How do I get access to the Safety Data?
You can obtain the Safety Data through the websites maintained by TRB’s contractors.
A team at Virginia Tech Transportation Institute is under contract to TRB to provide user support for the NDS data, and a team at Iowa State University is under contract to TRB to provide user support for the RID.
Naturalistic Driving Study
Roadway Information Database
What costs are associated with accessing the data?
Why are there access restrictions?
What is a Secure Data Enclave?
What is a Certificate of Confidentiality?
What is an IRB?
What Are Original Data and De-Identified Summary Data?
Do Any Similar Restrictions Apply to the Roadway Information Database?
Naturalistic Driving Study
The InSight Website provides information about the Naturalistic Driving Study and the types of data available. It also allows researchers to query the data to see how much might be available to answer their particular research questions. A data viewer function in InSight displays a small port of real data from the study that is de-identified. Also available is a downloadable training dataset with non-personally identifying data that mirrors the “real” data.
A variety of functions are available to users on InSight, including:
• Driver Descriptions and Assessments. Summary graphs and detailed records of driver assessments are provided addressing driver demographic background, physical, psychological, and medical condition.
• Custom Query Capability. Users can build custom queries to search for records matching criteria that span multiple datasets.
• Summary of Continuous Naturalistic Driving Data Collected. Graphs and detailed records describe data collection progress and characteristics of trips collected during the study.
• Naturalistic Driving Study Background Information. Users are able to access an overview of the SHRP 2 Naturalistic Driving Study project, data collection procedures, data dictionaries, and sample data.
• Vehicle Descriptions. Summary graphs and detailed records describe the types of vehicles involved in the study.
• Access to SHRP 2 NDS Forums. Users can join a community of SHRP 2 NDS Forum members to discuss available data, website functionality, and related topics.
InSight uses an authentication process to ensure that all researchers requesting to view the data are familiar with the ethical concerns associated with human subjects data. Once on the web site, users can establish an account, which provides basic access. Users who would like to gain additional functionality and feature access can choose to take on-line human research subjects training and, upon passing a quiz and providing an electronic certificate, become a “Qualified Researcher”. This training typically takes about one hour.
While InSight allows people to examine the NDS data, only a limited amount of the entire database is viewable on the website and there is no capability to download data. Data download requires a data use license (DUL), under which terms qualified researchers may access and work with the data. The licensee agrees not to make copies of the original data, not to share with unauthorized individuals, and to maintain the original data in accordance with data security protocol.
There are four licenses which represent the most common use cases.
1. InSight only
2. Export of data without personally identifying information (PII)
3. Secure data enclave use for data with PII
4. Algorithm run within the database
In the data use license, the requesting researcher provides details including the data specifications, proof of Institutional Review Board (IRB) approval, institutional signatures, and a data security plan. A reduced dataset without PII is then assembled for the licensee. Or if secure data enclave use is authorized, the user visits the secure data enclave established at Virginia Tech Transportation Institute to view and analyze personally identifying information and then export non-PII data.
Visit the InSight Website to explore what data is available and to contact the SHRP2 Naturalistic Driving Study database operator to start the data use licensing process.
Roadway Information Database
The Roadway Information Database Website provides information about how the Roadway Information Database was collected and designed, the types of data elements available, and metadata and data dictionaries. Also available is a downloadable sample Geodatabase and metadata file from Tampa, FL.
Interested users should download a sample data from the website and familiarize themselves with the available data (the mobile data, existing data from states, and supplemental data). Once the roadway attributes needed are identified, the database operator can assist users on deciding how to proceed. The user will need to complete a Terms of Use Agreement before receiving a copy of the RID or a reduced dataset.
Visit the RID Website to explore what data is available and to contact the SHRP2 Roadway Information Database operator to obtain a copy of the data.
What costs are associated with accessing the data?
There is no cost to use the InSight website. InSight makes a portion of the entire database available for users to view. There little to no cost associated with obtaining an export of data elements viewable on InSight under a data use license.
There is an associated cost for time and labor to export a reduced dataset of data elements not posted on InSight, to run an algorithm in the enclave and export results, or to utilize the secure data enclave for analysis. Costs vary greatly with the complexity of the data request. Visit the InSight website for more information on example costs for past and current research projects.
Why are there access restrictions?
The Naturalistic Driving Study data include sensitive and personally identifiable information (PII) such as face video and GPS traces (including home, work, and school locations). In addition, the data are considered human subjects data and must be protected in accordance with federal law. All privacy protections promised to participants in the research protocol and consent forms regarding their data continue even after the study ended. Still, the data need to be made widely available to researchers in order to maximize the financial and personal investment made by the sponsors and the participants.
In order to both uphold the privacy of study participants and promote the use of the data in important research, a process was created for researchers to establish a data use license (DUL) under which terms they may work with the data. In many cases, Institutional Review Board (IRB) approval of a qualified researcher’s plans for data use will be required due to the involvement of human subjects.
