Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 28
2 Safety Focus Area SHRP 2 Safety program goal: To prevent or reduce the severity of highway crashes through more accurate knowledge of driver behavior and other crash factors. D river behavior has been identified as the major factor in approximately 90 percent of roadway crashes (Sabey and Staughton 1975; Treat et al. 1979; Hendricks et al. 2001). Yet the driver remains the most difficult part of the system to study. Driving simulators and studies performed on test tracks or in special test vehicles have made important contributions to understanding the driver, but these methods do not always provide good representations of real-world driving conditions or behavior. Surveying drivers about their behavior and interviewing them after a crash fail to yield very accurate or reliable reports of real behavior, not only because people are biased in their assessments of their own driving prowess or fear liability in the case of a crash but also because so much driver behavior occurs with- out full conscious awareness. In the case of crashes, events occur so quickly and many responses are so automatic that drivers often cannot remember how they acted—whether they braked or accelerated, for example. Advances in sensors, cameras, computing, and communications tech- nologies have now made it possible to gather real-world driving data that have never before been accessible to highway safety professionals. Sophis- ticated instrumentation packages can be installed inconspicuously in vol- unteers’ vehicles to collect data on speed, acceleration, and braking; use of signals, lights, wipers, and safety belts; lane position; and more. Extremely small cameras can capture looking and motor behavior without the driver’s 28

OCR for page 28
safety focus area 29 noticing the cameras or being reminded of their presence. This informa- tion can be collected not just when a crash or unusual incident occurs, but throughout the course of ordinary driving. For the first time, then, objective, scientific information can be obtained about what happens when people crash, when they experience a near- crash, or when they drive without incident. This information will make it possible to draw confident conclusions about the crash risks posed by vari- ous factors and whether those factors are related to the vehicle, the driver, the roadway and traffic environment, or some interaction of these. This kind of safety research is sometimes referred to as a “naturalistic driving study” because it captures driving in natural or real-world circumstances. The benefits of this new knowledge will be realized in more effective use of existing safety countermeasures and in the development of entirely new safety strategies. An example of these benefits is the use of actual driving data as educa- tional feedback for younger drivers (see Box 2-1). The results of naturalistic driving studies are also valuable to states as they develop Strategic Highway Safety Plans. These plans call for data-driven, highly effective strategies that maximize safety benefits, in terms of reductions in deaths and serious injuries, for each dollar invested. In these initiatives, states are considering Box 2-1 iowa teen driver study Teen drivers exhibit high crash rates, especially early in their driving experience. The University of Iowa demonstrated the use of in-vehicle sensors and cameras as an educational tool to help parents coach their children in better driving skills. The instrumentation on the teenagers’ vehicles was configured to save data when certain thresholds of acceleration were triggered (in contrast to the SHRP 2 study, in which data will be collected and saved continuously). Each week the events cap- tured by the instrumentation were shared with the drivers and their parents as an educational opportunity. The teens with the highest frequency of safety-relevant events improved their driving by 72 percent in the first 9 weeks of the intervention (McGehee et al. 2007). Insurance companies have begun to offer vehicle instru- mentation as a service to parents insuring their teenage drivers.

OCR for page 28
30 implementing the results of the second strategic highway research program behavioral programs as well as roadway initiatives, bringing together “4E” (engineering, education, enforcement, and emergency medical services) organizations to determine what programs are needed across the board. To make these prioritization processes work well, states need not only more knowledge about the effectiveness of current behavioral programs but also knowledge concerning what new behavioral treatments are needed to further reduce the current crash toll. Similarly, vehicle manufacturers and their suppliers, the National Highway Traffic Safety Administration (NHTSA), and others can use the knowledge provided by naturalistic studies to target their safety efforts. shrp 2 safety research SHRP 2 is taking a systems approach to safety by examining how the driver interacts with the roadway, the vehicle, and environmental factors. Using technologies designed and tested in smaller-scale studies, SHRP 2 will instrument the vehicles of 4,000 volunteer drivers and record their driving for a year or more. The drivers will be men and women in various age groups, driving different types of light vehicles, from different socioeconomic strata, and from different geographic areas across the United States. In addition to having instrumentation in their vehicles, the drivers will take a battery of tests and respond to questionnaires concerning a number of factors that may be related to driving performance so these factors can be studied in relation to actual driving behavior. The study will collect data not only on the driv- ers’ behavior but also on their vehicles’ characteristics and performance. To capture the roadway element of the safety interaction, data on road type, geometry, shoulders, safety furniture, signage, pavement markings, and more will be collected for the roads used by the volunteer drivers during the study. Through the Global Positioning System (GPS), driver behavior at various locations will be correlated with roadway and roadside features at those locations. In addition, data on environmental variables such as traf- fic, lighting, and weather conditions will be collected to the extent feasible, whether from the in-vehicle data systems or by other means. The research described above—known as the SHRP 2 naturalistic driv- ing study—represents by far the largest study of its type ever undertaken and promises to be a resource for improving highway safety for decades to

