Does non-motorists' safety perception of autonomous vehicles vary across opinion change stemming from crash occurrence? Investigating perceptions using fixed and random parameter ordered logit models.

The perception of non-motorists toward autonomous vehicles (AVs) could change or remain the same after hearing about a fatal AV crash. This study aims to discern the differences between non-motorists whose opinions were influenced and those whose views remained unchanged after hearing about the fatal AV crash. Additionally, this study investigates how the operational competence of AV manufacturers affect non-motorist's safety perception. A fixed parameter ordered logit model was explored to explain non-motorists' safety perception for respondents who changed their opinion after a crash occurrence and those who did not. A random parameter ordered logit model was then explored to address the difference in residual variance. The finding from the random parameter ordered logit model indicates that two factors (i.e., future impact of AVs on traffic injuries and willingness to adopt Pittsburgh's streets as a proving ground) were significant for the respondents with negative change in opinion after the crash. The result also suggests that negative information such as the crash in Arizona may influence non-motorists' safety perception of AVs, but this effect can be mitigated by providing positive information about the AVs. The results also demonstrate that non-motorists who perceive the operational competence of the AV manufacturers to be high have higher safety perception of the AVs, while those who perceive the operational competence of the AV manufacturers to be low have lower safety perception of the AVs. This study offers recommendations for policies and interventions to enhance non-motorists' safety based on the examination of factors associated with safety perception among respondents who changed their opinion following a crash and those who maintained their initial stance.

Exploring the effect of road network, demographic, and land use characteristics on teen crash frequency using geographically weighted negative binomial regression.

Motor vehicle crashes are one of the leading causes of teen deaths worldwide. It is important to assess the environment and identify the risk factors influencing teen crashes for planning strategies and improving their safety. This research, therefore, focuses on exploring the effect of road network, demographic, and land use characteristics to compute teen crash frequency. Data for 201 spatially distributed road segments in Mecklenburg County, North Carolina, USA were considered for the evaluation. Data related to teen crashes were obtained from the Highway Safety Information System (HSIS). Demographic and land use characteristics were extracted around each selected road segment using two different buffer widths (0.25 miles and 0.5 miles). Teen crash frequency of each road segment was used as the dependent variable. The generalized linear models with the negative binomial distribution (GLM-based NB model) and geographically weighted negative binomial regression models (GWNBR and GWNBRg) were developed and compared. The annual average daily traffic (AADT), light commercial land use, number of household units, and number of pupils enrolled in public or private high schools are significant explanatory variables influencing teen crash frequency. Both methods have good predictive capabilities and can be used to compute teen crash frequency. However, the GWNBR and GWNBRg better capture the spatial dependency and spatial heterogeneity of associated risk factors influencing teen crash frequency.

Investigating the influence of contributing factors and predicting visibility at road link-level.

Data from weather stations at airports, far away locations or predictions using macro-level data may not be accurate enough to disseminate visibility related information to motorists in advance. Therefore, the objective of this research is to investigate the influence of contributing factors and develop visibility prediction models, at road link-level, by considering data from weather stations located within 1.6 km of state routes, US routes and interstates in the state of North Carolina (NC). Four years of meteorological data, from January 2011 to December 2014, were collected within NC. Ordinary least squares (OLS) and weighted least squares (WLS) regression models were developed for different visibility and elevation ranges. The results indicate that elevation and cloud cover are negatively associated with low visibility. The chances of low visibility are higher between six to twelve hours after rainfall when compared to the first six hours after rainfall. A visibility sensor was installed at four different locations in NC to compare hourly visibility from the selected regression model, High-Resolution Rapid Refresh (HRRR) data, and the nearest weather station. The results indicate that the number of samples with zero error range was higher for the selected regression model compared with the HRRR and weather station observations.

Influence of on-network, traffic, signal, demographic, and land use characteristics by area type on red light violation crashes.

