Truck Accident Data Research Articles

Analyzing truck accident data on the interurban road Ankara–Aksaray–Eregli in Turkey: Comparing the performances of negative binomial regression and the artificial neural networks models

ABSTRACTStatistical methods such as Poisson distribution, negative binomial regression (NB), and zero inflated negative binomial regression (ZINB) have generally been used in road safety studies to establish the complex relationships between variables. Over the last few years, the artificial neural networks (ANN) model has also been used. The ANN model does not have any prior limitations such as the equality condition of mean and variance observed in Poisson regression. However, though the ANN model has been used in the analysis of different accident types, to the best of our knowledge, no study has used the ANN model for establishing this relationship with truck accident data on divided multilane interurban roads. In this study, the road sections D750/07–D750/15 in Ankara–Aksaray–Eregli, Turkey, were considered and truck accident data from 2008 to 2011 were analyzed using NB and ANN. The analysis show that the ANN model has lower errors and higher R2 values than NB and performs slightly better than NB for predicting the number of trucks involved in accidents. Based on a comparison of performances the study concludes, that ANN could be used as an alternative model for analyzing truck accident data on divided multilane interurban roads in Turkey.

Route evaluation for hazmat transportation based on total risk – A case of Indian State Highways

Risk-based hazmat transportation route evaluation involves risk calculations taking into consideration the probability of collision related accident occurrence and detailed consequence analysis of various event scenarios. Probabilistic hazmat transportation risk assessment mainly depends on three important factors i.e. accident rate, Average Daily Traffic and population density besides route length which has a definite bearing on it. An effort has been made to estimate the route segment specific (location-specific) accident rate instead of aggregate National or State average values in order to bring specificity into the issue of decision making to avoid routes with higher accident rates. Instead of using default accident rate for different highway types developed with the US data, which are not well-comparable when used in Indian situations; the author used site-specific truck accident data. Subsequently, Loss of Containment (LOC) probabilities and spillage probabilities for different route segments have been computed and compared. Finally, route segment-wise total risk is estimated which is a convenient measure of the average number of persons likely to be exposed from all the possible consequence event scenarios resulting from releases of different hazmats being transported along the studied routes. The present study highlights the route evaluation carried out based on total risk computation, without going through detailed event based consequence analysis on two State Highway routes and one major urban road passing through important industrial corridors of Surat District in western India, to enable routing decisions by local authorities and also for planning emergency mitigation purposes.

Modeling vehicle accidents and highway geometric design relationships

The statistical properties of four regression models—two conventional linear regression models and two Poisson regression models—are investigated in terms of their ability to model vehicle accidents and highway geometric design relationships. Potential limitations of these models pertaining to their underlying distributional assumptions, estimation procedures, functional form of accident rate, and sensitivity to short road sections, are identified. Important issues, such as the treatment of vehicle exposure and traffic conditions, and data uncertainties due to sampling and nonsampling errors, are also discussed. Roadway and truck accident data from the Highway Safety Information System (HSIS), a highway safety data base administered by the Federal Highway Administration (FHWA), have been employed to illustrate the use and the limitations of these models. It is demonstrated that the conventional linear regression models lack the distributional property to describe adequately random, discrete, nonnegative, and typically sporadic vehicle accident events on the road. As a result, these models are not appropriate to make probabilistic statements about vehicle accidents, and the test statistics derived from these models are questionable. The Poisson regression models, on the other hand, possess most of the desirable statistical properties in developing the relationships. However, if the vehicle accident data are found to be significantly overdispersed relative to its mean, then using the Poisson regression models may overstate or understate the likelihood of vehicle accidents on the road. More general probability distributions may have to be considered.

Procedure for Developing Truck Accident and Release Rates for Hazmat Routing

To conduct risk assessments in routing studies for the highway transportation of hazardous materials (hazmat), estimates of accident and release rates are essential. Published literature underscores the importance of these rates in risk assessment, and the significant shortcomings of the available data. Truck accident rates are developed in this paper as a function of roadway type and area type (urban/rural) from state data on highway geometrics, traffic volume, and accidents. Release probabilities in accidents are derived from a combination of federal and state truck accident data bases. The procedure is presented in detail. A discussion of the dangers of combining data from different states is also presented, along with justification for the decision of the writers to do so. A revised model for the accident probability portion of the U.S. Department of Transportation (DOT) hazardous materials routing guidelines is recommended. Statistical tests are presented to determine whether accident rates based on site-specific data or systemwide values, such as those derived in this paper, should be used for any particular route segment. Language: en

Characteristics of Hazardous‐Material Accidents in Pennsylvania

Pennsylvania has accounted for the highest number of hazardous material incidents in the United States between 1971 and 1983, with 90% of the incidents being highway related. The Pennsylvania Department of Transportation has sponsored a research project that focuses on the risk assessment associated with the movement of hazardous materials and determination of corresponding vulnerable areas in the state. The risk assessment of hazardous‐material movements is data intensive, and its analysis depends on several data sources and on the quality of the input data. Three data bases are analyzed for this project: (1) The Research and Special Programs Administration (RSPA) hazardous material incident reports (HMIR); (2) the Bureau of Motor Carrier Safety (BMCS) truck accident data base; and (3) the Pennsylvania State traffic accident data base for hazardous material. In addition to the individual analysis, a comparison among these data bases was also made. The comparison was conducted to find out whether the characteristics of accidents and incidents are consistent in different data bases.

