validation dataset with only 6.5% false positive rate. Because of the superior classification performance of SGB and its minimal required data preparation, SGB is recommended as a promising technique for real-time risk assessment application.
  Authors: Ahmed, Mohamed M.; Abdel-Aty, Mohamed A.
  Authors: Ahmed, Mohamed M.; Abdel-Aty, Mohamed A.
  Year: 2013
  Document Type: Paper
  Subject: Data and Information Technology; Safety and Human Factors
  Session: 724
  Paper Number: 13-0410
• Analyzing Effect of All-Red Intervals in Crash Reduction: Case Study of Heckman Correction for Urban Signalized Intersection Crashes
  Abstract: All-Red (AR) interval is designed as a method of clearance interval to safely clear vehicles that enter the intersection dilemma zone. The provision of AR is generally expected to reduce the occurrence of crashes, though there are situations that AR is not proved to be effective because it is used at intersections with a higher potential for crashes. This controversial result however, does not indicate that the AR interval is a contributing cause of crashes. Therefore, the self-selection bias of signal designs needs to be corrected when estimating their effect in improving safety. To address the selection-bias problem at signalized intersections, a Heckman two-stage approach is adapted. First, a probit model is developed to explain the interrelationship between the AR interval and highway geometry, traffic volume, and environmental variables. Second, the selection bias term (or Heckman correction) is included in the second stage to build two negative binomial models for locations with and without an AR interval. Further, average treatment effects (ATE) and effect of treatment on the treated (TT) are estimated to examine the effect of AR intervals on the whole sample and treated sample, respectively. Three-year crash data on urban signalized intersections in the Detroit metro area is used to validate the proposed models. The results show that a random intersection with an AR interval will reduce crashes by 36 percent when compared to a non-AR interval intersection. For treated intersections (with AR interval) there is a 51 percent reduction of total crashes compared to intersections without treatment (if not designed with AR interval). The AR interval is a meaningful advance in reducing crashes by 15 percent. Key words: Self-selection bias, Heckman Two-step correction model, All-Red Interval, Probit Model, Negative Binomial Model
  Authors: Mishra, Sabyasachee; Zhu, Xiaoyu
  Authors: Mishra, Sabyasachee; Zhu, Xiaoyu
  Year: 2013
  Document Type: Paper
  Subject: Data and Information Technology; Safety and Human Factors
  Session: 724
  Paper Number: 13-0548
• Feasibility of Incorporating Reliability Analysis in Traffic Safety Investigation
  Abstract: In this paper, the method of reliability analysis has been employed to investigate the feasibility of using it in traffic safety analysis. The reliability analysis approach, frequently used to evaluate the probabilities of failures for a specific structural system, has two main outcomes which are the reliability index and design points. Two different approaches to use these two outcomes in traffic safety analysis have been presented in this paper. Data from a mountainous freeway in Colorado was used. The reliability index was utilized to evaluate the hazardous freeway segments by incorporating the traffic flow parameters provided by radar detectors. The design points were employed to predict the crash occurrence at the disaggregate level with weather parameters. Finally the results from both approaches have been compared to the results from a traditional method, and the reliability analysis method showed promising applications in traffic safety. By using the reliability indexes, the three most hazardous segments are consistent with the results from the crash rates segment ranking approach; for the design points, by utilizing these thresholds the accuracy rate of predicting crash occurrence could be improved by 10% compared to the logistic regression method.
  Authors: Yu, Rongjie; Shi, Qi; Abdel-Aty, Mohamed A.
  Authors: Yu, Rongjie; Shi, Qi; Abdel-Aty, Mohamed A.
  Year: 2013
  Document Type: Paper
  Subject: Data and Information Technology; Safety and Human Factors
  Session: 724
  Paper Number: 13-0477
• Crash-Type Propensity Analysis with Bayesian Models Using Microscopic Traffic and Weather Data
  Abstract: This study investigates a range of effects of microscopic traffic and weather factors and roadway geometry information on the specific crash type for a mountainous freeway. Crashes have been categorized as rear-end, sideswipe and single-vehicle crashes. Six-minute Automatic Vehicle Identification (AVI) segment average speed, real-time weather data and roadway geometry data are utilized as explanatory variables in this study. First, two binary logistic regression models were estimated by comparing single-vehicle to multi-vehicle crashes and sideswipe crashes to rear-end crashes. Then a full model which simultaneously fits two conditional logistic regression models (mixed logit model) for the three crash types has also been estimated. Results from the models indicate that single-vehicle crashes are more probable in the snow season, at moderate slopes, three-lane segments, under the free-flow conditions; while the sideswipe crash occurrence differs from rear-end crashes with the visibility situation, number of lanes, grades and their directions (up or down). Moreover, the results of the Bayesian random effects logistic regression models have been compared with the results from the classic logistic regression with the Frequentist and Bayesian inference techniques. It was demonstrated that the Bayesian random effects logistic regression outperforms the other two approaches with higher accuracy and lower Brier scores. The innovative way of estimating two conditional logistic regression models simultaneously in the Bayesian framework fits the data structure well. Conclusions from this study imply that different active traffic management strategies should be designed for three- and two-lane roadway sections and also considering the seasonal effects.