The focus of this research paper is on extraction of predictor variables pertaining to on-network, traffic, signal, demographic, and land use characteristics, by area type, and examining their influence on the number of red light violation crashes. Data for the city of Charlotte, North Carolina was extracted and used for analysis. Three different sets of signalized intersections were selected in the three different area types - Central Business District (CBD), urban, and suburban areas. Each set is comprised of sixty signalized intersections (total 180 signalized intersections). The number of red light violation crashes from January 2010 to December 2014, within the vicinity of each selected signalized intersection, was considered as the dependent variable to develop crash estimation models for each area type. The crash estimation models by area type were compared with the crash estimation model developed considering all the 180 signalized intersections together. Different predictor variables were found to be significant at a 95% confidence level in three different areas. Log-link model with Negative Binomial distribution was observed to best fit the data used in this research. Findings indicate that enforcement, either manually or using red light running cameras (RLCs), at signalized intersections with high traffic volume in the CBD area; at signalized intersections with high traffic volume, high all-red clearance time, near high density of horizontal mixed non-residential and open space/recreational type land uses in urban area; at signalized intersections with high traffic volume, speed limit on the major approach, the number of lanes on the minor approach, and all-red clearance time and areas surrounded with horizontal mixed non-residential and retail type land use in suburban areas, would lead to a reduction in the number of red light violation crashes.

Modeling and comparing injury severity of at-fault and not at-fault drivers in crashes.

This paper examines and compares the effect of selected variables on driver injury severity of, both, at-fault and not at-fault drivers. Data from the Highway Safety Information System (HSIS) for the state of North Carolina was used for analysis and modeling. A partial proportional odds model was developed to examine the effect of each variable on injury severity of at-fault driver and not at-fault driver, and, to examine how each variable affects these two drivers' injury severity differently. Road characteristics, weather condition, and geometric characteristics were observed to have a similar effect on injury severity in a crash to at-fault and not at-fault drivers. Age of the driver, physical condition, gender, vehicle type, and, the number and type of traffic rule violations were observed to play a significant role in the injury severity of not at-fault drivers when compared to at-fault drivers in the crash. Moreover, motorcyclists and drivers 70 years or older are observed to be the most vulnerable road users.

Predictability and interpretability of hybrid link-level crash frequency models for urban arterials compared to cluster-based and general negative binomial regression models.

Machine learning (ML) techniques have higher prediction accuracy compared to conventional statistical methods for crash frequency modelling. However, their black-box nature limits the interpretability. The objective of this research is to combine both ML and statistical methods to develop hybrid link-level crash frequency models with high predictability and interpretability. For this purpose, M5' model trees method (M5') is introduced and applied to classify the crash data and then calibrate a model for each homogenous class. The data for 1134 and 345 randomly selected links on urban arterials in the city of Charlotte, North Carolina was used to develop and validate models, respectively. The outputs from the hybrid approach are compared with the outputs from cluster-based negative binomial regression (NBR) and general NBR models. Findings indicate that M5' has high predictability and is very reliable to interpret the role of different attributes on crash frequency compared to other developed models.

Modeling crash injury severity by road feature to improve safety.

OBJECTIVE: The objective of this research is 2-fold: to (a) model and identify critical road features (or locations) based on crash injury severity and compare it with crash frequency and (b) model and identify drivers who are more likely to contribute to crashes by road feature. METHOD: Crash data from 2011 to 2013 were obtained from the Highway Safety Information System (HSIS) for the state of North Carolina. Twenty-three different road features were considered, analyzed, and compared with each other as well as no road feature. A multinomial logit (MNL) model was developed and odds ratios were estimated to investigate the effect of road features on crash injury severity. RESULTS: Among the many road features, underpass, end or beginning of a divided highway, and on-ramp terminal on crossroad are the top 3 critical road features. Intersection crashes are frequent but are not highly likely to result in severe injuries compared to critical road features. Roundabouts are least likely to result in both severe and moderate injuries. Female drivers are more likely to be involved in crashes at intersections (4-way and T) compared to male drivers. Adult drivers are more likely to be involved in crashes at underpasses. Older drivers are 1.6 times more likely to be involved in a crash at the end or beginning of a divided highway. CONCLUSIONS: The findings from this research help to identify critical road features that need to be given priority. As an example, additional advanced warning signs and providing enlarged or highly retroreflective signs that grab the attention of older drivers may help in making locations such as end or beginning of a divided highway much safer. Educating drivers about the necessary skill sets required at critical road features in addition to engineering solutions may further help them adopt safe driving behaviors on the road.

