Use of Event Data Recorder (EDR) Technology for Highway Crash Data Analysis (2005)

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.

Research Results Digest 298 July 2005 SUMMARY This digest summarizes the final re- port for NCHRP Project 17-24 (available as NCHRP Web-Only Document 75). This digest examines the benefits and the costs of using EDR data in highway crash data analysis and research. Although EDRs hold tremendous promise for improving high- way crash data analysis, this digest identi- fies several issues that may impede the use of EDR data for this purpose. These im- pediments include technological, legal, and consumer acceptability concerns. These is- sues were investigated in depth and rec- ommendations were developed for resolu- tion of these potential barriers to the use of EDR data. INTRODUCTION Widespread deployment of EDRs, sometimes called “black boxes,” promises a new and unique glimpse of the events that occur during a highway traffic collision. The EDR in a colliding vehicle can provide a comprehensive snapshot of the entire crash event—pre-crash, crash, and post- crash. In 2004, an estimated 40 million pas- senger vehicles were equipped with EDRs. By carefully collecting and analyzing the details provided by the growing number of EDR-equipped vehicles, state transpor- tation agencies, federal agencies, and the highway safety research community have an unprecedented opportunity to understand the interaction of the vehicle-roadside- driver system as experienced in thousands of U.S. highway crashes each year. State and federal transportation agen- cies can expect both immediate and longer term benefits from the collection of EDR data. The initial benefit for state transpor- tation agencies will be the use of EDR data from individual traffic crash investiga- tions as a powerful new form of evidence in legal proceedings, e.g., to defend against lawsuits or to recover costs of repairing collision damage to the highway infra- structure. With a more methodical system of EDR data collection, state and federal USE OF EVENT DATA RECORDER (EDR) TECHNOLOGY FOR HIGHWAY CRASH DATA ANALYSIS This digests presents the results of NCHRP Project 17-24, “Use of Event Data Recorder (EDR) Technology for Highway Crash Data Analysis.” This study developed recommendations for the enhancement of EDRs to meet the specific needs of highway crash data analysis. It also developed a rec- ommended EDR database format for agencies that seek to collect and sys- tematically store EDR data. This digest is based on a draft final report au- thored by the principal investigator, Hampton C. Gabler, and Douglas J. Gabauer and Heidi L. Newell, all of Rowan University, and Michael E. O’Neill of George Mason Law School. Subject Area: IVB Safety and Human Performance Responsible Senior Program Officer: Charles W. Niessner NATIONAL COOPERATIVE HIGHWAY RESEARCH PROGRAM

transportation agencies can expand this benefit to significantly improve the efficiency of data collec- tion for crash statistic databases. For example, in state crash databases designed to meet the Model Minimum Uniform Crash Criteria (MMUCC) for- mat, one-third (24 of 75) of the recommended data elements could be provided by EDRs. In the longer term, one of the crucial benefits of EDRs will be their influence on highway crash safety research. The ready availability of EDR data in a crash statis- tics database will enable highway safety researchers to address a number of elusive research questions that directly affect state transportation agencies, e.g., the relevancy of the NCHRP 350 crash test guide- lines for roadside safety features. RESEARCH OBJECTIVES The objectives of NCHRP Project 17-24 were (1) to recommend a minimum set of EDR data ele- ments for roadside safety analysis and (2) to recom- mend procedures for the retrieval, storage, and use of EDR data from vehicle crashes to include legal and public acceptability of EDR use. CONCLUSIONS AND RECOMMENDATIONS EDRs offer a remarkable new data source for im- provements in highway crash data analysis and re- search. However, several difficult issues may impede the use of EDR data for highway crash data analysis. These impediments include technological, legal, and consumer acceptability concerns. The following sub- sections summarize the benefits of EDR data as well as the impediments that must be overcome to use EDR data. The subsections also present recommen- dations that will permit transportation agencies and safety researchers to capitalize on the full potential of EDRs for highway crash data analysis. Benefits of Collecting EDR Data State and federal transportation agencies that collect EDR data can expect several benefits: • The initial benefit of EDR data for state trans- portation agencies will be improved investi- gation of individual crashes. EDR data are increasingly being used in the courtroom as another means of reconstructing aspects of the crash, such as vehicle speed. Many state and local law enforcement organizations already collect EDR data on a regular basis for fatal crash investigations. State transportation agen- cies will find EDR data to be a powerful new form of evidence in legal proceedings involv- ing collisions with roadside hardware—either to defend against lawsuits or to seek damages to recover costs of repairing roadside hard- ware. State transportation agencies are cau- tioned however that the use of EDR data to as- sign blame for a crash is precisely what the public finds least acceptable about EDR use. • State and federal transportation agencies that use EDR data can significantly improve the ef- ficiency of data collection for crash statistic databases. The use of EDR data can improve the accuracy of these databases and may, in the long term, reduce the costs of data collection. Based upon the methodical examination of eight existing crash databases and three rec- ommended database formats, the researchers conclude that a significant fraction of data ele- ments currently being collected could be pro- vided by either existing or future EDR data el- ements. For example, 56 of the 175 Fatality Analysis Reporting System (FARS) data ele- ments could be provided by EDRs. For state accident databases designed to meet MMUCC format, 24 of the 75 recommended data ele- ments could be provided by EDRs. • One of the crucial long-term benefits of EDRs will be their influence on highway crash safety research. The ready availability of EDR data in a crash statistics database will enable vehi- cle and roadside safety researchers to address several elusive, and often technically contro- versial, research questions. For example • How relevant are the impact conditions used in NCHRP 350? • For roadside crashes, is there a linkage be- tween vehicle acceleration and occupant in- jury? How realistic is the flail space model when evaluated against actual EDR crash pulses and hospital injury records? • Are current vehicle designs compatible with current roadside safety hardware designs? • Do impacts with soft roadside safety de- vices, e.g., crash cushions, lead to late airbag deployments? • Are advanced occupant restraint systems, e.g., dual-stage inflator systems, perform- ing as designed? 2