Unlike many databases typically used in transportation safety research, the SHRP 2 safety databases were developed using human research subjects. Studying human subjects—as opposed to using more traditional sources such as police report-based crash records—provides a much better opportunity to understand how drivers interact with their vehicles, other motorists, roadways, and environmental factors including weather.
As human subjects, all SHRP2 Naturalistic Driving Study participants were promised confidentiality in exchange for participation in the study and recording of their driving data, including various video streams inside and outside the vehicle. The SHRP 2 NDS data include some items that could be used to identify the human subjects who participated in the NDS. The most commonly used PII data elements include driver face video; full trip GPS traces which can be used to identify a person’s home, work, and school locations; and unaltered forward video of a crash. There are other less commonly used yet potentially identifying data elements. The operators of the NDS database have an ethical obligation to take that promise of participant confidentiality seriously. Data elements that contain any personally identifying information (PII) is available only within a secure data enclave (SDE) to protect the privacy of study participants. Qualified researchers who wish to view and analyze PII must meet eligibility criteria and agree to the requirements of a data use license (DUL).
What is a Secure Data Enclave?
A secure data enclave (SDE) is a physically isolated environment in which carefully controlled access to restricted data can be provided to qualified researchers. A physical SDE is a room or set of rooms which meets a set of security rules so that restricted data cannot be taken out of the room or copied. A remote (or virtual) SDE accomplishes this same level of security via electronic and physical means. There is currently one physical Secure Data Enclave available in which PII from the Naturalistic Driving Study can be used. That SDE is at the Virginia Tech Transportation Institute (VTTI) in Blacksburg, Virginia. However, the Transportation Research Board and VTTI are now working with the Turner-Fairbank Highway Research Laboratory outside Washington, DC to develop and pilot test a remote secure data enclave. That remote SDE, if successful, could be the model for other remote secure enclaves in the future, subject to availability of funding.
Use of any secure data enclave is subject to certain restrictions and scheduling constraints. The cost of secure data enclave use may be recovered from data licenses.
What is a Certificate of Confidentiality?
Research subjects who participated in the SHRP2 NDS are protected by a Certificate of Confidentiality issued by the Department of Health and Human Services (HHS) in accordance with the provisions of section 301(d) of the Public Health Service Act (42 U.S. C., 241(d)). All data licensees are considered to be a cooperating agency under the terms of the Confidentiality Certificate; as such, all Data Licensees, Principal Investigators, and Research Staff are required to protect the privacy of the individuals who are the subjects of SHRP2 NDS by withholding their identifying characteristics from all persons who are not signatories to a data use license. “Identifying characteristics” are considered to include those data defined as PII under the terms of the data use license.
What is an IRB?
An institutional review board (IRB), also known as an independent ethics committee (IEC), ethical review board (ERB) or research ethics board (REB), is a committee that has been formally designated to approve, monitor, and review research involving humans. In the case of the NDS, VTTI manages the review and approval process for new and amended data use licenses under authority granted by the Virginia Tech IRB and the Transportation Research Board. The National Academies of Sciences, Engineering, and Medicine IRB is also involved; this IRB does not review individual data use licenses, but does periodically review the process used to approve licenses to make sure that human subjects are protected. Finally, each data licensee needs to gain approval from the IRB (or equivalent in the case of non-US licensees) that governs human subject research for their institution.
What Are Original Data and De-Identified Summary Data?
The SHRP 2 NDS data include both original data (collected directly from the participant at enrollment, about the vehicle at installation, from each trip as the driver conducted normal daily driving, and about crashes via post-crash interviews) and de-identified summary data. The research protocol that was negotiated between the Transportation Research Board and the Institutional Review Boards states that any original data must be destroyed by any licensing agency at 30-40 years after data collection ended (depending on the type of data) and thus these data must be tracked. (Data use licenses being issued now typically call for the return or destruction of original data at two years past project completion, which is usually long enough to complete a research project and generate publications.) This sort of limitation on the life of original data is very common in situations where human subjects data are being used for research, for instance in medicine.
De-identified data have been processed to prevent participants’ identities from being disclosed. Summary data have been transformed in a significant way, for instance being presented as an average or standard deviation instead of original data. De-identified summary data from the NDS may be kept indefinitely. Most projects will request a combination of original and de-identified summary data. For these reasons, projects have a default licensee data retention length of two years past project completion. De-identified summary data (which cannot be reverted to original data by transformation) produced by the licensee as part of their work processes on this DUL may be retained and used by the licensee without licensing agency or data steward constraints.
Do Any Similar Restrictions Apply to the Roadway Information Database?
The Roadway Information Database (RID) that is available to complement the NDS contains no personally-identifying information. No research protocol was approved by an IRB for the RID since one was not required by Federal laws. Therefore the RID data can used with far fewer limitations than the NDS data and they do not have to be destroyed after 30 or 40 years. A simple terms of use agreement is required to track the use of the RID and to help prevent version control issues that might accompany unlimited copying of the data.