James L Brown

Matching Simulator Characteristics to Highway Design Problems

Driving simulators hold much promise for addressing roadway design issues. However, although simulators have demonstrated their value in experimental research addressing driver performance, their ability to support road design projects has not been as clearly established. This paper describes a design-centered framework to make simulators valuable for traffic engineers and geometric designers. This framework includes several steps: (a) identification of design issues that would benefit from driving simulators, (b) identification of simulator characteristics to match them to design issues, and (c) translation of driver performance data from the simulator to traffic behavior on the road. Several critical obstacles inhibit application of simulators to highway design. First, driving safety researchers and engineers comprise separate communities and their perspectives on how simulators can be applied to address road design issues often diverge. This paper seeks to reduce this divergence and make simulators useful to highway engineers. Interviews with engineers revealed important issues that simulators could address, such as intersection and interchange design. Second, driving simulators are often broadly defined as high fidelity, which provides little value in matching simulators to design issues. A survey of simulators and simulator characteristics clarifies the meaning of simulator fidelity and links it to road design issues. Third, simulators often produce data that do not correspond to data collected by traffic engineers. This mismatch can result from inadequate simulator fidelity, but can also arise from more fundamental sources—traffic engineers focus on traffic behavior and driving simulator researchers focus on driver behavior. Obstacles in using simulators for highway design reflect both technical and communication challenges.

Validations of Integrated Driver Vehicle Interface (DVI) Configurations

This project examines questions of discernibility and presentation methods for safety-critical driving messages. A driving simulator experiment tested two methods of providing safety messages: distinct (with all alerts having distinct auditory and visual components) and master (a common visual and auditory alert) presentations. Participants completed drives that contained a safety critical event, with and without an alert, and reported their perceptions of the alert’s meaning and hazard location. No significant differences were observed in participants’ ability to identify the location of the referent hazard. There were significant differences in participants’ ability to assess the meaning of the alert: the distinct group displayed higher overall performance as compared to the master group. Implications of the study for design guidance and potential future research topics are discussed.

Using Naturalistic Driving Data to Develop a Typology of Speeding Episodes

Speeding-related crashes continue to be a serious problem in the United States. A recently completed NHTSA project, Motivations for Speeding, collected data to address questions about driver speeding behavior. This naturalistic driving study used 1-Hz GPS units to collect data from 88 drivers in Seattle, Washington, to record how fast vehicles traveled on different roadways. The current project further developed this data set to redefine speeding in terms of speeding episodes, which were continuous periods in which drivers exceeded the posted speed limit by at least 10 mph. More than half of all study participants averaged less than one speeding episode per trip taken. Various characteristics of speeding episodes representing aspects such as duration, magnitude, variability, and overall form of speeding were examined. Cluster analyses conducted using these characteristics of speeding episodes identified six types of speeding. These included two types of speeding that occurred around speed-zone transitions (speeding up and slowing down), incidental speeding, casual speeding, cruising speeding, and aggressive speeding. Qualitative examination of the speeding types indicated that these types also differed in terms of the prevalence of additional risky situational characteristics.

Investigating Speeding Behavior with Naturalistic Approaches

Although speeding is a significant contributor to traffic fatalities, attempts to address this problem have not led to a significant reduction in speed-related fatalities. There are a number of inherent shortcomings in using primarily self-report surveys and crash data to learn more about why drivers speed and in selecting countermeasures that will most effectively address speeding behaviors. An emerging empirical approach is to study the speeding choices that drivers make under everyday driving conditions by using naturalistic driving methods. Such an approach has the potential to yield highly informative data about speeding. These data, however, are complicated and prone to analytical confusion and uncertain interpretation if some key conceptual and methodological issues are not addressed. In this paper, an overview is provided of a naturalistic driving study that was intended to (a) identify the reasons why drivers speed; (b) model the relative roles of situational, demographic, and personality factors in predicting travel speeds; (c) classify speeders; and (d) identify interventions, countermeasures, and strategies for reducing speeding behaviors. The focus here is on discussing lessons learned associated with three methodological issues in particular (defining speeding, identifying a way to measure exposure, and obtaining accurate posted speeds) that were crucial to successfully analyzing the data that this study provided and for generating useful results and conclusions. It is believed that careful consideration of these issues will greatly benefit the traffic safety community, especially as future analyses of naturalistic driving data are considered.

Human Factors Guidelines for Road Systems: Collection C: Chapters 16, 17, 18, 19, 20, 22 (Tutorials 4, 5, 6), 23 (Updated), 24, 25, 26 (Updated)

This report contains guidelines that provide human factors principles and findings for consideration by highway designers and traffic engineers. The guidelines allow the non-expert in human factors to more effectively consider the roadway users capabilities and limitations in the design and operation of highway facilities. The following chapters are included in collection C: (16) Special Considerations for Rural Environments; (17) Speed Perception, Speed Choice, and Speed Control; (18) Signing; (19) Changeable Message Signs; (20) Markings; (22) Tutorials (Tutorials 4, 5, 6); (23) References (Updated); (24) Glossary; (25) Index; and (26) Abbreviations (Updated).

Using a Task Analysis to Identify Potential Information Processing Bottlenecks in Intersection Driving Scenarios

A task analysis was conducted to identify the information processing bottlenecks that drivers face in specific intersection driving scenarios. These bottlenecks represent situations in which drivers may become “overloaded” by driving demands, possibly resulting in drivers improperly performing important driving tasks or skipping certain tasks altogether. The focus of this task analysis was on identifying the underlying information processing elements, including the perceptual, cognitive, and psychomotor subtasks associated with each individual driving task. A total of 7 distinct driving scenarios were investigated in the task analysis, with each scenario being successively decomposed into segments, tasks, and subtasks/information processing elements. Information processing bottlenecks were identified by analyzing converging information about workload levels, task sequencing and pacing, spatial distribution of required information, and other factors that mitigate or amplify taxing conditions.