• How accurate are the delta-V estimates in U.S. national crash databases? • What is the distribution of impact speeds as a function of roadside object struck? • Coupling EDR pre-impact data with high- way design data, what are the relationships between highway geometric design and the probability of a runoff road event? This research project conducted an extensive re- view of the roadside safety literature, which suggests that many of the data elements recommended for col- lection by previous research studies could either be obtained with current EDR devices or in future EDR designs. Examples of critical research data needs that could be met by either existing or near-future EDRs are pre-crash vehicle trajectory, post-crash vehicle trajectory, and the orientation of the vehicle (yaw, pitch, and roll) at the time of impact. Costs of Collecting EDR Data Both startup and operational costs are associated with EDR data collection. Startup costs include the purchase of EDR data retrieval units and the train- ing of the crash investigators or law enforcement personnel who will be performing the actual EDR downloads. In addition, EDR data collection will add somewhat to the time required for crash inves- tigation. These costs are summarized below: • Purchase of EDR Retrieval Equipment. To download EDR data from crashed vehicles, state DOTs will need to purchase an EDR data retrieval unit. Currently, the only pub- licly available EDR download device is the Vetronix Crash Data Retrieval (CDR) system. At the time of this report, the list price of the Vetronix CDR system is $2,500. In addition, use of the Vetronix system requires data down- load to a portable computer or laptop; some jurisdictions may need to purchase this equip- ment as well for their investigators. • Training. State and federal transportation agencies that wish to extract EDR data, for applications such as crash databases, should anticipate the need for specialized training in EDR data retrieval. NHTSA found that a key component of a successful EDR download pro- gram is specialized EDR training for its crash investigators. In 2002, approximately half of NHTSA’s unsuccessful EDR downloads were attributed to “technical/training problems.” In 2003, after conducting specialized EDR train- ing for crash investigators, “technical/training problems” was noted as the reason for an un- successful download in only 10% of the cases. • Implementation. State crash databases and many federal crash databases (e.g., FARS and National Automotive Sampling System [NASS]/General Estimates System [GES]) are based on data extracted from police crash re- ports. The exceptions are in-depth crash data- bases (e.g., NASS/Crashworthiness Data Sys- tem [CDS] and Longitudinal Barrier Special Study [LBSS]), which are based upon data col- lection by crash investigators. In the near term, the collection and use of EDR data is unlikely to be a widespread practice in police-level crash data collection because the initial costs associ- ated with the required equipment and training may present a formidable obstacle to police de- partments. In addition, the increased time re- quired at the scene would likely render EDR data collection unacceptable to many law en- forcement agencies for routine data collection. These startup costs however are only ex- pected to be a barrier to EDR data collection in the near term. As EDR data become more widely used in the courts and as EDRs be- come more widespread in the passenger ve- hicle fleet, legal incentives to collect EDR data will grow. In many severe crashes, EDR data collection is already commonplace. Many state police fatal accident investigation divi- sions collect EDR data, whenever possible, to aid in their accident reconstructions. The understanding of severe accidents would be greatly improved if EDR data collected in fatal crashes by state police and other law en- forcement agencies were included with case submissions to the FARS database. NHTSA is encouraged to retrieve EDR data from state accident investigators when available for storage with, and enhancement of, the FARS database. Recommendations for EDR Enhancement EDRs are a rapidly evolving and, in many ways, still immature technology. Although the first research studies using EDR data are confirming their po- tential, numerous technological issues must still be 3

