Empirical evaluation of alternative approaches in identifying crash hot spots

This study proposes a framework of a model-based hot spot identification method by applying full Bayes (FB) technique. In comparison with the state-of-the-art approach [i.e., empirical Bayes method (EB)], the advantage of the FB method is the capability to seamlessly integrate prior information and all available data into posterior distributions on which various ranking criteria could be based. With intersection crash data collected in Singapore, an empirical analysis was conducted to evaluate the following six approaches for hot spot identification: (a) naive ranking using raw crash data, (b) standard EB ranking, (c) FB ranking using a Poisson-gamma model, (d) FB ranking using a Poisson-lognormal model, (e) FB ranking using a hierarchical Poisson model, and (f) FB ranking using a hierarchical Poisson (AR-1) model. The results show that (a) when using the expected crash rate-related decision parameters, all model-based approaches perform significantly better in safety ranking than does the naive ranking method, and (b) the FB approach using hierarchical models significantly outperforms the standard EB approach in correctly identifying hazardous sites.

Bayesian hierarchical analysis on crash prediction models

Traditional crash prediction models, such as generalized linear regression models, are incapable of taking into account the multilevel data structure, which extensively exists in crash data. Disregarding the possible within-group correlations can lead to the production of models giving unreliable and biased estimates of unknowns. This study innovatively proposes a -level hierarchy, viz. (Geographic region level – Traffic site level – Traffic crash level – Driver-vehicle unit level – Vehicle-occupant level) Time level, to establish a general form of multilevel data structure in traffic safety analysis. To properly model the potential cross-group heterogeneity due to the multilevel data structure, a framework of Bayesian hierarchical models that explicitly specify multilevel structure and correctly yield parameter estimates is introduced and recommended. The proposed method is illustrated in an individual-severity analysis of intersection crashes using the Singapore crash records. This study proved the importance of accounting for the within-group correlations and demonstrated the flexibilities and effectiveness of the Bayesian hierarchical method in modeling multilevel structure of traffic crash data.

Comparison of moped, scooter and motorcycle crash risk and crash severity

The increased popularity of mopeds and motor scooters in Australia and elsewhere in the last decade has contributed substantially to the greater use of powered two-wheelers (PTWs) as a whole. As the exposure of mopeds and scooters has increased, so too has the number of reported crashes involving those PTW types, but there is currently little research comparing the safety of mopeds and, particularly, larger scooters with motorcycles. This study compared the crash risk and crash severity of motorcycles, mopeds and larger scooters in Queensland, Australia. Comprehensive data cleansing was undertaken to separate motorcycles, mopeds and larger scooters in police-reported crash data covering the five years to 30 June 2008. The crash rates of motorcycles (including larger scooters) and mopeds in terms of registered vehicles were similar over this period, although the moped crash rate showed a stronger downward trend. However, the crash rates in terms of distance travelled were nearly four times higher for mopeds than for motorcycles (including larger scooters). More comprehensive distance travelled data is needed to confirm these findings. The overall severity of moped and scooter crashes was significantly lower than motorcycle crashes but an ordered probit regression model showed that crash severity outcomes related to differences in crash characteristics and circumstances, rather than differences between PTW types per se. Greater motorcycle crash severity was associated with higher (>80 km/h) speed zones, horizontal curves, weekend, single vehicle and nighttime crashes. Moped crashes were more severe at night and in speed zones of 90 km/h or more. Larger scooter crashes were more severe in 70 km/h zones (than 60 km/h zones) but not in higher speed zones, and less severe on weekends than on weekdays. The findings can be used to inform potential crash and injury countermeasures tailored to users of different PTW types.

Methodology for developing real-time motorway traffic risk identification models using individual-vehicle data