Estimating truck accident rate and involvements using linear and Poisson regression models

During the past few years, vehicle miles of travel (VMT) for large trucks having six or more wheels in contact with the road and having a gross weight greater than 10,000 lbs, have been steadily increasing. This has resulted in an increased interaction between large trucks and other vehicles which is being manifested in an increasing rate of fatal accidents involving large trucks. For example, between 1982 and 1984, fatal accident rates for 100 million VMT for large trucks in Virginia, increased by 54% while the rate for other vehicles (passenger cars, vans and pickups) remained approximately constant at less than a 0.3% increase. In order to arrest this increasing trend in large truck fatal accident rates, it is necessary that appropriate counter measures be developed. For this to be done, however, it is necessary to identify those factors that are associated with large truck accidents. A recent analysis of large truck accident data in Virginia, indicated that certain traffic and highway geometric characteristics may be associated with the occurrence of large truck accidents. In order to identify these specific characteristics and determine the extent to which they influence accident rates, a study was conducted with the primary objective of establishing relationships between large truck accidents, and traffic and roadway geometric variables. This paper presents mathematical relationships obtained through multiple linear and Poisson Regression Analyses, relating the number of truck involved accidents per year at a section of highway with traffic and geometric variables. These models indicate that the slope change rate (absolute curve of slope changes in the vertical direction divided by the highway segment), the average daily traffic, the percent of trucks and the difference in speed between trucks and non‐trucks influence the number of truck involved accidents at a given stretch of highway.

Disaggregate model of highway accident occurrence using survival theory

The analysis of discrete accident data and aggregate exposure data frequently necessitates compromises that can obscure the relationship between accident occurrence and potential causal risk components. One way to overcome these difficulties is to develop a model of accident occurrence that includes accident and exposure data at a mathematically consistent disaggregate level. This paper describes the conceptual and mathematical development of such a model using principals of survival theory. The model predicts the probability of being involved in an accident at time t given that a vehicle has survived until that time. Several alternative functional forms are discussed including additive, proportional hazards and accelerated failure time models. Model estimation is discussed for the case in which both accident and nonaccident trips are included and for the case with only accident data. As formulated, the model has the distinct advantage of being able to consider accident and exposure data at a disaggregate level in an entirely consistent analytic framework. A conditional accident analysis is undertaken using truck accident data obtained from a major national carrier in the United States. Model results are interpretable and generally reasonable. Of particular interest is that segmenting accidents in several categories yields very different sets of significant parameters. Driver service hours seemed to most strongly effect accident risk: regularly scheduled drivers who take frequent trips are likely to have a reduced risk of an accident, particularly if they have a longer (greater than eight) number of hours off-duty just prior to a trip.

Analysis of Heavy Truck Accident Data—Exposure Based Approach

The measure used in most accident studies is as follows: Accident Rate = number of accidents divided by the vehicle miles travel. However, methodological problems would arise if the same measure is used in comparing accident data for different vehicle categroies, e.g. heavy trucks versus light trucks. This article addresses this issue, and presents a procedure for analyzing accident data involving trucks of varying sizes. The procedure developed was used to analyze truck and passenger car accidents in Michigan. It can also be used to assess the relative role of any two or three vehicular groups. The procedure is a modified version of an exposure-based method used by the principal author in an earlier study. It is used to conduct analyses of accident data involving heavy trucks. The procedure appears to be suitalble for analyzing heavy truck accident data and for the most part, lends itself to the use of data commonly available from state transportation agencies in the U.S. Department of Transportation and the U.S. Bureau of Census.

How reliable is the bureau of motor carrier safety truck accident data?

AbstractThis paper examines the degree of compliance with the accident reporting requirements imposed by the Bureau of Motor Carrier Safety (BMCS) of the U.S. Department of Transportation on motor carriers involved in interstate and foreign transport. It also examines the reliability and validity of the accident information reported by these carriers to the BMCS. The study shows that there is a high degree of accident underreporting, especially by the private and exempt carriers, and that the information reported on variables such as cargo weight, gross vehicle weights, and usage of seat belts at the time of accident, is biased toward the interests of the reporters of the accidents.

Analysis of truck accident rates using loglinear models

Alternative approaches for the analysis of factors affecting truck accident rates are assessed. The scope of the discussion is methodological in nature, with emphasis on exploring the limitations of each approach. Several important statistical concerns associated with the weighted least squares algorithm for calibrating loglinear models are addressed, including reduced cell membership in the contingency tables of accidents, scaling factors for measuring truck exposure, incompatibility between continuous exposure variables and categorical accident data, and sensitivity of calibrated coefficients to changes in accident characteristics for different data sets. An alternative approach for fitting loglinear models is proposed, which makes use of the generalized linear interactive models (GLIM) package. This approach assumes that the dependent variables in the contingency table of accidents behave in a Poisson-like process with values ranging from 0 to infinity. The calibration of the beta parameters in the loglinear expressions uses maximum likelihood techniques. Continuous exposure variates are incorporated into these models as offsets. As in the classical weighted least squares algorithm, this approach permits a stepwise statistical analysis of higher-order interactions for categorical accident frequency data, while adjusting directly for continuous measures of exposure. The results of a calibration of loglinear expressions for Ontario truck accident data are presented. Key words: truck accident rate, exposure, loglinear calibration, causal interaction.