resolved to promote the widespread use of EDR data. Following are recommendations for needed enhancements to EDRs: • Standardize EDR Data Elements. State trans- portation agencies are encouraged to actively support the NHTSA Notice of Proposed Rule- making (NPRM) on EDRs. Until recently, there have been no standards that govern EDR for- mat. This lack of standardization has been a significant impediment to national-level stud- ies of vehicle and roadside crash safety. Re- cently, SAE J1698 and IEEE 1616 were is- sued, which prescribe industry standards or recommended practices for EDRs. To date, however, no automaker has installed, or an- nounced plans to install, production EDRs that comply with these standards. Federal regula- tion appears to be the only alternative action that will result in standardization of EDR data elements. The proposed NHTSA rule requires that EDRs voluntarily installed in light vehicles record a minimum set of specified data ele- ments useful for accident investigation, analy- sis of occupant restraint systems, and auto- matic crash notification systems. As noted in the following paragraph, the researchers also recommend that NHTSA extend the proposed rule to include data elements that will assist roadside safety research in general. Should this not be possible in the near term however, simple adoption of the NHTSA rule in its cur- rent form would greatly advance state and federal efforts to collect EDR data to improve highway crash research. • Add Recommended Data Elements. NHTSA is encouraged to extend its proposed rule on EDRs to include data elements that will assist roadside safety research in general. Based on a comparison of EDR capabilities and high- way crash data analysis needs, this project has developed a catalog of 66 data elements rec- ommended for highway crash analysis. Nearly half of these data elements are already being stored in production vehicle EDRs. Thirty- eight of these elements are defined in the NHTSA NPRM on EDRs. In the near term, the researchers recommend adopting the data ele- ments proposed in the NHTSA NPRM and adding the following four priority roadside safety data elements: (1) crash location, (2) Ve- hicle Identification Number, (3) yaw rate, and (4) roll rate. In the longer term, NHTSA should require that the entire list of recommended data elements be stored in future EDRs. • Increase Recording Duration. To capture roadside feature crash performance, automak- ers should enhance future EDRs to record for a greater length of time than is the current prac- tice. Roadside safety analyses require knowl- edge of not only the pre-crash trajectory but also the post-crash trajectory. Currently, these data could be obtained if EDRs stored pre- crash parameters, such as vehicle velocity for 5 seconds, before and after a crash. Likewise, the recording duration of the crash pulse should be increased. Impacts with roadside features, such as a guardrail, are relatively long events in comparison with vehicle-to-vehicle crashes. To capture the entire vehicle-to-roadside event, the crash pulse should be recorded for a mini- mum duration of 300 milliseconds. This rec- ommendation is consistent with the NHTSA NPRM on EDRs. • Increase Number of Events Recorded. Au- tomakers should enhance EDRs to record a minimum of three crash events. EDRs that record only a single event (e.g., the current Ford design) lose approximately one-half of the events. EDRs that record only two events (e.g., the current GM design) lose approxi- mately 17% of the events. An EDR that records three events, on the other hand, would cap- ture 94% of the crash events. This recommen- dation is consistent with the NHTSA NPRM on EDRs. • Expand the Definition of an Event. Auto- makers are encouraged to extend the defini- tion of an “event” to include roadway depar- tures. Currently, an event is a crash. In addition to crashes, roadway departure, with or without an impact, is also an important event. Lane- keeping and roadway departure warning sys- tems, which could be adapted for this purpose, are now entering the market. Accurate record- ing and retrieval of roadway departure events would be invaluable for encroachment studies. Recommendations for Improved EDR Data Retrieval and Archival Methods Currently, there is no standardized method to download data from EDRs. Similarly, there is no stan- 4