Most of existing motorway traffic safety studies using disaggregate traffic flow data aim at developing models for identifying real-time traffic risks by comparing pre-crash and non-crash conditions. One of serious shortcomings in those studies is that non-crash conditions are arbitrarily selected and hence, not representative, i.e. selected non-crash data might not be the right data comparable with pre-crash data; the non-crash/pre-crash ratio is arbitrarily decided and neglects the abundance of non-crash over pre-crash conditions; etc. Here, we present a methodology for developing a real-time MotorwaY Traffic Risk Identification Model (MyTRIM) using individual vehicle data, meteorological data, and crash data. Non-crash data are clustered into groups called traffic regimes. Thereafter, pre-crash data are classified into regimes to match with relevant non-crash data. Among totally eight traffic regimes obtained, four highly risky regimes were identified; three regime-based Risk Identification Models (RIM) with sufficient pre-crash data were developed. MyTRIM memorizes the latest risk evolution identified by RIM to predict near future risks. Traffic practitioners can decide MyTRIM’s memory size based on the trade-off between detection and false alarm rates. Decreasing the memory size from 5 to 1 precipitates the increase of detection rate from 65.0% to 100.0% and of false alarm rate from 0.21% to 3.68%. Moreover, critical factors in differentiating pre-crash and non-crash conditions are recognized and usable for developing preventive measures. MyTRIM can be used by practitioners in real-time as an independent tool to make online decision or integrated with existing traffic management systems.

Impact of real-time traffic characteristics on freeway crash occurrence : systematic review and meta-analysis

A systematic literature review and a comprehensive meta-analysis that combines the findings from existing studies, was conducted in this thesis to analyse the impact of traffic characteristics on crash occurrence. Sensitivity analyses were conducted to investigate the quality, publication bias and outlier bias of the various studies, and the time intervals used to measure traffic characteristics were considered. Based on this comprehensive and systematic review, and the results of the subsequent meta-analysis, major issues in study design, traffic and crash data, and model development and evaluation are discussed.

Impact of real-time traffic characteristics on freeway crash occurrence: Systematic review and meta-analysis

The development of methods for real-time crash prediction as a function of current or recent traffic and roadway conditions is gaining increasing attention in the literature. Numerous studies have modeled the relationships between traffic characteristics and crash occurrence, and significant progress has been made. Given the accumulated evidence on this topic and the lack of an articulate summary of research status, challenges, and opportunities, there is an urgent need to scientifically review these studies and to synthesize the existing state-of-the-art knowledge. This paper addresses this need by undertaking a systematic literature review to identify current knowledge, challenges, and opportunities, and then conducts a meta-analysis of existing studies to provide a summary impact of traffic characteristics on crash occurrence. Sensitivity analyses were conducted to assess quality, publication bias, and outlier bias of the various studies; and the time intervals used to measure traffic characteristics were also considered. As a result of this comprehensive and systematic review, issues in study designs, traffic and crash data, and model development and validation are discussed. Outcomes of this study are intended to provide researchers focused on real-time crash prediction with greater insight into the modeling of this important but extremely challenging safety issue.

Crash involvement patterns of unlicensed drivers and riders in Queensland. Report to Queensland Department of Transport and Main Roads

Unlicensed driving remains a serious problem for road safety, despite ongoing improvements in traffic law enforcement practices and technology. While it does not play a direct causative role in road crashes, unlicensed driving undermines the integrity of the driver licensing system and is associated with a range of high-risk behaviours. This report examines official road crash data from Queensland for the years 2003-08 to compare the crash involvement patterns of unlicensed drivers with those of licensed drivers and explore the scope and nature of unlicensed driving. This study replicates and extends upon two previous studies examining the involvement of unlicensed drivers in crashes in Queensland (Watson, 2004a; 2004b; Watson & Steinhardt, 2006).

Heavy vehicle crash characteristics in Oman, 2009-2011

In recent years, Oman has seen a shift in the burden of diseases towards road accidents. The main objective of this paper, therefore, is to describe key characteristics of heavy vehicle crashes in Oman and identify the key driving behaviours that influence fatality risks. Crash data from January 2009 to December 2011 were examined and it was found that of the 22,543 traffic accidents that occurred within this timeframe, 3,114 involved heavy vehicles. While the majority of these crashes were attributed to driver behaviours, a small proportion was attributed to other factors. The results of the study indicate that there is a need for a more thorough crash investigation process in Oman. Future research should explore the reporting processes used by the Royal Oman Police, cultural influences on heavy vehicle operations in Oman, and improvements to the current licensing system.

Links between unlicensed and unregistered driving (Data analysis). Report to Queensland Department of Transport and Main Roads

There is a current lack of understanding regarding the use of unregistered vehicles on public roads and road-related areas, and the links between the driving of unregistered vehicles and a range of dangerous driving behaviours. This report documents the findings of data analysis conducted to investigate the links between unlicensed driving and the driving of unregistered vehicles, and is an important initial undertaking into understanding these behaviours. This report examines de-identified data from two sources: crash data; and offence data. The data was extracted from the Queensland Department of Transport and Main Roads (TMR) databases and covered the period from 2003 to 2008.

