Technologies for Improving Safety Data (2007)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

57 This synthesis summarizes the current state-of-the-practice and state-of-the-art utilization of technologies for efficient and effective collection and maintenance of data for highway safety analysis. Previous research reports have suggested technologies as providing a means by which to overcome many of the limitations surrounding safety data. This syn- thesis documents a number of successful implementations of technologies, whereby these six measures of safety data, timeliness, accuracy, completeness, comprehensiveness, efficiency, and integration, were improved. A multifaceted approach was used to define safety data requirements and identify technological solutions that enhance the timeliness, accuracy, efficiency, completeness, comprehensiveness, and integration of safety data sources. Six specific tasks were undertaken: • New safety analysis tools to identify data requirements were examined. • A survey of states was implemented to determine com- pliance with identified data requirements (crash, road- way inventory, and traffic operations) and to discover types of technologies used to collect, process, and manage the three types of data, as well as the extent of their use. • To ascertain additional information or clarify informa- tion on technology use, follow-up with the states was undertaken. • A matrix of available technologies to support collec- tion, processing, and maintenance within the three primary data categories was developed. • A literature search on technologies identified from the survey, general knowledge of the area, and through sug- gestions from the research review panel was completed. • A matrix matching technologies to specific data requirements to guide agencies seeking to employ new technologies to support safety information systems was developed. A comprehensive listing of safety data elements necessary for the use of the new safety analysis tools, Interactive High- way Safety Design Model (IHSDM) and SafetyAnalyst, was included in Appendix A of the report. The list is extensive and promises of new additions to the software packages will most likely lengthen the list in the coming months. Special attention was given to the differences in data requirements between IHSDM and SafetyAnalyst. IHSDM requires very detailed geometric data files for a specific section of road- way, whereas SafetyAnalyst requires comprehensive crash and road inventory data for a wide area or state. These dif- ferences leave little room for overlap in data requirements. Federally mandated data programs such as those for bridge structures, railroad crossings, and the Highway Per- formance Monitoring System have a strong effect on a state’s comprehensive collection of data. For bridge and pavement data elements, all states reported collecting these data ele- ments comprehensively. For elements such as road geomet- ric elements (horizontal and vertical curvature), however, fewer than half of the responding states indicated collecting the data elements comprehensively—most collect samples or on an as-needed basis, if at all. Overall, a number of data ele- ments required by safety programs (i.e., cross-slope, barriers, and intersection elements) are also underreported. The survey of states indicated that the use of technologies for collecting crash data was limited, and so too were efforts to research new methods for crash data collection and main- tenance. Although a number of states mentioned the capture of electronic crash reports, few had widespread use through- out the state. A number of technologies exist within these electronic reporting systems (i.e., barcode scanners, global positioning system location devices, etc); however, these too are not used extensively. Some states noted that other tech- nologies (i.e., laser or survey grade measurement and digital photography) are used for special studies, but that much of the data is filed away in hard copy limiting its long-term usefulness. As noted both in the crash database and road inventory surveys, route-milepost is the most common linear referenc- ing system in use by state departments of transportation DOTs. Unfortunately, the route-milepost system does not easily allow for multiple years of historical road inventory data to be readily available for analysis, making linkage to historical crash records difficult. Furthermore, roadway inventory databases were reported as dynamic databases that are constantly changing. These dynamic systems make it difficult to manage change dates for specific pieces of infor- mation (e.g., the date that raised pavement markers were added to a section of roadway or the date when a traffic sig- nal was added to an intersection). Most states archive the road inventory database each year, but not all roads are updated annually. Therefore, it is hard to establish a link CHAPTER SIX CONCLUSIONS