dardized format for storing EDR data in a crash statis- tics database. The following actions are recommended to alleviate these obstacles to implementation: • Standardize the EDR Retrieval Method. The state DOTs should actively support the proposed NHTSA EDR rule that mandates automakers make the contents of their EDRs accessible with publicly available tools. Cur- rently, there is no standardized method to download data from EDRs. Two automakers have awarded an exclusive license to the Vetronix Corporation to market an EDR re- trieval tool for their EDRs. The remaining au- tomakers use proprietary tools for EDR data re- trieval—effectively preventing EDR access by either state or federal transportation agencies. • Require a Crashworthy, Universal EDR Download Connector. NHTSA is encouraged to modify or extend its proposed rule on EDRs to require a uniform connection point for EDR download. NHTSA has found that, in a signif- icant percentage of crashes, crash investigators were unable to use the OBD-II port, the pri- mary Vetronix access point, to access the EDR data. Although investigators have the option to directly connect to the EDR, direct connection requires the partial dismantling of the crashed vehicle. Furthermore, direct connection re- quires the purchase of large numbers of dif- ferent EDR connection cables because there is no universal EDR connector. The researchers recommend that NHTSA either require a crash- worthy OBD-II connection to the EDR or mandate a universal connector for direct con- nection to the EDR. • Automate the Export of EDR Data. Vetronix, producer of the only publicly available EDR download tool, is strongly encouraged to modify its CDR software to allow electronic export of EDR data to crash databases such as NASS/CDS. Currently, all EDR data must be manually transcribed from Vetronix CDR screens into a database—a tedious and error- prone process. Vetronix is however developing a CDR-to-XML conversion program that has promise for federal and state DOTs with exist- ing or planned EDR databases. Vetronix is en- couraged to release a production version of this program to improve data entry efficiency and accuracy for mass users of EDR data, e.g., fed- eral and state transportation agencies. • Use the Recommended EDR Database For- mat. State and federal transportation agencies seeking to create an EDR database are en- couraged to use the recommended EDR data- base format developed by this project. The standardized EDR database was designed to (1) accommodate data from diverse existing EDR download formats including all publicly released GM and Ford formats, (2) store the fu- ture EDR data elements needed to comply with the NHTSA NPRM on EDRs, and (3) store the recommended list of data elements for high- way crash data analysis developed by this re- search project. Legal Acceptability of EDRs Although the preceding technological issues are challenging, they are solvable. More uncertain are the concerns that have been raised about the legal ac- ceptability of the widespread collection of EDR data. A special study was conducted as part of this research project to explore the legal issues surrounding the im- plementation and use of EDRs. The special study ad- dressed four specific issues: (1) whether the Fourth Amendment to the U.S. Constitution in any way bars the collection of data recorded by EDRs, (2) whether the U.S.DOT has the authority to mandate the instal- lation of EDRs in all new vehicles, (3) whether the data recorded by EDRs is admissible in court, and (4) whether the collection of such data violates pri- vacy rights. The report’s conclusions follow: • With respect to Fourth Amendment concerns, the police (or other government crash investi- gators) may properly seize EDR devices (or otherwise collect the EDR data) without a war- rant during post-crash investigations. This au- thority is premised upon one of two legal is- sues: either seizure of a required safety device does not constitute a search implicating the Fourth Amendment or, alternatively, seizure of a safety device qualifies under the exemptions for conducting a warrantless search. The po- lice’s authority to conduct warrantless searches may be affected by how soon after the crash the search occurs: the more immediately the search occurs following the crash, the greater the officers’ authority to conduct a warrant- less search. Absent a crash, however, police may not be able to seize such data routinely without either a warrant or express legislative 5

