The National Academies

NCHRP 17-126 [Pending]

Crash Prediction Models for Alternative and Unconventional Intersections

  Project Data
Funds: $750,000
Contract Time: 36 months
Staff Responsibility: Anne-Marie Turner


In 2010, AASHTO published the first edition of the Highway Safety Manual (HSM). HSM Part C includes methods to assess the safety performance of existing highway facilities, predict the safety performance of new facilities, and estimate the change in safety performance for proposed improvements. The forthcoming second edition of the Highway Safety Manual (HSM-2) expands crash prediction models for analyzing safety performance to include additional intersection and facility types. However, not all intersection types are addressed in the HSM or HSM-2; intersection types not addressed include

  • Intersections with frontage roads,
  • Protected intersections,
  • Offset intersections,
  • Quadrant intersections,
  • Through-cut intersections,
  • Median U-turn intersections,
  • Jughandle intersections,
  • Continuous green tee intersections,
  • Rural five-leg intersections, and
  • Six-or-more-leg intersections.

Research is needed to develop crash prediction models for additional intersection configurations and traffic control types that have not been previously studied for potential inclusion in future editions of the HSM. New crash prediction models need to be developed so comparisons can be made to the safety performance of intersection types in the latest edition of the HSM.


The objective of this research is to develop crash prediction models that are compatible with existing methods in HSM Part C and include a range of intersection configurations and traffic control types. 

Accomplishment of the project objective will require at least the following tasks.


PHASE I: Preliminary Research

Task 1. Conduct a literature review of relevant research and guidelines related to crash prediction models for intersection types neither in the HSM nor currently being studied by other NCHRP projects. The review shall include published and unpublished research conducted through the NCHRP, the FHWA, and other academic sources.

Task 2. Analyze and synthesize the results of the literature review to identify knowledge gaps and opportunities as related to the project objective. These gaps must be addressed in the final product or the recommended future research. 

Task 3. Develop a survey for state departments of transportation and other agencies. The draft survey and distribution list must be approved by the NCHRP before distribution. The survey shall address, at a minimum

  • Commonly proposed and existing types of alternative and unconventional intersections,
  • State and local needs for crash prediction models at intersection types not currently addressed in the HSM,
  • Attributes of crash prediction models that are most valuable, and
  • Identification of potential data collection sites.

Task 4. Develop methods to achieve the project objective to be executed in Phase II. Develop a prioritized list of at least eight intersection types, from which the NCHRP will select four or more to be addressed in Phases II and III of this study. At a minimum, development of the prioritized list and methods should consider the following factors

  • Intersection types identified in previous tasks;
  • The sufficiency and limitations of available data;
  • Methods for developing crash prediction models and crash modification/adjustment factors, including crash frequency, severity, and type;
  • Data collection methods;
  • Both urban and rural area types;
  • Both signalized and unsignalized intersection types;
  • Strategies for incorporating multimodal road users;
  • Strategies to consider crash exposure;
  • Strategies to validate crash prediction models;
  • Requirements of a spreadsheet-based tool(s); and
  • The potential for study results to be applicable nationwide.

Task 5. Prepare Interim Report No. 1 that documents Tasks 1 through 4 and provides an updated work plan for the remainder of the research. 

PHASE II: Development of Crash Prediction Models 

Task 6. Collect data and refine methods for developing crash prediction models. 

Task 7. Execute the methods to develop crash prediction models according to the approved Interim Report No. 1 as refined in Task 6. 

Task 8. Validate the proposed crash prediction models per the approved methods and document the model development and validation efforts. Organize relevant data with relevant data documentation in an electronic format.

Task 9. Refine and update the crash prediction models based on validation results and develop the implementation tool and supporting materials. 

Task 10. Develop a strategy and materials for conducting a workshop, pilot tests, or other method(s) to demonstrate the use of the models, applicability to practical use cases encountered by transportation agencies, and implementation support needed.

Task 11. Prepare Interim Report No. 2 that documents Tasks 6 through 10. Provide relevant data with documentation in an electronic format and the implementation tool with supporting materials.

PHASE III: Final Deliverables

Task 12. Execute the strategy developed in Task 10. Summarize successes and shortcomings any of proposed refinements to the tool(s), method(s) and other materials for practitioners.

Task 13. Prepare draft language for consideration by AASHTO for inclusion in its next update of the HSM. The draft content is due 6 months before the end of the contract.

Task 14. Prepare final deliverables, including

  • A research report, including an executive summary, that documents all research and project efforts;
  • Draft language for consideration by AASHTO for inclusion in its next update of the HSM;
  • An implementation tool(s) (including any annotated spreadsheet macro code, if used), user manual(s), and any other documentation;
  • All raw and cleaned data collected and used in this research (data should be provided in as close to its raw form as possible, based on contractual or legal restrictions, and must be organized and submitted in an electronic format): input data sets, fused and integrated research data sets, data dictionaries, data models, etc.;
  • A PowerPoint presentation with speaker notes that summarizes the project and distinctly illustrates for the audience how the research can be applied in their organization; and
  • A technical memorandum titled “Implementation of Research Findings and Products”.

The contractor has been selected. Contract pending.

To create a link to this page, use this URL: http://apps.trb.org/cmsfeed/TRBNetProjectDisplay.asp?ProjectID=5527