  Authors: Yu, Rongjie; Abdel-Aty, Mohamed A.; Ahmed, Mohamed M.; Wang, Xuesong
  Authors: Yu, Rongjie; Abdel-Aty, Mohamed A.; Ahmed, Mohamed M.; Wang, Xuesong
  Year: 2013
  Document Type: Paper
  Subject: Data and Information Technology; Safety and Human Factors
  Session: 724
  Paper Number: 13-0718
• Evaluation of Postencroachment Time as a Surrogate for Opposing Left-Turn Crashes
  Abstract: Highway safety evaluation has traditionally been performed using crash data though this method has limitations in terms of timeliness and efficiency. Previous studies show that the use of surrogate safety data allows for faster evaluation of safety in comparison to the significantly longer time horizon required for collecting crash data. However, the predictive capability of surrogate measures is still an area of ongoing research. Previous studies have often resulted in inconsistent findings for the relationship between surrogates and crashes, one of the primary reasons being inconsistent definitions of a conflict. This study evaluates the effectiveness of Post Encroachment Time (PET) as a surrogate measure for evaluating the propensity of crashes between left-turning vehicles and opposing through vehicles at 4-legged signalized intersections. The primary method of data collection is through video recording with post-processing using custom semi-automatic video processing software to reduce the video to a useable format ready for analysis. The study evaluates the effectiveness of PET as a surrogate measure by comparing three variations of PET measures with crash history. This comparison shows that a threshold value of PET plays an important role in establishing its correlation with crashes with the best results at a threshold as low as one second.
  Authors: Peesapati, Lakshmi; Hunter, Michael P.; Rodgers, Michael Owen
  Authors: Peesapati, Lakshmi; Hunter, Michael P.; Rodgers, Michael Owen
  Year: 2013
  Document Type: Paper
  Subject: Data and Information Technology; Safety and Human Factors
  Session: 724
  Paper Number: 13-0839
• Spatial Analysis of Fatal and Injury Crashes in Flanders, Belgium: Application of Geographically Weighted Regression Technique
  Abstract: Generalized Linear Models (GLMs) are the most widely used models utilized in crash prediction studies. These models illustrate the relationships between the dependent and explanatory variables by estimating fixed global estimates. Since the crash occurrences are often spatially heterogeneous and are affected by many spatial variables, the existence of spatial correlation in the data is examined by means of calculating Moran’s I measures for dependent and explanatory variables. The results indicate the necessity of considering the spatial correlation when developing crash prediction models. The main objective of this research is to develop different Zonal Crash Prediction Models (ZCPMs) within the Geographically Weighted Generalized Linear Models (GWGLM) framework in order to explore the spatial variations in association between Number of Injury Crashes (NOICs) (including fatal, severely and slightly injury crashes) and other explanatory variables. Different exposure, network and socio-demographic variables of 2200 Traffic Analysis Zones (TAZs) are considered as predictors of crashes in the study area, Flanders, Belgium. To this end, an activity-based transportation model framework is applied to produce exposure measurements while the network and socio-demographic variables are collected from other sources. Crash data used in this study consist of recorded crashes between 2004 and 2007. GWGLMs are developed using a Poisson error distribution and are often referred to as Geographically Weighted Poisson Regression (GWPR) models. Moreover, the performances of developed GWPR models are compared with their corresponding GLMs. The results show that GWPR models outperform the GLM models; this is due to the capability of GWPR models in capturing the spatial heterogeneity of crashes.
  Authors: Pirdavani, Ali; Brijs, Tom; Bellemans, Tom; Wets, Geert
  Authors: Pirdavani, Ali; Brijs, Tom; Bellemans, Tom; Wets, Geert
  Year: 2013
  Document Type: Paper
  Subject: Data and Information Technology; Safety and Human Factors
  Session: 724
  Paper Number: 13-1049
• Reducing Severity of Crashes During Holidays: Are We Targeting the Right Behaviors?
  Abstract: Holidays are times of recreation, rest and relaxation to be enjoyed with our loved ones. Hence, any road traffic accidents and the subsequent deaths and injuries tend to receive more media attention and evoke stronger public reactions because they are more tragic as they turn our merriment into grief and mourning. Consequently, many jurisdictions have implemented more aggressive police enforcement and publicity campaigns targeted at reducing risky driving behaviors during the holidays. However, relatively little formal research has been conducted to specifically identify the factors contributing to crashes during holidays. Using data from 1999-2008, this research endeavours to identify the behavioral factors that statistically and significantly contribute to the severity of holiday crashes involving two-vehicles. In addition, the impact of different control variables formed from crash, vehicle, road surface and other behavioral factors will also be explored. Our results indicate that drivers’ violation, drivers’ error, drivers’ intoxication and non-use of seat-belts significantly contribute to increasing the severity of holiday crashes. However, the impact of unsafe speeding is found to be insignificant in the study. The results obtained suggest that it may be time to consider a more balanced approach to the road safety blitzes conducted during holidays.
  Authors: Anowar, Sabreena; Yasmin, Shamsunnahar; Tay, Richard
  Authors: Anowar, Sabreena; Yasmin, Shamsunnahar; Tay, Richard
  Year: 2013
  Document Type: Paper
  Subject: Data and Information Technology; Safety and Human Factors
  Session: 724
  Paper Number: 13-1095
• Exploring Behavioral Responses of Motorists to Risk-Based Charging Mechanisms
  Abstract: This paper reports on the behavioural response of motorists to a variable rate charging scheme designed to encourage safer driving practices and reduce exposure to crash-risk – specifically kilometres driven, night-time driving and speeding. The study involved a five-week ‘before’ period of GPS monitoring to establish how motorists drove normally, followed by a five-week ‘after’ period of GPS monitoring in which charges were levied and changes assessed. Incentives were paid to motorists for the difference in the charges between the two five-week periods. Vehicle kilometres travelled (VKT) were reduced by ten percent, although the sample was evenly split by those increasing VKT compared to those decreasing VKT. The proportion of distance speeding fell by 4.7 percent, which when coupled with decreases in VKT, implied a net reduction of kilometres spent speeding of over 40 percent. Three-quarters of the sample reduced their speeding. Exit interviews with a cross-section of participants highlighted the practical difficulties of reducing kilometres, but (more encouragingly) reinforced the potential to reduce speeding.