OCR for page 28
safety focus area 31 come. The specific research projects included under the SHRP 2 Safety program are listed in Appendix B. In the original plan for SHRP 2 safety research (TRB 2003), the natural- istic, or in-vehicle, driving study was to form one track of research, while a second track would employ a “site-based” research strategy. The site-based approach involves instrumenting roadway segments or intersections with multiple overhead cameras and automating vehicle trajectory tracking to study the driving behavior of multiple vehicles at specific locations. These two tracks of research are complementary in nature. The naturalistic driving approach provides detailed information about a single driver and vehicle over an extended period of time but little information about surrounding vehicles and the behavior of their drivers. The site-based approach, on the other hand, provides little information about individual drivers (it would not reveal, for example, whether a driver was wearing a safety belt or using a cell phone) but useful information about the movements of all the vehicles passing through a targeted segment of the roadway. Thus the site-based technique makes it pos- sible to study the interactions among drivers in a complex driving scenario. Specific sites, rather than the roads chosen by the drivers, can be studied, and site characteristics can be varied to examine changes in such features as signal timing, speed limits, signage, and pavement markings. When it became clear that SHRP 2 funding would be significantly less than originally proposed, the SHRP 2 Safety Technical Coordinating Com- mittee chose to focus on the in-vehicle track of research more than on the site-based track for the following reasons: • The technical feasibility of the in-vehicle instrumentation had already been demonstrated in smaller-scale studies, while the site-based technol- ogy required some additional development before it could be used in large- scale field work. • The opportunity to conduct the naturalistic driving study under SHRP 2 is truly unique because the scope and scale of this study make it highly unlikely that it will be undertaken by anyone else, while the final develop- ment and use of the site-based technology could be accomplished more easily by others in smaller-scale studies if it proves technically feasible. • The in-vehicle strategy promises to provide unprecedented insights into driving that could lead to significant improvements in highway safety at the

OCR for page 28
32 implementing the results of the second strategic highway research program national level in the long run and that cannot be provided through any other research approach available today. At the same time, the site-based approach was considered promising enough to merit some investment by SHRP 2. The program has funded a project to further develop the site-based technology in hopes of creating a robust tool that can be easily deployed to study different types of high- way locations. Of particular interest are intersections, where 45 percent of reported crashes and 21 percent of fatalities occur (FHWA 2008). Intersec- tions are locations of multiple and complex interactions of many drivers, making them appropriate targets for the site-based approach. If the SHRP 2 project in this area is successful, university researchers and state and local safety engineers could use the site-based approach to study issues associ- ated with particular roadway geometries that appear to be crash prone. If SHRP 2 were to receive additional funding, follow-on projects for field deployment and analysis could be conducted within the program. promising products, and potential users, incentives, and barriers SHRP 2’s Safety focus area is expected to produce the following products: • Initial findings that can be used in developing new driver, vehicle, and roadway treatments to reduce deaths and injuries and in improving existing treatments. • A rich source of naturalistic driving data, linked with roadway data, of unprecedented size and diversity, as well as tools for the development and evaluation of potential crash countermeasures. Safety researchers and practitioners will be able to use these data and tools to improve highway safety for years, if not decades, into the future. • Analysis tools [including validated crash surrogates (defined later in the chapter)]; research protocols; and specifications for monitoring, recording, and encoding instrumentation that safety researchers will be able to use and build on. • A site-based video system for studying vehicle behavior on particular roadway segments, such as intersections.