Methods to rank traffic rule violations resulting in crashes for allocation of funds.

Education, enforcement and engineering countermeasures are implemented to make road users comply with the traffic rules. Not all the traffic rule violations can be addressed nor countermeasures be implemented at all unsafe locations, at once, due to limited funds. Therefore, this study aims at ranking the traffic rule violations resulting in crashes based on individual ranks, such as 1) frequency (expressed as a function of the number of drivers violating a traffic rule and involved in crashes), 2) crash severity, 3) total crash cost, and, 4) cost severity index, to assist transportation system managers in prioritizing the allocation of funds and improving safety on roads. Crash data gathered for the state of North Carolina was processed and used in this study. Variations in the ranks of traffic rule violations were observed when individual ranking methods are used. As an example, exceeding authorized speed limit and driving under the influence of alcohol are ranked 1st and 2nd based on crash severity while failure to reduce speed and failure to yield the right-of-way are ranked 1st and 2nd based on frequency. To minimize the variations and capture the merits of individual ranking methods, four different composite ranks were computed by combining selected individual ranks. The computed averages and standard deviations of absolute rank differences between composite ranks is lower than those obtained from individual ranks. The weights to combine the selected individual ranks have a marginal effect on the computed averages and standard deviations of absolute rank differences. Combining frequency and crash severity or cost severity index, using equal weights, is recommended for prioritization and allocation of funds.

Risk drivers pose to themselves and other drivers by violating traffic rules.

OBJECTIVE: Violation of traffic rules is a major contributing factor in both crashes and fatalities in the United States. This study aims at quantifying risk that drivers pose to themselves and other drivers by violating traffic rules. METHOD: Crash data from 2010 to 2013 were gathered for the state of North Carolina. Descriptive analysis was carried out to identify frequent traffic violations and who were committing the traffic violations that resulted in crashes. A multinomial logit model was then developed to examine the relation between different traffic violations and driver injury severity. Additionally, odds ratios were estimated to identify the likelihood (probability) of severe or moderate injury to the driver and other drivers due to a driver violating a traffic rule that led to a crash. RESULTS: Exceeding the speed limit is more likely to result in severe injury compared to disregarding traffic signals. However, going the wrong way is more likely to result in severe injury to other drivers when compared to any other traffic violation. Driving under the influence of alcohol is 2 times more likely to result in severe injury than driving under the influence of drugs. These 2 traffic violations by a driver are almost equally likely to result in severe injury to other drivers. CONCLUSIONS: Drivers often perceive that violating traffic rules will not result in a crash or severe injury. However, the results from this study show that a majority of the traffic violations lead to severe injury to the violator as well as to other drivers. The findings from this study serve as documented evidence to educate drivers about the risk they pose to themselves and to other drivers by violating traffic rules and encourage the adaptation of safe driving behavior in order to contribute toward reaching the "zero traffic deaths" vision. They also help make policy changes pertaining to penalty points and fines for violating a traffic rule.

Evaluating the effectiveness of on-street bicycle lane and assessing risk to bicyclists in Charlotte, North Carolina.