  Authors: Greaves, Stephen; Fifer, Simon
  Authors: Greaves, Stephen; Fifer, Simon
  Year: 2013
  Document Type: Paper
  Subject: Data and Information Technology; Safety and Human Factors
  Session: 724
  Paper Number: 13-1128
• Identifying Primary and Secondary Accidents from Spatiotemporal Accident Impact Analysis
  Abstract: The identification of secondary accidents is accompanied by the definition of the primary accident impact area. Although the accident impact area varies with the geometric characteristics of roads and periodic characteristics of traffic flow as well as with accident types, most previous studies used a fixed boundary to identify secondary accidents and primary accidents. Thus, the objective of this research is to develop a method to define the spatio-temporally different boundaries varying with different types of accident. Based on the developed boundaries, the secondary accident is identified in the primary accident location as well as in its opposite direction. Secondary accidents in the same and opposite directions were identified to be 8.1% and 3.7% of total primary accidents, respectively. Also, only 0.4% of total primary accidents were connected with the secondary accident both in the same and opposite directions. Although the proposed method seems to be complicated, the results from the method will be useful to understand secondary accident characteristics in more realistic analysis through the spatio-temporal accident impact area in the accident direction as well as in its opposite direction. Specifically, they can be used by public sector transportation agencies in making operational strategies for reducing the secondary accidents on freeways.
  Authors: Chung, Younshik
  Authors: Chung, Younshik
  Year: 2013
  Document Type: Paper
  Subject: Data and Information Technology; Safety and Human Factors
  Session: 724
  Paper Number: 13-1148
• Redesigning Black Spots in Traffic: Effect Evaluation
  Abstract: This study evaluates the safety effects of an extensive black spot program that has been implemented in Flanders-Belgium. Based on their crash history, 800 locations were selected as black spots. The study evaluates 134 locations that were redesigned between 2004 and 2007. The adopted approach is an Empirical Bayes before-and-after study that accounts for effects of general trends and for the stochastic nature of crashes, including regression to the mean. Two different comparison groups were established. Dependent on the applied comparison group, the analyses showed a decrease in the number of injury crashes of 24 to 27%, significant at the 1%-level. A separate analysis for crashes with serious or fatal injuries showed a decrease of 40 to 52%, also significant at the 1% level. ANOVA-analyses were made to check whether differences in effects occur depending on the characteristics of the location or the implemented intersection design. The results suggest a more favourable evolution for intersections that were priority controlled in the before situation compared with signal-controlled intersections. Crash reductions were also higher at locations with a lower traffic volume compared to locations with a higher volume.
  Authors: De Pauw, Ellen; Daniels, Stijn; Brijs, Tom; Hermans, Elke; Wets, Geert
  Authors: De Pauw, Ellen; Daniels, Stijn; Brijs, Tom; Hermans, Elke; Wets, Geert
  Year: 2013
  Document Type: Paper
  Subject: Data and Information Technology; Safety and Human Factors
  Session: 724
  Paper Number: 13-1465
• Crash Fault Analysis of Out-of-State Drivers in Vermont
  Abstract: This study examined single- and two-vehicle police-reported crashes in Vermont between 2003 and 2008. It evaluated the likelihood of being at fault for Vermont drivers versus out-of-state drivers. Analysis using odds ratios estimated that out-of-state drivers are 98% more likely to be at-fault for a single-vehicle crash and 9% more likely to be at-fault for a two-vehicle crash.Age, gender, season of year, light conditions, and road type were statistically significant interactions for Vermont and out-of-state drivers for single-vehicle crashes. Male drivers and driving during the winter months had more pronounced effects of increasing single-vehicle crash fault for out-of-state drivers than for Vermont drivers. Vermont drivers, on the other hand, were more apt to cause a crash on gravel roads.The interactions were less pronounced for two-vehicle crashes. Being male or an older driver increased crash odds for both groups. Driving during the summer months increased out-of-state drivers crash odds by 21%, while it was insignificant for Vermont drivers. The other factors tested were insignificant for both groups.The crash evaluation of fault for “foreign” drivers’ crashes has been understudied in the United States. Previous research, conducted mostly in other countries, has been limited but has shown that foreign drivers are more likely to be involved in a crash. This study in Vermont strongly suggests the need for further study of this factor as well as identification of associated interventions.
  Authors: Harootunian, Kristine; Aultman-Hall, Lisa; Lee, Brian H. Y.
  Authors: Harootunian, Kristine; Aultman-Hall, Lisa; Lee, Brian H. Y.