OCR for page 28
safety focus area 33 Initial Findings The current SHRP 2 effort will provide answers to a limited number of high-priority safety questions. Safety practitioners and researchers nation- wide were consulted to develop a list of nearly 400 safety questions that can be addressed with the use of SHRP 2 data. Box 2-2 contains examples of the types of questions being considered. As of this writing, the questions were still being prioritized to determine which ones will actually be addressed within SHRP 2. The answers to these questions will provide safety profes- sionals with guidance for more effective safety strategies and are likely to Box 2-2 selected research questions generated for shrp 2 Lane-Keeping and Run-off-the-Road Crashes • Does lane-keeping vary with driver age, gender, or vehicle type? • Does risk of lane departure vary by road type, traffic volume, superelevation, or presence of opposing traffic? • How are run-off-the-road crashes affected by different roadway geometries, shoulder widths, speed limits, signage, or pavement markings? • What is the role of following distance in lane-change/merge crashes? • How does aggressive driving behavior affect the risk of crashes or near-crashes? • Is there a relationship between aggressive driving behaviors and run-off-the- road crashes or near-crashes? Intersection Safety • How do roadway design, traffic control variables, and signage influence be- havior at intersections, such as braking, gap acceptance, and decision to turn at intersections? • How does roadway design influence compliance with traffic controls at inter- sections? • What is the effect of access points near the intersection? In-Vehicle Technologies • Is the frequency of use of infotainment or navigation devices affected by road type or traffic volume? • What is the influence of road geometry on driver behavior during technology- related tasks?

OCR for page 28
34 implementing the results of the second strategic highway research program yield input to future editions of the Highway Safety Manual and the Compre- hensive Human Factors Guidelines for Road Systems (TRB 2005). The studies conducted for the chosen questions will also serve as a demonstration of the uses and usefulness of SHRP 2 data. Box 2-3 provides examples of findings from other naturalistic driving studies performed at the Virginia Tech Transportation Institute, while Box 2-4 describes an array of uses of similar data from field operational tests per- formed at the University of Michigan Transportation Research Institute. These studies are much smaller than the SHRP 2 naturalistic driving study and do not involve as wide a variety of drivers and geographic locations or the more detailed roadway data that SHRP 2 will have. Nonetheless, their find- ings are illustrative of what can be learned from such studies. In addition to developing new findings, the larger SHRP 2 study will produce databases that can be used to verify the findings of smaller studies and test the robustness of these findings under different scenarios. In-Vehicle Data The 2-year SHRP 2 naturalistic driving study will produce a number of data- bases for use by researchers and practitioners from universities, highway agencies, automobile manufacturers, and others concerned with highway safety. The raw data—from cameras, vehicle instrumentation, and roadway data collection—will be preserved for researchers who have interest and skills in mining these data. Reduced data sets—more accessible and more user-friendly versions of the raw data or some subsets thereof—will be use- ful to both researchers and practitioners. In many cases, safety practitioners will be indirect users of the data in that they will engage researchers to be the direct users. Three main components will collect data on or from the volunteers’ vehi- cles. The first will collect data on speed, acceleration, braking, seat belt use, and other vehicle-related factors. The second will be a set of inconspicuous cameras that will capture the volunteer’s face and the driver’s view out the front, rear, and side of the vehicle. The third will be a radar unit to capture the presence of other vehicles or objects near the subject vehicle. The data from these components will be encrypted, be transmitted to a secure loca- tion, undergo quality checks, and be stored for reduction and analysis. In addition to these data, as noted above, the volunteer drivers will be asked