The objectives of this manuscript are (1) to evaluate the effectiveness of on-street bicycle lane in reducing crashes involving bicyclists on urban roads, (2) to quantify and compare risk to bicyclists on road segments with and without on-street bicycle lane, (3) to evaluate the effect of on-street bicycle lane on other road network users (all crashes), and, (4) to assess the role of on-network characteristics (speed limit, the number of lanes, the width of on-street bicycle lane, the width of the right-most travel lane, and, the numbers of driveways, unsignalized approaches and signalized intersections per unit distance) on risk to bicyclists. Data for thirty-six segments with on-street bicycle lane and twenty-six segments without on-street bicycle lane in the city of Charlotte, North Carolina were extracted to compute and compare measures such as the number of bicycle crashes per center-lane mile, the number of bicycle crashes per annual million vehicle miles traveled (MVMT), the number of all crashes per center-lane mile, and the number of all crashes per MVMT. The results obtained from analysis indicate that bicyclists are three to four times at higher risk (based on traffic conditions) on segments without on-street bicycle lane than when compared to segments with on-street bicycle lane. An analysis conducted considering all crashes showed that on-street bicycle lanes do not have a statistically significant negative effect on overall safety. An increase in annual MVMT (exposure) and the number of signalized intersections per mile increases the number of bicycle crashes, while an increase in on-street bicycle lane width or right-most travel lane width (if on-street bicycle lane cannot be provided) decreases the number of bicycle crashes. Installing wider on-street bicycle lanes, limiting driveways to less than 50 per mile and unsignalized approaches to less than 10 per mile, increasing spacing between signalized intersections, and, facilitating wider right-most travel lane if on-street bicycle lane cannot be provided reduces occurrence of bicycle crashes and lowers risk to bicyclists on roads.

Pedestrian and motorists' actions at pedestrian hybrid beacon sites: findings from a pilot study.

This paper focuses on an analysis of pedestrian and motorists' actions at sites with pedestrian hybrid beacons and assesses their effectiveness in improving the safety of pedestrians. Descriptive and statistical analyses (one-tail two-sample T-test and two-proportion Z-test) were conducted using field data collected during morning and evening peak hours at three study sites in the city of Charlotte, NC, before and after the installation of pedestrian hybrid beacons. Further, an analysis was conducted to assess the change in pedestrian and motorists' actions over time (before the installation; 1 month, 3 months, 6 months, and 12 months after the installation). Results showed an increase in average traffic speed at one of the pedestrian hybrid beacon sites while no specific trends were observed at the other two pedestrian hybrid beacon sites. A decrease in the number of motorists not yielding to pedestrians, pedestrians trapped in the middle of the street, and pedestrian-vehicle conflicts were observed at all the three pedestrian hybrid beacon sites. The installation of pedestrian hybrid beacons did not have a negative effect on pedestrian actions at two out of the three sites. Improvements seem to be relatively more consistent 3 months after the installation of the pedestrian hybrid beacon.

Assessing the effect of introducing a permitted phase through the use of flashing yellow arrow signal for left-turning vehicles.

This paper evaluates the direct and indirect effects of introducing a permitted phase, through the use of flashing yellow arrow (FYA) signal for left-turning vehicles, in reducing crashes at intersections. Data for 18 study intersections in the city of Charlotte, NC, USA were used to conduct a before-after comparison study through the use of Empirical Bayes (EB) method and examine the effects. The estimated number of left-turn crashes, had the FYA signal not been installed, was compared to the actual number of left-turn crashes to assess the direct effect, while the estimated total number of crashes, had the FYA signal not been installed, was compared to the actual total number of crashes to assess the indirect effect. Only left-turn crashes along a selected FYA leg were used to examine the direct effect as the number of legs (approaches) with the FYA signal varied between the selected study intersections. The results obtained indicate that the FYA signal helps reduce the left-turn crashes (direct effect). It does not lead to any negative consequences. Instead, the FYA signal has the potential to indirectly lower the total number of crashes (indirect effect) and contribute to improved safety at intersections.

Effectiveness of red light running camera enforcement program in reducing crashes: evaluation using "before the installation", "after the installation", and "after the termination" data.

