This paper presents an overview of current issues in road safety, describes some promising areas of research and provides an introduction for the papers contained in this special issue of the Journal of Transportation Engineering.

Our road safety future is shaped by decisions that affect the amount of trip making, mode of travel used, kinds of infrastructure on which travel takes place, vehicle fleet, technology in use, and the prevailing norms of behavior. While in the past most such decisions were based on intuition and judgement, there is an obvious trend toward decisions based on fact and science. This transition from a “pragmatic” to a more “rational” style of road safety management is hungry for factual knowledge and for professionals to be its purveyors. Consequently, a broad class of professionals, those who influence the future of road safety, needs to be trained in what fact-based road safety knowledge exists. In addition, a vibrant, competent community of road-safety researchers has to be created. They need to be trained in the same road safety knowledge as well as in research methods. Above all they need to be freed from the constraints imposed on them by a myopic class of research administrators. The best interest of society is to move toward the gradual establishment of the rational style of road safety management; it is the engineer’s professional obligation to promote this societal interest.

There is a large and growing literature on road safety. Unfortunately, not all that is published is sound. This puts on the reader the onus of deciding what to believe. To help the reader, three common sources of error are discussed in this article. However, one cannot make the reader into an expert on research method. The main remedy is in improved peer review.

A critical part of any risk assessment is identifying how to represent exposure to the risk involved. Recent research shows that the relationship between crash count and traffic volume is nonlinear; consequently, a simple crash rate computed as the ratio of crash count to volume is not suitable for comparing the safety of sites with different traffic volumes. To solve this problem, we describe a new approach for relating traffic volume and crash incidence. Specifically, we disaggregate crashes into four types: (1) single-vehicle, (2) multivehicle same direction, (3) multivehicle opposite direction, and (4) multivehicle intersecting, and then define candidate exposure measures for each (as a function of site traffic volumes) that we hypothesize will be linear with respect to each crash type. This article describes investigation using crash and physical characteristics data for highway segments from Michigan, California, Washington, and Illinois obtained from the Highway Safety Information System. We have used a hierarchical Bayesian framework to fit zero-inflated-Poisson regression models for predicting counts for each of the above crash types as a function of the daily volume, segment length, speed limit and lane/shoulder width using Markov Chain Monte Carlo methods. We found that the relationship between crashes and the daily volume is nonlinear and varies by crash type, and is significantly different from the relationship between crashes and segment length for all crash types. Significant differences in exposure functions by crash type are proven using analysis of variance and Tukey tests.

In the Midwest, three states in recent years have re-examined their grade crossing model. Illinois conducted a study that developed a new model, the expected accident frequency formula, using a nonlinear regression analysis procedure. Kansas also recently re-examined their model and replaced their design hazard rating formula with a simpler one developed for New Hampshire. Finally, the Missouri Department of Transportation began a similar assessment in 2001. In this paper, we present a framework for assessing grade-crossing models and selecting a preferred model. The methodology consists of five steps that can be adapted to fit most any situation. After presenting this framework, we apply it to Missouri’s evaluation of its exposure index. A major strength of the five-step approach is that it brings together key stakeholders and involves them in the process of establishing characteristics of the “best” model. This dialog strengthens the credibility of any decision to keep or revise a state’s grade crossing model. We are confident that the five-step framework used in this study can be used by any agency seeking to upgrade its grade crossing model.

Unlike conventional traffic safety studies that focused on histrionic data analyses, this study attempts to identify traffic conditions that might lead to a traffic accident from real-time freeway traffic data. An innovative feature of the study is to apply the concept, real-time and preaccident, to accident studies by integrating real-time capabilities in advanced traffic management and information systems (ATMIS). In this study, the traffic conditions leading to more accidents are defined as real-time accident likelihood, and the accident likelihood is estimated by employing a nonparametric Bayesian model. The main goal of the study is to remove hazardous traffic condition prior to accident occurrence by incorporating the real-time accident likelihood into ATMIS. This study estimates real-time accident likelihood from empirical data on I-880 freeway in California, and shows its applicability as an accident precursor.

Road maintenance in Greece has not sufficiently addressed safety problems in the past. In an effort to assist maintenance agencies in justifying safety improvement and resource allocation decisions, a computerized system has been developed that provides suggestions for maintenance priority setting according to accident characteristics and proposals of likely countermeasures to specific safety hazards. The priority list is established based on accident and fatality rates as well as on high occurrence frequencies of a single accident type. Proposed countermeasures result from a knowledge base that has been developed based on past experience and expert judgments. The system effectiveness in providing appropriate solutions can be evaluated in time through before-and-after analyses and results are fed back to improve the knowledge base. The system includes a database where information about accident characteristics, road environment characteristics, and maintenance history is stored. Data input and result presentation forms have been developed to accommodate user interface. Analysis tools are employed for black section assessment (of roads in poor condition) and before-and-after analysis.

Well-designed roadways should equally consider and address safety, mobility, and community needs. The concept of context sensitive design is an emerging project development and geometric design approach that aims to meet the objectives of these elements. Central to the design process is the selection of key design elements that have the potential to influence design choices and roadway elements. Design speed is probably the most influential factor and current practices on how this speed is selected and used may need to be revisited. The need to design a roadway that could provide all required clues to the driver in order to drive at desired operating speeds is also an issue that requires attention. Design consistency along with compatibility between design and operating speed feature prominent among the issues that should be considered within the context sensitive design. This paper focuses on various design elements and discusses their relevance, importance, and influence on context sensitive design.

Although the single point interchange (SPI) has been around for approximately $3\text{decades}$ , very few attempts have been made to evaluate its safety performance with a similar configuration, the tight diamond interchange (TDI). This paper provides a crash comparison between the TDI and the SPI interchanges for intersection related crashes on the cross road only. The SPIs considered in this study are without frontage roads. Data from 27 diamond interchanges (DIs) in Washington were used in building a negative binomial model to predict total crashes and injury/fatality based on the off-ramp flow, cross-road flow, and separation distance between left and right ramp terminals. Crash analysis of 13 SPI sites was used to compare with newly developed crash models of diamond interchanges. Wilcoxon signed-rank test was applied for safety comparisons between the SPI and TDI. The safety comparison did not reveal a significant difference between the two types of interchanges neither for total crash. However, the single point urban interchanges were found to be safer than the comparable DIs for injury/fatality frequencies.

Analysis of actual accident experience in the United States from 1995 to 2000 has revealed that railroad commuter services using a combination of high and low level platforms exhibit a higher injury rate in boarding and alighting accidents than systems with either all high or all low platforms. This paper examines the safety features of these different types of station platforms and the corresponding railroad car entranceways used on these systems. New designs that would incorporate features of the lower accident rate high level only or low level only systems are identified and examined. Estimates of the reduction in the number of injuries that would result from use of these designs are made. The connection to other important performance features impacted by the designs are briefly considered, including cost, speed, meeting Americans with Disabilities Act requirements, and compatibility with freight service. Conclusions regarding the incorporation of boarding and alighting accident considerations in the evaluation of the various car and platform design options as rail passenger lines are upgraded or new services introduced are presented.