Exposing the Dangers of Distance: Mining Crash Narratives to Explore Why Pedestrians Face Severe Injury and Death Far From Home

Outcomes/Impacts

This project will generate new knowledge, tools, and methods that directly inform the design and implementation of safer transportation systems for pedestrians. By integrating structured crash data with narrative-based insights, the study moves beyond traditional analyses that focus solely on crash frequency and severity. Instead, it identifies why pedestrians are traveling in high-risk environments, what purposes their trips serve, and how broader shifts in urban form and demographics influence crash exposure. These insights create an evidence base that can guide more targeted, context-specific safety strategies at both the neighborhood and city levels.

A central output of this work is the development of an AI-driven classification and analysis system for crash narratives. Hosted locally to protect sensitive information, this system will allow researchers, agencies, and practitioners to systematically extract latent information from narrative records (e.g. trip purposes, environmental conditions, or contextual circumstances) that is rarely accessible through conventional crash databases. This tool represents a significant methodological advancement for transportation safety research. Automating narrative analysis enables large-scale, reproducible assessments that would otherwise be infeasible for human analysts. Over time, this system can be adapted for other states or integrated into federal crash reporting practices, potentially influencing how data is collected, processed, and used nationwide.

Another significant contribution is the home-based classification framework for pedestrian crashes, which uses both residence locations and urban structure (city centers, suburban peripheries, etc.) to categorize crash contexts. This system will allow practitioners to distinguish between near-home and far-from-home crashes, and to identify patterns linked to suburbanization, commuting, or limited access to essential services. Such insights can directly inform local planning decisions, from where to prioritize crosswalks and traffic calming to how transit stops and pedestrian access routes are designed.

In terms of policy and practice, the findings will provide actionable recommendations aligned with safety principles, focusing on arterials where pedestrians are most at risk. Results can inform the prioritization of roadway redesign projects, the placement of pedestrian infrastructure (e.g., midblock crossings, lighting, refuge islands), and targeted enforcement or education strategies. By highlighting where and why crashes occur farther from home, the study also equips agencies to anticipate and mitigate risks associated with population shifts, suburban expansion, and changes in travel behavior.

The anticipated impacts extend across multiple dimensions of the transportation system. In safety terms, the outputs support reductions in severe and fatal pedestrian crashes, particularly in suburban contexts where risks are escalating. Beyond direct infrastructure applications, this research has the potential to shape state and federal policy discussions around pedestrian safety, data collection practices, and roadway design standards, ultimately contributing to a safer, more resilient transportation system for all users. Lastly, the peer-reviewed journal publication will ensure the research findings are rigorously validated, widely disseminated, and positioned to influence future transportation safety research and practice.