The focus of this manuscript is to evaluate and assess the effectiveness of red light running camera (RLC) enforcement program in reducing crashes at signalized intersections. Data from January 1997 to December 2010 for thirty-two signalized intersections in the city of Charlotte, North Carolina, where RLCs were installed between August 1998 and August 2000 and terminated in fall 2006, were gathered, analyzed, and compared for "before the installation", "after the installation", and "after the termination" periods. Descriptive analysis and paired t-tests were performed using rear-end, sideswipe, left-turn, angle, and right-turn crashes as well as the number of total crashes. The expected number of total crashes, had RLC enforcement program not been implemented, was computed using the empirical bayes (EB) method and compared to the actual number of total crashes for "after the installation" and "after the termination" periods. Results obtained indicate that RLC enforcement program leads to an increase in sideswipe and rear-end crashes at ≥ 50% of the signalized intersections. It is effective in reducing total crashes at 50% and 16% of the thirty-two signalized intersections when analyzed considering "before the installation - after the installation" and "before the installation - after the termination" scenarios, respectively. Benefits due to reduction in the number of total crashes may be higher if RLC enforcement program is implemented at signalized intersections with (1) total entering vehicles per day less than 40,000, (2) fewer than 20 rear-end crashes per year, or (3) fewer than 5 sideswipe crashes per year.

Traffic analysis zone level crash estimation models based on land use characteristics.

The objective of this paper is to develop crash estimation models at traffic analysis zone (TAZ) level as a function of land use characteristics. Crash data and land use data for the City of Charlotte, Mecklenburg County, North Carolina were used to illustrate the development of TAZ level crash estimation models. Negative binomial count models (with log-link) were developed as data was observed to be over-dispersed. Demographic/socio-economic characteristics such as population, the number of household units and employment, traffic indicators such as trip productions and attractions, and, on-network characteristics such as center-lane miles by speed limit were observed to be correlated to land use characteristics, and, hence were not considered in the development of TAZ level crash estimation models. Urban residential commercial, rural district and mixed use district land use variables were observed to be correlated to other land use variables and were also not considered in the development of the models. Results obtained indicate that land use characteristics such as mixed use development, urban residential, single-family residential, multi-family residential, business and, office district are strongly associated and play a statistically significant role in estimating TAZ level crashes. The coefficient for single-family residential area was observed to be negative, indicating a decrease in the number of crashes with an increase in single-family residential area. Models were also developed to estimate these crashes by severity (injury and property damage only crashes). The outcomes can be used in safety conscious planning, land use decisions, long range transportation plans, and, to proactively apply safety treatments in high risk TAZs.

Pedestrian crash estimation models for signalized intersections.

The focus of this paper is twofold: (1) to examine the non-linear relationship between pedestrian crashes and predictor variables such as demographic characteristics (population and household units), socio-economic characteristics (mean income and total employment), land use characteristics, road network characteristics (the number of lanes, speed limit, presence of median, and pedestrian and vehicular volume) and accessibility to public transit systems, and (2) to develop generalized linear pedestrian crash estimation models (based on negative binomial distribution to accommodate for over-dispersion of data) by the level of pedestrian activity and spatial proximity to extract site specific data at signalized intersections. Data for 176 randomly selected signalized intersections in the City of Charlotte, North Carolina were used to examine the non-linear relationships and develop pedestrian crash estimation models. The average number of pedestrian crashes per year within 200 feet of each intersection was considered as the dependent variable whereas the demographic characteristics, socio-economic characteristics, land use characteristics, road network characteristics and the number of transit stops were considered as the predictor variables. The Pearson correlation coefficient was used to eliminate predictor variables that were correlated to each other. Models were then developed separately for all signalized intersections, high pedestrian activity signalized intersections and low pedestrian activity signalized intersections. The use of 0.25mile, 0.5mile and 1mile buffer widths to extract data and develop models was also evaluated.

Are pedestrian countdown signals effective in reducing crashes?