authorization (unless current thought changes on the individual’s reasonable expectation of privacy regarding EDR data). Of course, po- lice should have little trouble in obtaining a warrant to seize EDR data (or even the device itself). • U.S.DOT may require the installation of devices that demonstrably improve highway safety or advance some other significant pub- lic policy interest. The public policy interest in installing EDRs seems beyond peradventure. As a consequence, the U.S.DOT presumably enjoys the authority to mandate the installation of such devices on new automobiles. • Although the data (and the recorder itself) may be “owned” by the automobile’s owner or lessee, that data may almost certainly be used as evidence against that owner (or other driver) in either a civil or a criminal case. Cer- tainly nothing within the Federal Rules of Ev- idence (FRE) or the Fifth Amendment’s pro- tection against compelled self-incrimination would exclude the use of data recorded by the EDRs. Similarly, owners might be prohibited from tampering with the data if litigation is pending. • Finally, the issue here is not one so much of legal authority to use EDR data in court, but instead what the public will accept. Although the statutory authority to require EDRs may exist, the public may not want a device in- stalled in their automobiles that appears to en- croach upon their personal privacy interests. Understood in this way, the problem is less a legal concern than it is a battle to mold public perception. Not every life-saving device that is deployed with the best of intentions will be accepted by the public. Personal privacy and public safety must exist within the same sphere. Occasionally, respecting privacy rights will mean that harmful things may occur, but this is the cost of living in a free society. Public Acceptability of EDRs Paralleling the concerns over legal acceptability of EDRs are concerns over public acceptability. A consumer revolt against the installation of EDRs could negatively impact sales and/or lead many man- ufacturers to offer owners the option to turn off their EDRs or even to stop installation of them altogether. These options would seriously limit the amount of EDR data collected for research by personnel in law enforcement, insurance, government, manufacturing, and education. A special study was conducted as part of this re- search project to determine the public awareness and level of acceptance of EDRs. The consumer accept- ability study was conducted in two phases. In the first phase, a questionnaire, designed for this study, was mailed to 10,000 licensed drivers. In the second phase, focus groups were conducted, with a smaller sample of licensed drivers, to follow up on the sur- vey results. The survey results provided several key find- ings. A significant majority of all respondents were unaware of EDRs and their use. Most felt that EDRs would be beneficial in accident investigations, lower insurance rates for safe drivers, and encourage mon- itored drivers to behave more safely. Respondents expressed a preference for the use of EDRs to be op- tional and to maintain control of the data. The opin- ions expressed were reasonably consistent across demographic groups, but older, more affluent, Cau- casian males were more likely to report concerns about control of EDR data and their use. The results from the focus groups matched well with those obtained through the survey and from other related studies in the literature. Many of the par- ticipants were unaware of EDRs until contacted about participating in the focus groups, which matched the results of a survey that indicated most Americans are unaware of EDRs. Both survey and focus group par- ticipants felt that the devices could prove useful in accident investigations. Survey and focus group opinions mirrored what GM found in a recent sur- vey: self-labeled safer drivers are more accepting of EDRs than those who admitted to some minor traf- fic “transgressions.” Finally, as the results of this project’s survey and focus group are in line with the results of similar surveys, clearly the gathered opin- ions reflect the general population’s opinions, some of which are based on misconceptions. To alleviate public concerns about EDRs, those organizations or agencies wanting to use EDR data should promote education about what EDRs can ac- tually do. The automotive manufacturers, U.S.DOT, and state transportation agencies are encouraged to 6

conduct a more thorough public education campaign to inform the public about the presence of EDRs in passenger vehicles and about the safety and research benefits for the motoring public. REPORT AVAILABILITY The complete report for NCHRP Project 17-24 is available on TRB’s website as NCHRP Web-Only Document 75. ACKNOWLEDGEMENTS This study was conducted under NCHRP Proj- ect 17-24, “Use of Event Data Recorder Technology for Highway Crash Data Analysis.” It was guided by NCHRP Project Panel 17-24, consisting of Mack O. Christensen (chair), Richard R. Peter, Augustus Chidester, Keith A. Cota, Deborah Freund, Tom Kowalick, King K. Mak, Sarah McComb, Edward O’Hara, Kenneth R. Stack, Martin W. Hargrave, John A. Hinch, and Richard Pain. 7

These digests are issued in order to increase awareness of research results emanating from projects in the Cooperative Research Programs (CRP). Persons wanting to pursue the project subject matter in greater depth should contact the CRP Staff, Transportation Research Board of the National Academies, 500 Fifth Street, NW, Washington, DC 20001. Transportation Research Board 500 Fifth Street, NW Washington, DC 20001

PRSRT First Class U.S. Postage PAID Washington, DC Permit No. 8970 Transportation Research Board 500 Fifth Street, NW Washington, DC 20001 These digests are issued in order to increase awareness of research results emanating from projects in the Cooperative Research Programs (CRP). Persons wanting to pursue the project subject matter in greater depth should contact the CRP Staff, Transportation Research Board of the National Academies, 500 Fifth Street, NW, Washington, DC 20001.