between road characteristic changes and safety improve- ments without chasing a significant paper trail of contract documents. According to the road inventory survey, technologies were used sparingly except for pavement management and video-log. Where federal programs require more compre- hensive data, states have adopted technologies to increase their efficiency with these data collection activities. How- ever, many of the data elements are still collected manually with pen and paper or pulled from plan and profile sheets. Expansion of this type of data collection system would be cost-prohibitive. Additionally, much of the data collected in hard-copy format must be entered or scanned, resulting in additional labor costs. The availability of AASHTOWare products for data man- agement for federal requirements is viewed as a benefit to the states. These software packages are typically designed to be flexible for use in most states, while allowing nationwide analy- sis of exported data. Further development of the anticipated Transportation Safety Information Management System, although daunting, is expected to be an important venture. The states had great difficulty in responding to the traffic operations data survey. However, it is clear that additional efforts to define a more comprehensive system for traffic operations data collection and maintenance are needed. Sev- eral states indicated that multiple databases were maintained to house traffic operations data—some for Intelligent Trans- portation System data archives, some for traffic signal data, and others to support volume requirements for the Highway Performance Monitoring System. One of the overarching requirements of both IHSDM and SafetyAnalyst is the need for volume data. More integrated and effective data systems will be needed to meet these demands. Although several states mentioned plans for deployment of portable nonintrusive traffic detection systems in the near future (5–10 years), few states have been early adopters. The use of portable nonintrusive technologies will be required to expand the traffic data needs of future safety analysis programs. Many technologies exist that can provide numerous effi- ciencies in safety data acquisition, processing, and manage- ment. Unfortunately, many of these technologies are not being fully or even partially integrated into the safety data process. This document focused on the review of technolo- gies that had been used successfully in one or more states. Often, specific evaluations of the technology implementa- tions were not completed; however, it could be assumed that the technologies had produced noticeable benefits. There are key inexpensive technologies such as global positioning systems and geographic information systems that are critical to data collection and maintenance for almost all databases (crash, roadway inventory, and traffic operations) 58 that have not been fully capitalized on. Most states do maintain a geographic information system; however, they incorporate only a small portion of the data that they collect. Furthermore, outdated procedures of locating crash locations in a post hoc manner using the police-reported description and route-milepost data are unnecessary and time-consuming. However, without support of management and significant investment in the technologies, little improvement in updat- ing processes for safety data is likely to occur. The choices regarding linear referencing systems and the use of dynamic versus static road characteristics databases have dramatic effects on a states’ ability to use data systems for safety analysis. Some are much better than others, yet some choices have positive implications for one database and negative implications for others. These choices are difficult and must be consciously considered upfront in the design and implementation of a safety data system. Unfortunately, most of the data sets that are currently being used for safety data analysis were developed for purposes other than safety and as such do not support safety analysis effectively. The lack of federally mandated safety data programs in past years has hindered this integration. Advances in wireless communications and mobile com- puting allow for ease of mobile data capture and reporting; however, paper and pen remains a top medium for crash data collection. In a time when it is possible to send a package across the country in less than 24 h and receive immediate e-mail confirmation of delivery with an attached digital sig- nature, it is feasible to upgrade and expedite our crash records systems. Several states have begun to deploy business-like systems for the capture and delivery of this critical safety data. Not only are these states able to recoup the costs of the data system, but they are also able to recoup costs for damage to infrastructure owing to the timeliness of the data. Of all of the required data elements, key design elements (horizontal and vertical curvature and grade) are missing from road inventory databases. There is no doubt that some of these elements have strong ties to crash occurrence; how- ever, serious attention must be given to the requirements of comprehensive collection and accuracy of any such data. Several commercial pavement profiling and video-log sys- tems provide road design data, but the accuracy of the data is not clear. A recent development by FHWA, the Digital High- way Measurement Vehicle, provides extremely accurate and comprehensive road design data. Streamlining this system or a similar system and deploying it in a widespread fashion will be necessary to meet the needs of safety analysis in the near future. Finally, several exceptions should be noted with regard to technology implementation: • Technology alone cannot solve all of the problems associated with safety data systems, especially those

59 related to inadequate institutional cooperation. Organi- zational issues should be addressed before considering technological advancements. • Technology implementation and maintenance can be capital-intensive, requiring significant funding and pro- grammatic support—therefore, the benefits and costs must be clearly evident across all affected agencies. • Technologies are constantly evolving; therefore, agen- cies should seek to employ technologies that allow for flexibility. Additionally, a practical plan for maintain- ing and upgrading the technologies, as well as assess- ing their continued effectiveness, should be developed before initial investment and implementation of the technology. • No single technology will allow for the collection and maintenance of the varied databases required for safety analysis—hence, an array of technologies should be considered. In conclusion, technology is a means by which to over- come many of the limitations set forth by Pfefer et al. in NCHRP Report 430. Given institutional support, cooperative agreements between agencies, and necessary resources positive change can occur in safety data systems.