OCR for page 28
safety focus area 35 Box 2-3 examples of findings from naturalistic driving studies The Virginia Tech Transportation Institute (VTTI) has conducted a number of small naturalistic driving studies, most notably a study in which 100 volunteers drove instrumented vehicles for 1 year. VTTI has also conducted studies focused on teen- age drivers and on truck drivers. The findings described here illustrate the types of information that can be gleaned from naturalistic driving studies and the kinds of countermeasures that may be developed as a result. Driver inattention—particularly looking away from the roadway just before an unexpected event or condition—is the largest contributing factor to unsafe events such as crashes and near-crashes. Secondary tasks (those unrelated to driving) and external distractions account for most inattention-related risk. The highest risk is associated with looking away many times or for long periods of time. The kinds of secondary tasks associated with such frequent or prolonged inattention include dial- ing a cell phone, text messaging, manipulating an MP3 player, and using the Inter- net. Teenage drivers are four times more likely than adult drivers to be involved in a crash or near-crash while performing a secondary task. Such findings contributed to a Virginia law banning the use of handheld electronic devices by teen drivers. High-tech countermeasures, such as driver “eyes-forward” monitors, in combina- tion with other crash avoidance technology, could be used to warn drivers of unsafe behaviors and changing circumstances. Driver drowsiness is a contributing factor in 15 to 20 percent of crashes and other safety-related incidents involving long-haul trucking, local and short-haul trucking, and light-vehicle driving, and it can occur at any time of day. Crash and near-crash risk is 6 to 8 times higher when driving while drowsy than when driving while alert. A study of truck drivers found that long-haul “team” truck drivers have poorer sleep quality but sleep for longer periods and are generally safer than “single” drivers. Among local and short-haul truckers, drowsiness was found to be most closely associated with beginning the workweek in a tired state. Potential counter- measures include adjustments to hours-of-service regulations, electronic log books for truckers, and high-tech driver alertness monitors for all types of drivers. Evaluation of countermeasures can be carried out by using naturalistic driving data. For example, in-vehicle forward collision warning algorithms were compared by overlaying data from VTTI’s 100-car study on actual crash and near-crash data. The use of real-world scenarios makes it possible to test whether the driver could have avoided a crash and under what circumstances the system would fail to activate for a real crash. These results can be used to refine existing countermeasures and to estimate the crash avoidance benefits of countermeasures under development. SOURCE: VTTI (cf. Hanowski et al. 2000; Dingus et al. 2001; Dingus et al. 2005; Dingus et al. in progress; Blanco et al. in progress).

OCR for page 28
36 implementing the results of the second strategic highway research program Box 2-4 additional examples of the usefulness of in-vehicle data The University of Michigan Transportation Research Institute has collected natu- ralistic driving data in a number of field operational tests. More than 800,000 miles of driving data have been obtained from both light passenger vehicles and heavy trucks, including hundreds of data channels from the vehicle; from video; and from add-on instrumentation such as radar, lidar, and inertial motion sensors. These data were originally collected to evaluate the performance and use of new types of safety systems on the vehicles, but the majority of the data are applicable to research and analysis across a wide spectrum of topics involving vehicles and highways. For example, General Motors used these data to develop an algorithm for a forward col- lision warning system and to assess the safety impact of adaptive cruise control. The University of Michigan has used the data to study highway safety and vehicle use patterns relative to the urban environment, to model driving behavior, and to ana- lyze fuel efficiency to support the design of future hybrid vehicles. A project funded by the Alzheimer’s Association used the data as a benchmark for evaluating perfor- mance and likely safety concerns for drivers suffering from early-stage dementia. In most cases, the information obtained from these data cannot be obtained from any other source. SOURCE: University of Michigan Transportation Research Institute. to take a number of tests and respond to questionnaires to provide infor- mation about themselves—such as their visual acuity, psychomotor skill, physical capability, risk taking and risk perception, sleep hygiene, medical conditions and medications, and driving knowledge and history—that can be anonymously correlated with the driving data. Another data component of the naturalistic driving study—a component never before included at this level of detail in such a study—is extensive road- way data on curves and grades, intersection locations and characteristics, pavement and shoulder types, presence and types of signs, pavement mark- ings, guardrails, and the like. Although states routinely collect roadway data for infrastructure assessment, inventory, and asset management purposes, these data are limited. They do not always cover all roads in a jurisdiction, they may not cover locally owned roads, and they may not include all the elements needed for safety analyses. Having these data will allow SHRP 2

