Enhancing Traffic Safety Data
Challenge
NHTSA wants to transform fragmented crash data into a unified resource for safety analysis.
Comprehensive, timely, and accessible crash data are critical in advancing effective traffic safety initiatives and policy decisions. However, the National Highway Traffic Safety Administration (NHTSA) faces significant challenges with its existing crash data systems, including time lags in data availability, variations in implementation of data standards across jurisdictions, and difficulties integrating supplemental data sources. These challenges hinder the ability to promptly identify emerging crash trends and patterns, limiting the effectiveness of responsive policymaking and countermeasure development.
To maximize the utility of crash data for safety improvements, NHTSA needs to enhance its current data processing capabilities, implement advanced statistical and geospatial analysis techniques, and create more robust, user-friendly, and analytically powerful systems that serve both internal needs and those of state highway safety offices and other transportation agencies.
Solution
NORC and its partners will use advanced analytics and visualization techniques to unlock insights hidden in crash data.
Over the last decade, NHTSA has engaged in a concerted data modernization effort to improve its crash data collection systems and analytic capabilities. To support that effort, NORC has partnered with the Virginia Tech Transportation Institute (VTTI), the largest transportation research institute in the United States, to deliver a comprehensive solution for NHTSA’s crash data challenges. Our approach includes:
- Data Imputation and Quality Enhancement: Developing sophisticated techniques to handle missing values across different crash datasets, dramatically improving the completeness and reliability of analyses
- Data Integration: Integrating supplemental data sources beyond the NHTSA to provide deeper insights and uncover relationships that may otherwise remain hidden, enabling a more holistic understanding of crash-related factors and outcomes
- Advanced Computational Methods: Implementing hierarchical forecasting models and machine learning algorithms to improve crash projections and anomaly detection capabilities
- Geospatial Analysis Tools: Creating specialized methods to understand spatial patterns of crashes and their relationship to road infrastructure, allowing for targeted safety improvements
- Interactive Dashboards and Query Tools: Designing user-friendly interfaces customized for different stakeholder groups, from NHTSA analysts to state department of transportation officials and the public
- Statistical Surveillance Systems: Building monitoring systems that can automatically detect emerging crash trends, hotspots, and anomalies in near real-time
Our partnership leverages VTTI’s established relationship with NHTSA and decades of experience analyzing diverse safety data, including NHTSA’s National Center for Statistics and Analysis (NCSA) data and state crash databases. Combined with NORC’s innovative statistical approaches and data visualization expertise, we’re developing tools that will transform how stakeholders access, understand, and utilize crash data to inform safety policies.
Result
The data tools we develop will empower decision-makers with accurate, accessible information to improve roadway safety.
NORC’s data imputation methodologies will substantially improve the completeness and reliability of NHTSA’s crash data, particularly for critical elements like blood alcohol concentration and roadway surface information. The implementation of advanced forecasting models will enable more accurate and timely projection of crash trends, allowing NHTSA to identify emerging safety concerns much earlier than previously possible.
The interactive dashboards we are developing will transform how NHTSA and state transportation agencies access and utilize crash data. State departments of transportation will have customizable tools to analyze specific safety concerns in their jurisdictions, while NHTSA can more effectively monitor national trends. These tools will be instrumental in identifying high-risk locations and driver behaviors, directly informing the development of countermeasures and policy initiatives.
Our comprehensive documentation and training materials will ensure the long-term sustainability of these enhancements, allowing NHTSA to maintain and expand these capabilities as new data sources become available. The project establishes a foundation for continuous improvement in traffic safety analysis that will benefit the nation’s road users for years to come.
Related Tags
Project Leads
-
Don Jang
Vice PresidentCo-Principal Investigator -
Ned English
Associate DirectorSenior Staff -
Stas Kolenikov
Principal StatisticianSenior Staff -
Kiegan Rice
Senior StatisticianSenior Staff -
Weihuang Wong
Principal Data ScientistSenior Staff -
Josiane Bechara
Senior Research MethodologistSenior Staff
External Partners
- Feng Guo, Co-Principal Investigator (VTTI)
- Shu Han, Project Lead (VTTI)
- Michael Porter, Consultant (University of Virginia)