International survey of bicycling exposure, crash involvement, behaviors, and attitudes: Preliminary results

This paper presents some results from preliminary analyses of the data of an international online survey of bicycle riders, who reported riding at least once a month. On 4 July 2015, data from 7528 participants from 17 countries was available in the survey, and were subsequently cleaned and checked for consistency. The median distance ridden ranged from 30 km/week in Israel to 150 km/week in Greece (overall median 54 km/week). City/hybrid bicycles were the most common type of bicycle ridden (44%), followed by mountain (20%) and road bikes (15%). Almost half (47%) of the respondents rode “nearly daily”. About a quarter rode daily to work or study (27%). Overall, 40% of respondents reported wearing a helmet ‘always’, varying from 2% in the Netherlands to 80% in Norway, while 25% reported ‘never’ wearing a helmet. Thus, individuals appeared to consistently either use or not use helmets. Helmet wearing rates were generally higher when riding for health/fitness than other purposes and appeared to be little affected by the type of riding location, but some divergences in these patterns were found among countries. Almost 29% of respondents reported being involved in at least one bicycle crash in the last year (ranging from 12% in Israel to 53% in Turkey). Among the most severe crashes for each respondent, about half of the crashes involved falling off a bicycle. Just under 10% of the most severe crashes for each respondent were reported to police. Among the bicycle-motor vehicle crashes, only a third were reported to police. Further analyses will address questions regarding the influence of factors such as demographic characteristics, type of bicycle ridden, and attitudes on both bi-cycle use and helmet wearing rates.

Sleep-related crash characteristics: Implications for applying a fatigue definition to crash reports

Sleep-related (SR) crashes are an endemic problem the world over. However, police officers report difficulties in identifying sleepiness as a crash contributing factor. One approach to improving the sensitivity of SR crash identification is by applying a proxy definition post hoc to crash reports. To identify the prominent characteristics of SR crashes and highlight the influence of proxy definitions, ten years of Queensland (Australia) police reports of crashes occurring in ≥100 km/h speed zones were analysed. In Queensland, two approaches are routinely taken to identifying SR crashes. First, attending police officers identify crash causal factors; one possible option is ‘fatigue/fell asleep’. Second, a proxy definition is applied to all crash reports. Those meeting the definition are considered SR and added to the police-reported SR crashes. Of the 65,204 vehicle operators involved in crashes 3449 were police-reported as SR. Analyses of these data found that male drivers aged 16–24 years within the first two years of unsupervised driving were most likely to have a SR crash. Collision with a stationary object was more likely in SR than in not-SR crashes. Using the proxy definition 9739 (14.9%) crashes were classified as SR. Using the proxy definition removes the findings that SR crashes are more likely to involve males and be of high severity. Additionally, proxy defined SR crashes are no less likely at intersections than not-SR crashes. When interpreting crash data it is important to understand the implications of SR identification because strategies aimed at reducing the road toll are informed by such data. Without the correct interpretation, funding could be misdirected. Improving sleepiness identification should be a priority in terms of both improvement to police and proxy reporting.

Learning from insurance data: Injuries to other road users in motorcyclist at-fault crashes

In multi-vehicle motorcycle crashes, the motorcycle rider is less likely to be at-fault but more commonly severely injured than the other road user. Therefore, not surprisingly, crashes in which motorcycle riders are at-fault and particularly the injuries to the other road users in these crashes have received little research attention. This paper aims to address this gap in the literature by investigating the factors influencing the severity of injury to other road users in motorcyclist-at-fault crashes. Five years of data from Queensland, Australia, were obtained from a database of claims against the compulsory third party (CTP) injury insurance of the at-fault motorcyclists. Analysis of the data using an ordered probit model shows higher injury severity for crashes involving young (under 25) and older (60+) at-fault motorcyclists. Among the not at-fault road users, the young, old, and males were found to be more severely injured than others. Injuries to vehicle occupants were less severe than those to pillions. Crashes that occurred between vehicles traveling in opposite directions resulted in more severe injuries than those involving vehicles traveling in the same direction. While most existing studies have analyzed police reported crash data, this study used CTP insurance data. Comparison of results indicates the potential of using CTP insurance data as an alternative to police reported crash data for gaining a better understanding of risk factors for motorcycle crashes and injury severity.