OCR for page 28
safety focus area 37 researchers to develop greatly improved knowledge about the relationship of these features to safety outcomes. This knowledge can be used to develop better tools for the design and selection of safety countermeasures. SHRP 2 will collect data of unprecedented quantity and detail on motor vehicle use, near-collisions, and collisions. These data will enable public health researchers and motor vehicle safety practitioners to probe and understand the conjunction of events and conditional circumstances that lead to collisions and near-collisions in a way never before possible. The data will make it possible to identify and prioritize roadway traffic safety challenges, develop and assess potential countermeasures, and deploy those countermeasures that provide a net societal benefit. It is not unreasonable to believe that the countermeasures identified, developed, and deployed as a result of the SHRP 2 study could reduce motor vehicle–related fatalities and serious injuries by tens of thousands annually. The incident-free data collected—which safety professionals refer to as “exposure”—will provide the basis for comparisons of crash risk under different circumstances and involving different factors. Objective data on what really occurs before and during a crash or near-crash—how fast the driver was really going; when he or she braked; and how speed and braking are affected by cell phone use, fatigue, or other factors—will provide better understanding than biased and incomplete postcrash recollections of drivers and observers that until now have served as the primary source of such data. SHRP 2 data will also make it possible to study how drivers react to differ- ent roadway and environmental features and how their reactions affect crash risk. A strength of the SHRP 2 data, never available previously, will be the synthesis of highly detailed roadway, environmental, and driver data. The SHRP 2 data will enable study of how such factors as driving speed, braking, steering, and attention change with changes in roadway and environmen- tal features—such as lane width, signing, pavement type, pavement mark- ings, shoulder type and width, lighting conditions, and prevailing weather conditions. This in turn should allow some roadway countermeasures to be explored analytically, with a wide variety of drivers, by taking advantage of features already in place in various areas instead of implementing the features in an experimental situation and waiting a long time to acquire crash data. For example, SHRP 2 will help in understanding when and where speed- ing is most likely to occur; when it poses the greatest crash risk; and which

OCR for page 28
38 implementing the results of the second strategic highway research program countermeasures, such as innovative signing and roadway treatments, are most effective in reducing unsafe speeds. Such insights, paired with tech- nology (such as digital maps), will provide the opportunity to deliver notice to drivers of changing roadway and environmental conditions and thereby allow advanced driver response. Driver distraction is another important area in which SHRP 2 research can contribute by identifying the sources of distraction and calculating the crash risk associated with different sources. Cell phones, in-vehicle devices, and other technologies are hypothesized to be a major source of distraction—hypotheses that can be tested by SHRP 2. Cell phones in particu- lar are a rapidly growing concern. Approximately 230 million cell phones are now in use in the United States. Current NHTSA surveys show that about 6 percent of drivers are using cell phones at any given moment (NHTSA 2008). The SHRP 2 data will make it possible to examine differences in crash risk while drivers are using a cell phone under various driving conditions (e.g., road type, traffic volume, speed) or by demographic identifiers (age or driving experience, for example). Results of research on these questions could lead to a number of safety countermeasures. Rumble strips or other roadway design features could help bring a driver’s attention back to the driving task. In-vehicle technology could be developed to prevent cell phones from func- tioning while drivers are operating in challenging traffic and environmental conditions with high cognitive demand and to enable safe cell phone use in low-demand conditions. Laws on the use of cell phones (for example, hands- free versus handheld and restrictions on younger drivers) could be based on empirical evidence of risk. Education programs and new technologies could perhaps be tailored to particular audiences and driving behaviors. Other nomadic devices that can be brought into the vehicle—including television, laptops, MP3 players, texting devices, and GPS devices—are a rapidly growing issue, though far less well documented than cell phones. There is some evidence that they pose a greater problem for young drivers who are the early adopters of these devices, use them frequently, and may think they can multitask without being distracted to a level that compro- mises the cognitive capacity they devote to the driving task. SHRP 2 data will allow the risks associated with these devices to be studied and quanti- fied so that appropriate countermeasures, possibly similar to those sug- gested for cell phones, can be effectively applied.