  Year: 2013
  Document Type: Paper
  Subject: Data and Information Technology; Safety and Human Factors
  Session: 724
  Paper Number: 13-1597
• Indexing Crashworthiness and Crash Aggressiveness by Major Car Brands
  Abstract: This study aims at indexing crash worthiness and crash aggressivity of 23 major car brands in Florida with consideration of the brand origin. It contributes to the literature by proposing a method for redefining the safety performance of cars by taking into account the cars¡¯ hazardousness imposed to counterpart cars that are involved in the same crashes. A Bayesian hierarchical ordered logistic model was applied to relate the injury severity level of drivers to crash compatibility of car brands. In the models, we assume that the driver injury depends on the difference of the striking cars¡¯ aggressivity and the struck cars¡¯ self-protectiveness in two-vehicle crashes with external factors controlled. A total of 17,178 two-vehicle-crash records with 34,356 car involvements in Florida were used in the investigation. The results show that most of the premium cars such as Volvo, Cadillac, Infiniti and Lexus possess excellent crash worthiness and relatively low crash aggressivity. Self-protection abilities of popular car brands such as Ford, Toyota, Honda and Chevrolet vary considerably, but their hazardousness perform similarly and are lower than the average level. European cars perform relatively good self-protection but are also more hazardous to the counterpart cars when crashes occur. Japanese cars show lower worthiness and aggressivity than American cars, while South Korean cars are associated with the lowest crash worthiness and mean crash aggressivity.
  Authors: Huang, Helai; Hu, Shuiyan; Abdel-Aty, Mohamed A.
  Authors: Huang, Helai; Hu, Shuiyan; Abdel-Aty, Mohamed A.
  Year: 2013
  Document Type: Paper
  Subject: Data and Information Technology; Safety and Human Factors
  Session: 724
  Paper Number: 13-1754
• Assessment of Exposure Proxies for Macroscopic Road Safety Prediction
  Abstract: Road safety is a major global health problem and no effort should be spared in trying to limit its impacts. Modeling road safety is a complex task, which needs to consider both the quantifiable impact of specific parameters, as well as the underlying trends that cannot always be measured or observed. Macroscopic data are often not available, or not in the form that they are desired. Therefore, it is often required to attempt to consider alternative sources of data, which may be correlated with the modeled phenomenon. The objective of this research is to investigate the suitability of alternative proxy variables for macroscopic road safety modeling, using three suitable exposure proxies: (i) number of vehicles in circulation, (ii) GDP and (iii) fuel consumption. Several structural time-series models have been developed for each proxy for two Mediterranean countries with many similar socio-economic characteristics: Greece and Cyprus.Based on the findings of this analysis, a number of observations can be drawn. Proxy variables can provide reasonable results, when exposure data are not available. Furthermore, even in two countries with many similarities the selected proxy measure differs. This suggests that the underlying conditions that make a variable a suitable proxy for exposure is complex and needs further investigation.
  Authors: Antoniou, Constantinos; Yannis, George
  Authors: Antoniou, Constantinos; Yannis, George
  Year: 2013
  Document Type: Paper
  Subject: Data and Information Technology; Safety and Human Factors
  Session: 724
  Paper Number: 13-1788
• Analyzing Relationship Between Car Generation and Severity of Motor-Vehicle Crashes in Denmark
  Abstract: While in the last 40 years Danish roads have observed a decrease in the number of fatalities, research has not investigated the contribution of legislative, enforcement, technological, infrastructural and human factors to this reduction. In the context of a Danish car market with remarkably high registration tax causing potential buyers to hold longer onto old cars, the relationship between technological enhancements of vehicles and severity of crashes requires particular attention. The current study investigates the relationship between car generation (i.e., car’s first registration year) and injury severity sustained by car drivers involved in accidents in Denmark between 2004 and 2010. A generalized ordered logit model is estimated while controlling for several characteristics of the crash, the vehicle and the persons involved, and scenario analysis is performed for assessing the effect of car generation on drivers’ injury severity. Results illustrate that newer car generations are associated to significantly lower probability of injury and fatality, and that replacing older cars with newer ones introduces significant and not to be overlooked benefits for both population and society.
  Authors: Rich, Jeppe Husted; Prato, Carlo Giacomo; Hels, Tove; Lyckegaard, Allan; Kristensen, Niels Buus
  Authors: Rich, Jeppe Husted; Prato, Carlo Giacomo; Hels, Tove; Lyckegaard, Allan; Kristensen, Niels Buus
  Year: 2013
  Document Type: Paper
  Subject: Data and Information Technology; Safety and Human Factors
  Session: 724
  Paper Number: 13-1246
• Predicting Freeway Crash Likelihood and Severity with Real-Time Loop Detector Data
  Abstract: Real-time crash risk prediction using traffic data collected from loop detector stations is useful in dynamic safety management systems aimed at improving traffic safety through application of proactive safety countermeasures. The major drawback of most of the existing studies is that they focus on the crash risk without consideration of crash severity. This paper presents an effort to develop a model that predicts the crash likelihood at different levels of severity with a particular focus on severe crashes. The crash data and traffic data used in this study were collected on the I-880 freeway in California, United States. This study considers three levels of crash severity: fatal/incapacitating injury crashes (KA), non-incapacitating/possible injury crashes (BC), and property-damage-only crashes (PDO). The sequential logit models were developed to link the likelihood of crash occurrences at different severity levels to various traffic flow characteristics derived from detector data. The fitness and prediction capability of the forward and backward versions of the models were compared to select a better alternative. The results show that the sequential structure (forward vs. backward) does not have considerable impact on the model¡¯s fitness and predictive capabilities. More interestingly, the traffic flow characteristics contributing to crash likelihood were quite different at different levels of severity. The PDO crashes were more likely to occur under congested traffic flow conditions with highly variable speed and frequent lane changes, while the KA and BC crashes were more likely to occur under less congested traffic flow conditions. High speed, coupled with a large speed difference between adjacent lanes under uncongested traffic conditions, was found to increase the likelihood of severe crashes (KA). This study applied the 20-fold cross-validation method to estimate the prediction performance of the developed models. The validation results show that the model¡¯s crash prediction performance at each severity level was satisfactory. The findings of this study can be used to predict the probabilities of crash at different severity levels, which is valuable knowledge in the pursuit of reducing the risk of severe crashes through the use of dynamic safety management systems on freeways.