OBJECTIVE: The time left to cross the street displayed on pedestrian countdown signals can be used by pedestrians as well as drivers of vehicles, though these signals are primarily provided to help pedestrians make better crossing decisions at signalized intersections. This article presents an evaluation of the effect of pedestrian countdown signals in reducing vehicle-pedestrian crashes and all crashes at signalized intersections. METHODS: A before-and-after study approach was adopted to evaluate the effect considering pedestrian countdown signals installed over a 5-month period at 106 signalized intersections in the city of Charlotte, North Carolina. RESULTS: Analysis conducted at 95 percent confidence level showed that there has been a statistically insignificant decrease in vehicle-pedestrian crashes but a statistically significant decrease in all (includes vehicle-pedestrian and vehicle(s) only involved) crashes after the installation of pedestrian countdown signals. No negative consequences were observed after the installation of pedestrian countdown signals. Sixty-eight percent of the signalized intersections saw a decrease in the total number of all crashes, and 4 percent of the signalized intersections have not seen any change in the number of all crashes after the installation of pedestrian countdown signals. Improvements in terms of decrease in the total number of all crashes was high at signalized intersections with greater than 10 crashes per year during the before period. Likewise, decrease in the number of all crashes was high at signalized intersections with traffic volume between 7 AM to 7 PM greater than 20,000 vehicles during the before period. CONCLUSIONS: Based on results obtained, it can be concluded that pedestrians as well as drivers are making better decisions using the time left to cross the street displayed on pedestrian countdown signals at signalized intersections in the city of Charlotte, North Carolina.

Motorist actions at a crosswalk with an in-pavement flashing light system.

OBJECTIVE: An in-pavement flashing light system is used at crosswalks to alert motorists and pedestrians of possible conflicts and to influence their behavior to enhance safety. The relative behaviors of the drivers and the pedestrians affect safety. An evaluation of motorist behavior at a pedestrian crosswalk with an in-pavement flashing light system is presented in this manuscript. METHODS: Field observations provide the basis to evaluate motorist behavior at a crosswalk with an in-pavement flashing light system. Outcomes of pedestrian and motorists actions were observed to quantify measures of effectiveness (MOEs) such as yielding behavior of motorists, vehicle speeds, and yielding distance from the crosswalk. A before-and-after study design was used. The before condition was prior to the activation of the in-pavement flashing light system and the after condition was after the activation of the in-pavement flashing light system. The study was conducted on a relatively low-volume roadway located in the Henderson, Nevada. The significance of the differences in the MOEs between the 2 study periods was evaluated using statistical analysis tools such as a one-tailed test for proportions and the Welch-Satterthwaite t-test. RESULTS: The results show that the installation of the in-pavement flashing light system increased the yielding behavior of motorists significantly (P < 0.001). The vehicular speeds decreased when pedestrians were waiting at the curb to cross and when they were crossing (P < 0.001). Motorists yielded to pedestrians on an average about 3 m (∼10 feet) upstream from the yield markings and the yielding distances were consistent in both directions. CONCLUSION: The in-pavement flashing light system is seen to be effective to improve motorists' yielding behavior and the speeds of vehicles were also observed to decrease in the presence of pedestrians.

Assessing the role of pavement macrotexture in preventing crashes on highways.

OBJECTIVE: The objective of this article is to assess the role of pavement macrotexture in preventing crashes on highways in the State of North Carolina. METHODS: Laser profilometer data obtained from the North Carolina Department of Transportation (NCDOT) for highways comprising four corridors are processed to calculate pavement macrotexture at 100-m (approximately 330-ft) sections according to the American Society for Testing and Materials (ASTM) standards. Crash data collected over the same lengths of the corridors were integrated with the calculated pavement macrotexture for each section. Scatterplots were generated to assess the role of pavement macrotexture on crashes and logarithm of crashes. Regression analyses were conducted by considering predictor variables such as million vehicle miles of travel (as a function of traffic volume and length), the number of interchanges, the number of at-grade intersections, the number of grade-separated interchanges, and the number of bridges, culverts, and overhead signs along with pavement macrotexture to study the statistical significance of relationship between pavement macrotexture and crashes (both linear and log-linear) when compared to other predictor variables. RESULTS: Scatterplots and regression analysis conducted indicate a more statistically significant relationship between pavement macrotexture and logarithm of crashes than between pavement macrotexture and crashes. The coefficient for pavement macrotexture, in general, is negative, indicating that the number of crashes or logarithm of crashes decreases as it increases. The relation between pavement macrotexture and logarithm of crashes is generally stronger than between most other predictor variables and crashes or logarithm of crashes. CONCLUSIONS: Based on results obtained, it can be concluded that maintaining pavement macrotexture greater than or equal to 1.524 mm (0.06 in.) as a threshold limit would possibly reduce crashes and provide safe transportation to road users on highways.