OCR for page 28
safety focus area 39 New technologies increasingly being added to the vehicle also raise questions about distraction. Do onboard navigation devices contribute to distraction? Do in-vehicle warning systems (lane departure, collision, speed, and traction warning systems, for instance) contribute to distrac- tion, or are they a potential solution to all driver distraction issues? SHRP 2 data will allow algorithms for these systems to be developed and tested. For example, developers of lane departure or collision warning systems will want to minimize the number of false positives (i.e., the warning sounds when it is not needed, annoys the driver, and possibly leads to ignoring the warning when it is real) and false negatives (i.e., the warning does not sound until it is too late for the driver to react). The task for technology and system designers is to integrate new systems that provide value and func- tion without adding distracting operational features and to restrict access to operational features during periods of high driver workload and cogni- tive demand. SHRP 2 data will be able to be matched to the typology of crashes developed by the Volpe National Transportation Systems Center (Najm and Smith 2007) and enable an understanding of collision morphol- ogy (causal factors accumulating to cause a crash or near-crash). As they develop a profound understanding of how collisions occur and under what contributing conditions, safety professionals will be able to better conceive, develop, and assess potential countermeasures. The major barrier or challenge to the implementation of SHRP 2 Safety results is the need to house and maintain the data—updating hardware and software as appropriate; providing security; and staffing a center that pro- vides access to the data, assistance, and other services to researchers and practitioners. Significant costs are involved, and funding must be stable over time to avoid the loss of data availability, to provide training and sup- port, and to avoid hardware or software obsolescence. In addition, funding is needed to help ensure exposure of the data to both safety researchers and researchers in other fields who might have unique ideas about how to use them to derive new knowledge and to help in providing web- and university- based training courses. Research and Analysis Tools In carrying out the naturalistic driving study, SHRP 2 will develop or improve on a number of tools that can be used by both safety researchers and

OCR for page 28
40 implementing the results of the second strategic highway research program planners. Researchers may adopt the research protocols and management approaches developed for the study as guidelines, or perhaps even de facto standards, for the conduct of similar studies in the future. Because the SHRP 2 study will involve thousands of volunteers throughout the coun- try and make use of multiple contractors, it will advance the highway safety research community’s knowledge concerning studies with human subjects and how to work with multiple institutional review boards (IRBs) for such research.1 All of these tools and the knowledge gained will contribute to a new way of approaching highway safety based on more scientific understanding of risk factors and driver behavior. They also may lead to revised data needs with respect to crash reporting, roadway inventory, driver characteristics, and other safety-related factors as part of a comprehensive approach to asset management. The safety community currently experiences great dif- ficulty in achieving and maintaining adequate attention to and investment in high-quality safety data collection programs. SHRP 2 results should help define the most relevant sets of data that should be gathered by state and local entities so that data collection programs can be cost-effective and fully justified to leadership. Another analysis tool expected from SHRP 2 is crash surrogates—events or conditions that precede crashes, happen more frequently than crashes, and are highly correlated with crashes. For example, unintentional lane departures may be a surrogate for run-off-the-road crashes. The establish- ment of surrogates for major crash types could revolutionize the evalua- tion of safety countermeasures. Currently, to evaluate the effectiveness of a countermeasure in preventing crashes, a safety professional must spend years collecting crash data before and after the countermeasure has been applied. Using surrogates would take less time because surrogates take place more frequently than crashes; their use would therefore reduce, if not eliminate, the need to wait for actual crashes to occur to ascertain the effectiveness of countermeasures. 1 An IRB is an ethics board that reviews and approves research involving human subjects to ensure that the subjects’ rights and safety are protected. Universities and other research institutions that perform medical and social science research typically have one or more IRBs. The SHRP 2 naturalis- tic driving study will involve as many as a dozen IRBs because of the number of contractors engaged in the research.