  Authors: Xu, Chengcheng; Tarko, Andrew P.; Wang, Wei; Liu, Pan; Bai, Lu
  Authors: Xu, Chengcheng; Tarko, Andrew P.; Wang, Wei; Liu, Pan; Bai, Lu
  Year: 2013
  Document Type: Paper
  Subject: Data and Information Technology; Safety and Human Factors
  Session: 724
  Paper Number: 13-1715
• Surrogate Safety Measure for Simulation-Based Conflict Study
  Abstract: This paper proposes a surrogate measure named Aggregated Crash Propensity Index (ACPI) for simulation-based conflict studies. A Probabilistic model named Crash Propensity Model (CPM) is developed to determine the crash probability of simulated conflicts, by introducing the distributions of reaction time and maximum braking rates. This CPM is able to generate ACPI for three different types: crossing, rear-end and lane change. A field validation effort is conducted by simulating three major arterials (twelve intersections) in simulation package (VISSIM). Surrogate Safety Assessment Model (SSAM) is utilized to extract useful conflict data as the entry into CPM model to get ACPI. The Spearman rank tests indicate that ACPI is able to identify the relative safety among traffic facilities/treatments. Notably, ACPI outperforms the Highway Safety Manual (HSM) procedures in both correlation and rank tests. Both linear and non-linear regression models are well fitted for ACPI and real crash frequency, suggesting its potential to be directly linked to real crash.
  Authors: Wang, Chen; Stamatiadis, Nikiforos
  Authors: Wang, Chen; Stamatiadis, Nikiforos
  Year: 2013
  Document Type: Paper
  Subject: Data and Information Technology; Safety and Human Factors
  Session: 724
  Paper Number: 13-1512
• Urban-Rural Difference of Gasoline Price Effects on Traffic Safety
  Abstract: A large literature base has found that economic factors have important effects on traffic crashes. A small but growing branch of literature also examines the role of gasoline prices in the occurrence of traffic crashes. However, no studies have investigated the possible difference of these effects between urban and rural areas. In this study, we used the monthly traffic crash data from 1998–2007 at the county level in Minnesota to investigate the possibly different effects gasoline prices may have on traffic crashes per million vehicle miles traveled in urban versus rural areas. The results indicate that gasoline price effects on total crashes, property-damage-only crashes, and injury crashes are stronger in rural areas than in urban areas. Gasoline prices also significantly affect fatal crashes in both urban and rural areas; however, the difference is not significant. The results concerning the differences between urban and rural areas have important policy implications for traffic safety planners and decision makers.
  Authors: Chi, Guangqing; Quddus, Mohammed A.; Huang, Arthur; Levinson, David M.
  Authors: Chi, Guangqing; Quddus, Mohammed A.; Huang, Arthur; Levinson, David M.
  Year: 2013
  Document Type: Paper
  Subject: Data and Information Technology; Safety and Human Factors
  Session: 724
  Paper Number: 13-1688
• Analysis of Factors Affecting Winter Collision Severity
  Abstract: This paper presents the results of an analysis aiming at identifying the main injury severity factors associated with road collisions that occur during snowstorms, including traffic conditions, road geometry and environment, pavement surface conditions as well as vehicle and driver characteristics. A multilevel multinomial logit model is introduced for capturing the hierarchical nature of the collision data between individual collisions and the vehicles and persons involved. Different from past studies, the modeling effort focuses on the collisions that occurred over snowstorms so that the effect of weather related factors are not masked due to the imbalance of data sample between collisions occurred under normal conditions and those under snowstorms. This approach is also necessary for ensuring that the incremental effect of different weather severity, as well as winter road maintenance operations, could be captured. Collisions occurred on a number of highway routes from the province of Ontario, Canada, over six winter seasons (2000-2006), were selected for this analysis. It was found that factors related to drivers (age, sex, condition), road characteristics (number of lanes, speed limit, road surface conditions), vehicle type, position in vehicle, use of safety belt, and traffic volume have statistically significant effects on collision severity outcome. In general, the modeling results indicate that good road surface conditions, high traffic volume, young and male drivers and new vehicles are associated with reduced injury severity levels. Our analysis, however, did not confirm the main finding from literature, that is, severer weather, such as higher precipitation intensity and wind speed, is associated with lesser collision severity.
  Authors: Usman, Taimur; Fu, Liping; Miranda-Moreno, Luis Fernando
  Authors: Usman, Taimur; Fu, Liping; Miranda-Moreno, Luis Fernando
  Year: 2013
  Document Type: Paper
  Subject: Data and Information Technology; Safety and Human Factors
  Session: 724
  Paper Number: 13-1669
• Analyzing Crash Severity Based on Vehicle Damage and Occupant Injuries
  Abstract: In recent years, the reduction of injury crashes has been heralded as a great success. Improvements in federally mandated safety standards and advancements made by automotive industries to enhance the vehicle safety can be partially credited with the decline. Now, the national strategy on highway safety is to move “Toward Zero Deaths”. From this “Vision Zero” perspective, one of the appropriate strategies is to manage kinetic energy in crashes and collisions, i.e. minimizing the energy transferred to the human body, because the kinetic energy is responsible for occupant injuries and fatalities. Vehicle damage conditions are an unbiased indicator of kinetic energy in collisions while injury severities are the ultimate measure of the occupant risks. In this study, the vehicle damage and occupant injury models were developed for single-vehicle (SV) and multiple-vehicle (MV) crashes, respectively. The results of these models provide a complete view of the crash severity determinants and how they affect the occupant injuries and vehicle damage. Some factors have consistent impact across both injury severities and vehicle damage, while others are contradictory. Combining information from both occupants and vehicles is valuable for an impartial evaluation of specific components in highway design and an accurate assessment of the impacts of occupant characteristics, driver behavior, and errors on the resultant bodily injuries.