Evaluating the role of weaving section characteristics and traffic on crashes in weaving areas.

OBJECTIVE: The likelihood of being involved in a crash on a freeway, in general, is greater on weaving sections than on basic freeway sections and in ramp influence areas. This is due to possible crossing of entry and/or exit traffic over a short distance while traveling in the same direction without the aid of traffic control devices resulting in potential conflicting situations and crashes. This article focuses on evaluating the role of weaving section characteristics (configuration type, length and the number of required lane changes by weaving traffic) and traffic variables (entry volume, exit volume, and non-weaving volume) on crashes in weaving areas. METHODS: Data collected for 25 weaving sections in the Las Vegas metropolitan area are used to study the relationship between crashes and weaving section characteristics and traffic variables. The relationship between (1) crashes by selected collision types and contributing factors and (2) weaving section characteristics and traffic variables are also examined. Descriptive and statistical analysis techniques were used. RESULTS: The number of crashes tends to decrease with increase in length of weaving section. Increase in entry volume increases crashes due to improper lane change and ran off roadway crashes. On the other hand, increase in exit volume increases rear-end crashes, crashes due to following too closely, and crashes due to inattentive driving. Non-weaving volume, in general, also appears to play a prominent role in explaining most crash types and contributing factors on weaving sections. CONCLUSIONS: Type "A" weaving section tends to be relatively safer when compared to other weaving configuration section types. Retrofitting weaving sections with short lengths or designing sections with longer lengths, in addition to increased use of in-vehicle warning systems, may reduce crashes and improve safety. Ramp metering, enforcement, changeable message signs (speed signs), and capacity improvements are other solutions for consideration.

New methods to identify and rank high pedestrian crash zones: an illustration.

Identifying and ranking high pedestrian crash zones plays a key role in developing efficient and effective strategies to enhance pedestrian safety. This paper presents (1) a Geographical Information Systems (GIS) methodology to study the spatial patterns of pedestrian crashes in order to identify high pedestrian crash zones, and (2) an evaluation of methods to rank these high pedestrian crash zones. The GIS based methodology to identify high pedestrian crash zones includes geocoding crash data, creating crash concentration maps, and then identifying high pedestrian crash zones. Two methods generally used to create crash concentration maps based on density values are the Simple Method and the Kernel Method. Ranking methods such as crash frequency, crash density, and crash rate, as well as composite methods such as the sum-of-the-ranks and the crash score methods are used to rank the selected high pedestrian crash zones. The use of this methodology and ranking methods for high pedestrian crash zones are illustrated using the Las Vegas metropolitan area as the study area. Crash data collected for a 5-year period (1998-2002) were address matched using the street name/reference street name intersection location reference system. A crash concentration map was then created using the Kernel Method as it facilitates the creation of a smooth density surface when compared to the Simple Method. Twenty-two linear high crash zones and seven circular high crash zones were then identified. The GIS based methodology reduced the subjectivity in the analysis process. Results obtained from the evaluation of methods to rank high pedestrian crash zones show a significant variation in ranking when individual methods were considered. However, rankings of high pedestrian crash zones were relatively consistent with little to no variation when the sum-of-the-ranks method and the crash score method were used. Thus, these composite methods are recommended for use in ranking high pedestrian crash zones instead of individual methods.