OCR for page 28
safety focus area 41 Site-Based Video System If SHRP 2 can demonstrate the technical feasibility of the site-based safety research tool—the system of overhead cameras and automated vehicle tra- jectory tracking—this tool could be used by highway safety professionals to study driving behavior at locations of particular interest. The system, which is intended to be deployable at different types of locations, could be used by state and local agencies to evaluate new intersection geometries, different traffic signal timing, various warning signs, pavement markings at hazardous curves or at entrances to work zones, or other countermeasures. This could be accomplished by setting up the system before and after such a change has been made and studying its effect on speed, lane changing, gap acceptance, or other behaviors of interest. As in the naturalistic driving study, the site-based tool could be used to discover crash surrogates, which could be analyzed in lieu of actual crashes to determine the safety impacts of a countermeasure. At its current funding level, SHRP 2 will not be con- ducting these analyses itself. If the system can be demonstrated to work, however, such studies would be within the capacity of many states’ research programs, as well as those of the Federal Highway Administration, NHTSA, the National Cooperative Highway Research Program, the American Auto- mobile Association’s Foundation for Traffic Safety, the Insurance Institute for Highway Safety, and others. The incentives to implement this technology will be its flexibility in deploy- ment to different types of locations and its ability to record and automatically identify trajectories of multiple vehicles, making it possible to measure the safety performance of locations such as intersections. Once the initial tech- nical challenges have been overcome (mainly by using state-of-the-art tech- nology and the latest tracking algorithms), wide-scale implementation can follow. A growing number of companies and products use this type of video image processing technology for simpler applications in traffic management and incident detection. The SHRP 2 project will set benchmarks for the development of new systems that can be used to assess safety performance with accurate trajectory measurements. In the short to medium term, robust research-level prototypes will become available to highway agencies and others interested in site-based safety analysis. These prototypes will use off- the-shelf hardware components (cameras, networking, workstations), with low-volume manufacture of the data acquisition hardware. Software needed

OCR for page 28
42 implementing the results of the second strategic highway research program to run the system will also be made available. At this stage, the number of users and programs involved will be somewhat limited. For the longer term, commercial companies are already interested in further developing this type of technology to make it generally available and fully supported for the broader user community. In this context, the project can be seen as a catalyst for a new generation of safety evaluation systems, one that has established feasibility, generates awareness and interest, and sets standards that must be attained or surpassed by future commercial products. Other Potential Users The majority of the users of SHRP 2 Safety products are expected to be highway safety practitioners and researchers. Nevertheless, as knowledge of this research becomes more widespread, additional sectors of the high- way community are expressing interest in using the data to address other critical issues. For example, highway planners are interested in the route choices made by the volunteers in the safety study. Traffic operations profes- sionals want to see whether the data can help them understand how drivers behave when they travel in work zones or drive past crash scenes. The site- based video tool could even be set up in work zones or near special events to study traffic patterns under these circumstances. In the naturalistic driving study, volunteers’ daily driving behavior will be recorded, and these data could be used to generate travel time distributions for repeated trips on the same route. This information, together with data on weather, crashes, or special events, could be used to study the impacts of such events on travel time reliability and possibly the detour choices made by drivers. The data derived from the study could also be used to examine driver behavior and vehicle performance related to air quality, such as cold starts. SHRP 2’s Reliability focus area (see Chapter 4) is carrying out a study to determine the feasibility of using data from safety research such as SHRP 2’s natural- istic driving study to address some of these other issues. conclusion The SHRP 2 Safety program represents the most ambitious driver-centered, system-oriented safety research effort conducted to date. Its main compo- nent is a naturalistic driving study involving approximately 4,000 volunteers

OCR for page 28
safety focus area 43 over a period of 2 years. This study will yield hitherto unobtainable data on driver behavior and the interaction among driver performance, vehicle performance, and roadway conditions before, during, and after crash events, as well as during near-crashes and normal driving. Crash risks associated with various factors will be determined, and this information will form the basis for better use of existing safety countermeasures and development and application of new countermeasures. Identification of crash surrogates will make evaluation of countermeasures significantly faster and safer. The data that make all this possible will be a major national resource. However, the potential benefits of this SHRP 2 Safety research will not be fully realized without a continued investment of human and financial resources to ensure that this major national resource is made available to users long after SHRP 2 has ended. references Abbreviations FHWA Federal Highway Administration NHTSA National Highway Traffic Safety Administration TRB Transportation Research Board Blanco, M., R. J. Hanowski, D. Bowman, J. S. Hickman, A. Nakata, M. Greening, P. Madison, R. L. Olson, and G. T. Holbrook. In progress. The Naturalistic Truck Driving Study—Commercial Vehicle Data Collection and Countermeasures Assessment Research Project, Phase II: Investigating Critical Incidents, Driver Re-Start Period, Sleep Quantity, and Crash Countermeasures in Commercial Vehicle Operations Using Naturalistic Data Collec- tion. Interim Project Report for DTFH61-01-C-00049, Task Order 23. Federal Highway Administration, Washington, D.C. Dingus, T. A., S. G. Klauer, and S. Lee. In progress. Research on Driving Risk Among Novice Teen Drivers—Study 1. National Institutes of Health, Washington, D.C. Dingus, T. A., S. G. Klauer, V. L. Neale, A. Petersen, S. E. Lee, J. Sudweeks, M. A. Perez, J. Hankey, D. Ramsey, S. Gupta, C. Bucher, Z. R. Doerzaph, J. Jermeland, and R. R. Knipling. 2005. The 100-Car Naturalistic Driving Study: Phase II—Results of the 100- Car Field Experiment. Interim Project Report for DTNH22-00-C-07007, Task Order 6. Report No. DOT HS 810 593. National Highway Traffic Safety Administration, Wash- ington, D.C. Dingus, T. A., V. L. Neale, S. A. Garness, R. J. Hanowski, A. S. Keisler, S. E. Lee, M. A. Perez, G. S. Robinson, S. M. Belz, J. G. Casali, E. F. Pace-Schott, R. A. Stickgold, and