  Authors: Qin, Xiao; Wang, Kai; Cutler, Chase E.
  Authors: Qin, Xiao; Wang, Kai; Cutler, Chase E.
  Year: 2013
  Document Type: Paper
  Subject: Data and Information Technology; Safety and Human Factors
  Session: 724
  Paper Number: 13-2072
• Modeling Traffic Accidents on Auckland Motorway, New Zealand
  Abstract: This paper investigates motorway safety by developing accident prediction models that link accident frequencies to their non-behavioural contributing factors, including traffic conditions, geometric and operational characteristics of road, and weather conditions. The study used a sample of accidents occurred from 2004 through 2010 on a 74 km long section of Auckland motorway. A number of accident prediction models were developed and assessed for their predictive ability using negative binomial regression models under three categories: first for the whole of the motorway, second for rural and urban motorway segments separately and third for motorway segments without ramp, with on-ramp and with off-ramp separately. The results uncovered the safety impacts of different non-behavioural contributing factors, in which segment length, AADT per lane and the number of lanes always have the most profound effects on accident frequency. The findings make the recommendation of effective countermeasures on motorway safety to be possible.
  Authors: Chngye, Pan; Ranjitkar, Prakash
  Authors: Chngye, Pan; Ranjitkar, Prakash
  Year: 2013
  Document Type: Paper
  Subject: Data and Information Technology; Safety and Human Factors
  Session: 724
  Paper Number: 13-1976
• Methodology for Determining Traffic Accident Risk Zones
  Abstract: In Costa Rica, the traffic accident database is still under development. Due to the limited quantity of information it is very difficult for the DOT to the accurately locate the road sections with significant concentration of accidents, also known as “blackspots”. The National Laboratory of Materials and Structural Models of the University of Costa Rica (LanammeUCR) has developed a methodology that initially assesses the potential risk of accidents associated with a combination of four different parameters related to road infrastructure and the environment. The study was performed in four of the Country’s main highways, for a study length of over 1,000 km of roads. The parameters considered in the methodology were: pavement friction, retro-reflectivity of the road marking, geometrical and topographical alignment of the roadway and climatic factors. The experimental parameters associated with each category were measured directly based on NDT testing. The climatic factors were based on current and historical weather station information. The proposed methodology consists of a combination of values for each individual parameter, which finally result in a susceptibility profile for the road, which is related to the risk that an accident will occur. All of the data was plotted in geo-referenced maps to be available for road users and the government. Finally, the results were correlated with accident data to verify for the sensitivity of the method.
  Authors: Aguiar-Moya, José Pablo; Barrantes-Jimenez, Roy; Sanabria, Jairo; Loria-Salazar, Luis
  Authors: Aguiar-Moya, José Pablo; Barrantes-Jimenez, Roy; Sanabria, Jairo; Loria-Salazar, Luis
  Year: 2013
  Document Type: Paper
  Subject: Data and Information Technology; Safety and Human Factors
  Session: 724
  Paper Number: 13-2286
• Effect of Sun Glare on Traffic Accidents in Japan
  Abstract: This study aims to clarify effect of sun glare on traffic accident occurrence. Traffic accidents analyses were carried out to calculate the position of the sun relative to the first vehicle concerned (i.e., the vehicle most responsible for causing the accident) at the accident time and spot by using the traffic accident database of Chiba Prefecture. Daytime traffic accidents that occurred during fine weather were extracted for analysis. The traffic accident rate was found to increase when the viewing angle decreased to less than 90 degrees. Daytime traffic accidents during fine weather were extracted, and traffic accidents in which this viewing angle was less than 90 degrees were regarded as sun-glare-related ones, and all others were regarded as sun-glare-unrelated ones. Logistic regression analyses were carried out, with the viewing angle as the dependent variable and certain traffic accident data items as the independent variables. When the sun was in front of the first vehicle concerned, the accident rate was much higher for pedestrian accidents, bicycle accidents and accidents at intersection and slightly higher for right-turning accidents and accidents in winter. However, the tendency for vehicle drivers to be affected adversely by sun glare was not observed to increase with increases in vehicle speed. The sun glare tended to cause drivers to not see pedestrians and cyclists at signalized intersections. Traffic safety measures against such kinds of accidents are needed.
  Authors: Hagita, Kenji; Mori, Kenji
  Authors: Hagita, Kenji; Mori, Kenji
  Year: 2013
  Document Type: Paper
  Subject: Data and Information Technology; Safety and Human Factors
  Session: 724
  Paper Number: 13-2388
• Analysis of Factors Affecting Freeway Traffic Crash Frequency Under Different Light Conditions with Random Parameter Count Models
  Abstract: This research develops a random parameter count model of crash frequency on freeways with a speed limit of 110 km/h in Korea and performs a comparison between time periods (daytime, nighttime, twilight, and the whole 24 hour period). Data for crashes in 2007-2010, excluding vehicle factors such as engine overheating and malfunction in damping device and human factors such as drunk driving and dosing off at the wheel, was drawn from Korea freeway crash data. The results show several factors having random effects on crashes: traffic share of light vehicles, number of lanes, urban area, and foggy area. While some factors are statistically significant regardless of the time period (e.g., traffic share of light vehicles, number of lanes, urban area, frequent fog in area, and number of days with snowfall), some factors have statistical effects only during certain time periods (e.g., number of interchanges/junctions and number of bridges during daytime, traffic share of heavy vehicles during nighttime and the whole 24 hour period, and short tangent (<1,421 m) and number of crest vertical curves during twilight). The results indicate that the effect of roadway geometrics on crash frequency differs by time of day which can be used in driver information systems to supply different information to drivers about the road ahead based on time of day. For example, during daytime drivers need more information about upcoming interchanges/junctions. The results indicate that roadway design should try to avoid combining horizontal and sag vertical curves.