Secondary Collisions Revisited: Real-World Crash Characteristics and Relationship to Crash Test Criteria

Previous research conducted in the late 1980’s suggested that vehicle impacts following an initial barrier collision increase severe occupant injury risk. Now over twenty-five years old, the data used in the previous research is no longer representative of the currently installed barriers or US vehicle fleet. The purpose of this study is to provide a present-day assessment of secondary collisions and to determine if full-scale barrier crash testing criteria provide an indication of secondary collision risk for real-world barrier crashes. The analysis included 1,383 (596,331 weighted) real-world barrier midsection impacts selected from thirteen years (1997-2009) of in-depth crash data available through the National Automotive Sampling System (NASS) / Crashworthiness Data System (CDS). For each suitable case, the scene diagram and available scene photographs were used to determine roadside and barrier specific variables not available in NASS/CDS. Binary logistic regression models were developed for second event occurrence and resulting driver injury. Barrier lateral stiffness, post-impact vehicle trajectory, vehicle type, and pre-impact tracking conditions were found to be statistically significant contributors toward secondary event occurrence. The presence of a second event was found to increase the likelihood of a serious driver injury by a factor of seven compared to cases with no second event present. Twenty-four full-scale crash test reports were obtained for common non-proprietary US barriers, and the risk of secondary collisions was determined using recommended evaluation criteria from NCHRP Report 350. It was found that the NCHRP Report 350 exit angle criterion alone was not sufficient to predict second collision occurrence for real-world barrier crashes.

Development of safety performance functions for empirical Bayes estimation of crash reduction factors

Crash reduction factors (CRFs) are used to estimate the potential number of traffic crashes expected to be prevented from investment in safety improvement projects. The method used to develop CRFs in Florida has been based on the commonly used before-and-after approach. This approach suffers from a widely recognized problem known as regression-to-the-mean (RTM). The Empirical Bayes (EB) method has been introduced as a means to addressing the RTM problem. This method requires the information from both the treatment and reference sites in order to predict the expected number of crashes had the safety improvement projects at the treatment sites not been implemented. The information from the reference sites is estimated from a safety performance function (SPF), which is a mathematical relationship that links crashes to traffic exposure. The objective of this dissertation was to develop the SPFs for different functional classes of the Florida State Highway System. Crash data from years 2001 through 2003 along with traffic and geometric data were used in the SPF model development. SPFs for both rural and urban roadway categories were developed. The modeling data used were based on one-mile segments that contain homogeneous traffic and geometric conditions within each segment. Segments involving intersections were excluded. The scatter plots of data show that the relationships between crashes and traffic exposure are nonlinear, that crashes increase with traffic exposure in an increasing rate. Four regression models, namely, Poisson (PRM), Negative Binomial (NBRM), zero-inflated Poisson (ZIP), and zero-inflated Negative Binomial (ZINB), were fitted to the one-mile segment records for individual roadway categories. The best model was selected for each category based on a combination of the Likelihood Ratio test, the Vuong statistical test, and the Akaike's Information Criterion (AIC). The NBRM model was found to be appropriate for only one category and the ZINB model was found to be more appropriate for six other categories. The overall results show that the Negative Binomial distribution model generally provides a better fit for the data than the Poisson distribution model. In addition, the ZINB model was found to give the best fit when the count data exhibit excess zeros and over-dispersion for most of the roadway categories. While model validation shows that most data points fall within the 95% prediction intervals of the models developed, the Pearson goodness-of-fit measure does not show statistical significance. This is expected as traffic volume is only one of the many factors contributing to the overall crash experience, and that the SPFs are to be applied in conjunction with Accident Modification Factors (AMFs) to further account for the safety impacts of major geometric features before arriving at the final crash prediction. However, with improved traffic and crash data quality, the crash prediction power of SPF models may be further improved.

Implementation of a driving diary intervention to reduce aberrant driving behaviours

Crash data involving taxis indicates that such drivers are over represented in crashes and are one to two times more likely to be involved in a fatality crash. This study reports on the pre intervention survey to provide a baseline measure of the self-reported attitudes and corresponding driving behaviours of a sample of taxi drivers. Results indicate that some taxi drivers willingly admit to engaging in unsafe driving practices. In addition, preliminary results of a post intervention survey revealed that taxi drivers’ safety perceptions, attitude and behaviours improved after completing a Driving Diary intervention.