OCR for page 28
44 implementing the results of the second strategic highway research program J. A. Hobson. 2001. Impact of Sleeper Berth Usage on Driver Fatigue: Final Project Report. Contract No. DTFH61-96-00068. Federal Highway Administration, Washington, D.C. FHWA. 2008. Crash Types and Causes. U.S. Department of Transportation, Washington, D.C. http://safety.fhwa.dot.gov/intersections/inter_facts.htm#. Accessed July 9, 2008. Hanowski, R. J., W. W. Wierwille, S. A. Garness, and T. A. Dingus. 2000. Impact of Local/ Short Haul Operations on Driver Fatigue. Draft Final Report, Contract No. DTFH61-96- C-00105. Virginia Tech Transportation Institute, Blacksburg. Hendricks, D. L., J. C. Fell, and M. Freedman. 2001. The Relative Frequency of Unsafe Driving Acts in Serious Injury Accidents. Veridian Engineering, Buffalo, N.Y. McGehee, D. V., M. Raby, C. Carney, J. D. Lee, and M. L. Reyes. 2007. Use of Video Feed- back in Rural Teen Driving: An Intervention Study. Proc., 2007 Mid-Continent Transpor- tation Research Symposium, Iowa State University, Ames, Aug. Najm, W. G., and D. Smith. 2007. Definition of a Pre-Crash Scenario Typology for Vehicle Safety Research. Presented at 20th Enhanced Safety of Vehicles Conference: Inno- vations for Safety: Opportunities and Challenges, Lyons, France, June 18–21, 2007. http://www-nrd.nhtsa.dot.gov/pdf/nrd-01/esv/esv20/07-0412-o.pdf. NHTSA. 2008. Driver Electronic Device Use in 2007. Traffic Safety Facts Research Note. DOT HS 810 963. U.S. Department of Transportation, Washington, D.C., June. Sabey, B. E., and G. C. Staughton. 1975. Interacting Roles of Road, Environment, and Road User in Accidents. Presented at 5th International Conference of the International Asso- ciation for Accident and Traffic Medicine and the 3rd International Conference on Drug Abuse of the International Council on Alcohol and Addiction, London, Sept. 1–5. TRB. 2003. NCHRP Report 510: Summary Report: Interim Planning for a Future Strategic Highway Research Program. National Academies, Washington, D.C. TRB. 2005. NCHRP Research Results Digest 296: Comprehensive Human Factors Guidelines for Road Systems. National Academies, Washington, D.C., April. http://onlinepubs.trb. org/Onlinepubs/nchrp/nchrp_rrd_296.pdf. Treat, J. R., N. S. Tumbas, S. T. McDonald, D. Shinar, R. D. Hume, R. R. Mayer, R. L. Stan- sifer, and N. J. Castellan. 1979. Tri-Level Study of the Causes of Traffic Accidents: Final Report. Volume I: Causal Factor Tabulations and Assessments. Report No. DOT HS 805 085. U.S. Department of Transportation, Washington, D.C.