  Authors: Hong, Sungmin; Kim, Joon-Ki; Oh, Cheol; Ulfarsson, Gudmundur Freyr
  Authors: Hong, Sungmin; Kim, Joon-Ki; Oh, Cheol; Ulfarsson, Gudmundur Freyr
  Year: 2013
  Document Type: Paper
  Subject: Data and Information Technology; Safety and Human Factors
  Session: 724
  Paper Number: 13-2397
• Modeling Frequency of Traffic Conflicts at Signalized Intersections Using Generalized Linear Regression Models
  Abstract: The primary objective of this study was to identify the potential of using conflict prediction models to predict the frequency of traffic conflicts at signalized intersections. The opposing left-turn conflicts were selected for the development of conflict prediction models. Using data collected at thirty approaches at twenty signalized intersections where the permitted left-turn phases were used, the underlying distributions of the conflict frequency for different volume regimes in different time intervals were examined. It was found that the conflict frequency generally followed a negative binominal distribution. Different conflict prediction models were developed, including a linear regression model, an overall negative binomial model, and separate models developed for four traffic scenarios which were defined based on the volume to capacity ratio of the conflicting traffic flows. The prediction performance of different models was compared. It was found that the linear regression model was not appropriate for modeling the conflict frequency data. In addition, drivers behaved differently under different traffic conditions. Thus, the effects of conflicting traffic volumes on conflict frequency were different in different traffic conditions. The generalized linear regression models developed for different traffic scenarios provided the best estimates for the field measured conflicts.
  Authors: Zhang, Xin; Liu, Pan
  Authors: Zhang, Xin; Liu, Pan
  Year: 2013
  Document Type: Paper
  Subject: Data and Information Technology; Safety and Human Factors
  Session: 724
  Paper Number: 13-2427
• Systematic Approach for Hazardous Intersection Identification and Countermeasure Development
  Abstract: Safety performance functions (SPFs) are typically used to correlate geometric, traffic and environmental characteristics with total crashes and to identify hotspots which have high overall crash frequencies. However, with a distinct conflict pattern in vehicle maneuvers, each crash type is likely to associate with different risk factors. This study developed approach-level SPFs using a full Bayesian method to assess the safe effects of specific risk factors for rear-end, left-turn, right-angle, sideswipe and total crashes. To account for the spatial correlations among approaches at the same intersection, a random intersection-specific effect term was incorporated into each model. It was affirmed that these models were helpful in identifying high risk intersections with specific safety problems, and could serve as useful complements to general hotspot analyses using expected crash totals. In addition, it was found that certain variables (e.g. number of through lanes, median, and left-turn protection all on the entering approach) could have even contrary effects on crash occurrence of different types. Approach-level crash type models provide valuable insights in developing countermeasures aimed at reducing certain crash types and an improved ability in identifying deficiencies related to geometric and traffic characteristics for each intersection approach.
  Authors: Wang, Xuesong; Xie, Kun; Abdel-Aty, Mohamed A.; Tremont, Paul J.; Chen, Xiaohong
  Authors: Wang, Xuesong; Xie, Kun; Abdel-Aty, Mohamed A.; Tremont, Paul J.; Chen, Xiaohong
  Year: 2013
  Document Type: Paper
  Subject: Data and Information Technology; Safety and Human Factors
  Session: 724
  Paper Number: 13-2698
• Hot-Spot Identification: Categorical Binary Model Approach
  Authors: Ferreira, Sara
  Authors: Ferreira, Sara
  Year: 2013
  Document Type: Presentation; Poster
  Subject: Data and Information Technology; Safety and Human Factors
  Session: 724
  Paper Number: 13-0095
• Analyzing Relationship Between Car Generation and Severity of Motor-Vehicle Crashes in Denmark
  Authors: Hels, Tove
  Authors: Hels, Tove
  Year: 2013
  Document Type: Presentation; Poster
  Subject: Data and Information Technology; Safety and Human Factors
  Session: 724
  Paper Number: 13-1246
• Surrogate Safety Measure for Simulation-Based Conflict Study
  Authors: Wang, Chen
  Authors: Wang, Chen
  Year: 2013
  Document Type: Presentation; Poster
  Subject: Data and Information Technology; Safety and Human Factors
  Session: 724
  Paper Number: 13-1512
• Modeling Traffic Accidents on Auckland Motorway, New Zealand
  Authors: Ranjitkar, Prakash
  Authors: Ranjitkar, Prakash
  Year: 2013
  Document Type: Presentation; Poster
  Subject: Data and Information Technology; Safety and Human Factors
  Session: 724
  Paper Number: 13-1976
• Analyzing Crash Severity Based on Vehicle Damage and Occupant Injuries
  Authors: Qin, Xiao
