BACKGROUND
Quantitative safety analysis is a relatively new field, and crash prediction modeling and other data-driven safety analysis procedures traditionally focus on the project level within state departments of transportation (DOTs). Many decisions affect safety or constrain safety improvement options later in a project’s lifecycle, and these decisions could benefit from quantitative safety analysis. Macro-level crash prediction models (CPMs) are a quantitative safety analysis method that predict an average crash frequency, by crash type and severity, for a defined area, such as a census block group, tract, traffic analysis zone, or county.
NCHRP Research Report 1044: Development and Application of Quantitative Macro-Level Safety Prediction Models describes macro-level CPMs designed for a quantitative consideration of safety in planning-level decisions. This supports a proactive and comprehensive approach to addressing safety needs for all road users. Macro-level CPMs were developed for inclusion in the AASHTO Highway Safety Manual, second edition, in a chapter titled, “Areawide Approach to Roadway Safety Management.” The development process involved utilizing state- and metropolitan planning organization (MPO)-level data from three MPOs (Richmond, VA; St. Louis, MO; and Chicago, IL) and two states (Illinois and Utah).
NCHRP Research Report 1044 observed a high level of consistency between the individual agency models, as well as the ability to integrate data and models into a combined model that incorporated more than one agency. Supplementing NCHRP Research Report 1044 are (a) a spreadsheet tool to support the testing and implementation of the macro-level CPMs; and (b) NCHRP Web-Only Document 348: Macro-Level Analysis of Safety Planning and Crash Prediction Models: A Guide, which provides guidance on how to use the spreadsheet tool.
There is a need to provide technical guidance and training to state DOTs to implement the CPMs from NCHRP Research Report 1044 to address safety needs across diverse scenarios.
OBJECTIVES
The objectives of this research are to (a) provide technical help to state DOTs to implement the macro-level CPMs from NCHRP Research Report 1044, (b) conduct pilot projects with state DOTs, and (c) develop supplemental materials for the macro-level CPMs based on the results of the pilot projects.
Accomplishment of the project objective(s) will require at least the following tasks.
TASKS
PHASE I
Task 1. Identify at least four state DOTs to participate in pilot projects of the macro-level CPMs. This project will focus on CPMs relevant to state DOTs, but state DOTs can invite their MPO partners to participate if feasible. The NCHRP will make the final determination of which states will participate in this research project.
Deliverable: A brief memorandum on the identified states and the process used to identify them.
Task 2. Working with each state DOT, use a sample dataset to validate the macro-level CPMs and spreadsheet tool, and document this validation in individual case studies. Validation includes activities such as verifying availability and integrity of data; determining how the data can be used in the existing models; and identifying concerns related to crash severity, coding, and socioeconomic indicators that may have changed since the previous research was completed. Data should be obtained from multiple sources and integrated into one dataset according to the methods prescribed in NCHRP Research Report 1044. Propose supplemental materials to the NCHRP Research Report 1044 spreadsheet tool.
Deliverable: At least 4 case studies summarizing the validation process and findings and a report on the proposed supplemental materials.
Task 3. Conduct a virtual workshop with participating agencies to review their data development processes. Information to be shared includes keys to success, challenges, lessons learned, and feedback on the data development processes from participating agencies and the project team. The workshop will also serve as an opportunity to discuss proposed supplemental materials to the macro-level CPMs and spreadsheet tool.
Deliverable: A brief memorandum summarizing the virtual workshop.
Task 4. Prepare an interim report to document Tasks 1 through 3 and provide an updated work plan for the reminder of the research. The updated work plan must include calculated goodness-of-fit measures to determine if Phase II should use calibrated models or agency-based models.
Deliverable: Interim report.
PHASE II
Task 5. Working with staff from each pilot state, apply the models and spreadsheet tool to data at the statewide level according to the interim report.
Deliverable: A brief memorandum summarizing the results of Task 5.
Task 6. Using the results of Task 5, develop a process to prioritize areas for further safety analysis for each pilot project. Develop transferability guidelines and recommendations for how states can bridge the gap from model outputs to identifying priority areas. Transfer compiled data for macro-level CPMs and results to the participating agencies.
Deliverables: Documentation of processes, transferability guidelines, and recommendations.
Task 7. Conduct a virtual peer exchange workshop with participating agencies to review results and discuss project findings. Discussions in the peer exchange will focus on experiences learned using the macro-level CPMs. This peer exchange will also help refine implementation best practices to supplement the existing research.
Deliverable: A brief memorandum summarizing the virtual peer exchange.
Task 8. Prepare final deliverables including:
- A conduct of research report that documents the entire research effort and case studies of each pilot;
- A PowerPoint presentation of project results to be used at conferences and in webinars;
- A technical memorandum that (a) describes how state DOTs can use macro-level CPMs for further safety analysis, (b) identifies processes state DOTs can use, and (c) documents barriers to and lessons learned from implementation; and
- Any other materials developed during the course of this project.
STATUS: This project is active.
