2018 Research Project
Developing a Taxonomy of Human Errors and Violations that Lead to Crashes
Principal Investigator
Asad J. Khattak
University of Tennessee, Knoxville
View Bio
Co-Investigator
Eric Dumbaugh
Florida Atlantic University
View Bio
Full Report
Project Slide Deck
Research Brief
Summary
Human errors and violations are highly relevant to the safe systems approach as human error tends to dominate crash occurrence, contributing to 80%-90% of crashes. A better understanding of “critical reasons for the critical pre-crash events” has significant potential in reducing deadly behaviors on roadways. A key gap in the literature relates to the origin of the different types of human errors, e.g., whether they begin with intentional actions or unintentional actions, and how they relate to the built environment.
The project will be organized into the following tasks:
Task A: Review relevant work that will be integrated into a taxonomy: As part of identifying gaps in the relevant literature, we will identify data limitations that may pose a threat to study results. After we develop a taxonomy based on NDS data, we will match the potential countermeasures to the crash contributory factors at hand, based on insights from the literature.
Task B: Process and prepare the naturalistic driving study database and develop framework: In addition to theoretical developments, new data analytic methods will be used to extract valuable information about errors and violations and their respective mechanisms from driving and crash data. While a substantial amount of the work will be conceptual, data from SHRP Naturalistic Driving Study (NDS) will be used.
Task C: Develop a methodology to classify crash-contributing errors: The project will explore the role of human cognition, information acquisition, processing and use, and rational behavior as they relate to human errors in crashes. A classification of the full range of errors and violations can help focus on predominant error types and facilitate in the design of effective prevention and mitigation tools and strategies.
Task D: Develop a “safety matrix” that quantifies the contributions of different factors for different scenarios: In this task, we will develop safety matrices to quantify the proportions of different crash-contributing factors. The relationships between error-producing environments and errors recorded in crashes will add a new dimension to our existing understanding of errors.
Task E: Apply a modeling approach for analyzing relationships: The overall framework would be a Structural Equation Modeling (SEM) approach. Correlations between human errors and various factors will be quantified.
Task F: Use the safety matrix to explore implications for road safety in the future: In this task, we will explore the implications of the results for improving current and future road safety. The aim is to expand our knowledge-base for implementation of more informed decisions about safe vehicles, safe people, safe speeds, and safe environments.
Task G: Final report and dissemination: After completion of the project, the research team will document the results so they can be used by engineers and planners in the future.
Project Details
| Project Type: | Research |
| Project Status: | Active |
| Start Date: | 5-1-2018 |
| End Date: | 1-31-2021 |
| Contract Year: | Year 2 |
| Total Funding from CSCRS: | $79,722 |
| Co-sponsor: | Tennessee DOT |
| Collaborating Organizations: | Florida Atlantic University; University of North Carolina, Chapel Hill |