  Authors: Qin, Xiao
  Year: 2013
  Document Type: Presentation; Poster
  Subject: Data and Information Technology; Safety and Human Factors
  Session: 724
  Paper Number: 13-2072
• Effect of Sun Glare on Traffic Accidents in Japan
  Authors: Hagita, Kenji
  Authors: Hagita, Kenji
  Year: 2013
  Document Type: Presentation; Poster
  Subject: Data and Information Technology; Safety and Human Factors
  Session: 724
  Paper Number: 13-2388
• Application of Stochastic Gradient Boosting Technique to Enhance Reliability of Real-Time Risk Assessment Using Automatic Vehicle Identification and Remote Traffic Microwave Sensor Data
  Authors: Ahmed, Mohamed
  Authors: Ahmed, Mohamed
  Year: 2013
  Document Type: Presentation; Poster
  Subject: Data and Information Technology; Safety and Human Factors
  Session: 724
  Paper Number: 13-0410
• Feasibility of Incorporating Reliability Analysis in Traffic Safety Investigation
  Authors: Yu, Rongjie
  Authors: Yu, Rongjie
  Year: 2013
  Document Type: Presentation; Poster
  Subject: Data and Information Technology; Safety and Human Factors
  Session: 724
  Paper Number: 13-0477
• Crash-Type Propensity Analysis with Bayesian Models Using Microscopic Traffic and Weather Data
  Authors: Yu, Rongjie
  Authors: Yu, Rongjie
  Year: 2013
  Document Type: Presentation; Poster
  Subject: Data and Information Technology; Safety and Human Factors
  Session: 724
  Paper Number: 13-0718
• Crash-Type Propensity Analysis with Bayesian Models Using Microscopic Traffic and Weather Data
  Authors: Ahmed, Mohamed
  Authors: Ahmed, Mohamed
  Year: 2013
  Document Type: Presentation; Poster
  Subject: Data and Information Technology; Safety and Human Factors
  Session: 724
  Paper Number: 13-0718
• Crash Fault Analysis of Out-of-State Drivers in Vermont
  Authors: Harootunian, Kristine
  Authors: Harootunian, Kristine
  Year: 2013
  Document Type: Presentation; Poster
  Subject: Data and Information Technology; Safety and Human Factors
  Session: 724
  Paper Number: 13-1597
• Indexing Crashworthiness and Crash Aggressiveness by Major Car Brands
  Authors: Hu, Shuiyan
  Authors: Hu, Shuiyan
  Year: 2013
  Document Type: Presentation; Poster
  Subject: Data and Information Technology; Safety and Human Factors
  Session: 724
  Paper Number: 13-1754
• Urban-Rural Difference of Gasoline Price Effects on Traffic Safety
  Authors: Chi, Guangqing
  Keywords: poster presentation; poster design; poster template
  Authors: Chi, Guangqing
  Year: 2013
  Document Type: Presentation; Poster
  Subject: Data and Information Technology; Safety and Human Factors
  Session: 724
  Paper Number: 13-1688
• Predicting Freeway Crash Likelihood and Severity with Real-Time Loop Detector Data
  Authors: Xu, Chengcheng
  Authors: Xu, Chengcheng
  Year: 2013
  Document Type: Presentation; Poster
  Subject: Data and Information Technology; Safety and Human Factors
  Session: 724
  Paper Number: 13-1715
• Methodology for Determining Traffic Accident Risk Zones
  Authors: Aguiar-Moya, Jose
  Authors: Aguiar-Moya, Jose
  Year: 2013
  Document Type: Presentation; Poster
  Subject: Data and Information Technology; Safety and Human Factors
  Session: 724
  Paper Number: 13-2286
• Analysis of Factors Affecting Freeway Traffic Crash Frequency Under Different Light Conditions with Random Parameter Count Models
  Authors: Oh, Cheol
  Authors: Oh, Cheol
  Year: 2013
  Document Type: Presentation; Poster
  Subject: Data and Information Technology; Safety and Human Factors
  Session: 724
  Paper Number: 13-2397
• Spatial Analysis of Fatal and Injury Crashes in Flanders, Belgium: Application of Geographically Weighted Regression Technique
  Authors: Pirdavani, Ali
  Authors: Pirdavani, Ali
  Year: 2013
  Document Type: Presentation; Poster
  Subject: Data and Information Technology; Safety and Human Factors
  Session: 724
  Paper Number: 13-1049
• Assessment of Exposure Proxies for Macroscopic Road Safety Prediction
  Authors: Antoniou, Constantinos
  Authors: Antoniou, Constantinos
  Year: 2013
  Document Type: Presentation; Poster
  Subject: Data and Information Technology; Safety and Human Factors
  Session: 724
  Paper Number: 13-1788
• Reducing Severity of Crashes During Holidays: Are We Targeting the Right Behaviors?
  Authors: Anowar, Sabreena
  Authors: Anowar, Sabreena
  Year: 2013
  Document Type: Presentation; Poster
  Subject: Data and Information Technology; Safety and Human Factors
  Session: 724
  Paper Number: 13-1095
• Systematic Approach for Hazardous Intersection Identification and Countermeasure Development
  Authors: Wang, Xuesong
  Authors: Wang, Xuesong
  Year: 2013
  Document Type: Presentation; Poster
  Subject: Data and Information Technology; Safety and Human Factors
  Session: 724
  Paper Number: 13-2698
• Redesigning Black Spots in Traffic: Effect Evaluation
  Authors: De Pauw, Ellen
  Authors: De Pauw, Ellen
  Year: 2013
  Document Type: Presentation; Poster
  Subject: Data and Information Technology; Safety and Human Factors
  Session: 724
  Paper Number: 13-1465