At the Transportation Research Board (TRB) Annual Meeting, our research team presented findings from our project for the 2025 INRIX x MetroLab Challenge during the poster session hosted by INRIX.
The project brought together transportation engineers and urban planners from Utah State University, UNC Chapel Hill, UCLA, Cal Poly, and the University of Maryland, united by a shared focus on mobility, equity, and data-driven methods. Our work addressed a critical but understudied challenge in emergency transportation planning: vehicle abandonment during wildfire evacuations.
The Problem: When Evacuations Grind to a Halt
During large wildfire evacuations, congestion can escalate rapidly. In some cases, evacuees are forced to abandon their vehicles and continue on foot, an outcome that not only endangers individuals but also severely disrupts evacuation traffic and emergency response operations. This issue gained national attention during the Palisades wildfires in January 2025, when media coverage documented evacuees leaving cars behind on major corridors.
Why We Chose This Project
Our team selected this project in direct response to those reports of vehicle abandonment. While vehicle abandonment is often mentioned anecdotally after disasters, it is rarely measured quantitatively. Understanding its spatial and temporal patterns is essential for improving evacuation design, traffic management, and emergency response strategies.
The 2025 INRIX x MetroLab Challenge provided a unique opportunity to test whether high-resolution trajectory data could be used to move beyond anecdotes and toward actionable insights.
Data and Methods: Turning Trajectories into Signals
The backbone of our analysis was INRIX trajectory data, which allowed us to observe individual vehicle movements over time. We identified potential vehicle abandonment by detecting trips that ended and did not resume during the evacuation period. A major challenge was distinguishing truly abandoned vehicles from those that temporarily stopped or resumed travel later.
Key Findings: Patterns That Match Reality
Our analysis revealed clear and compelling insights:
- Spatial and temporal clustering of abandoned vehicles emerged along Sunset Boulevard and Palisades Drive, closely aligning with locations highlighted in the news reports.
- As evacuations progressed, vehicle abandonment shifted northeast along Sunset Boulevard, indicating that abandonment risk changes dynamically over time rather than remaining fixed.
- Even a relatively small number of abandoned vehicles can significantly disrupt evacuation traffic flows.
The strong alignment between our results and real-world reporting gave us confidence that trajectory data can reliably capture this phenomenon.
From Research to Real-World Impact
These methods from the project can be used to:
- Identify roadway features and evacuation conditions associated with higher abandonment risk
- Anticipate future problem locations before evacuations occur
- Support more proactive evacuation design and traffic management
Key stakeholders that would help improve evacuations with updated evacuation design and planning include emergency managers, transportation agencies, wildfire planners, and local governments.
Feedback from TRB and What Comes Next
The two most common points of feedback after our presentation was interest in:
- How this approach could be scaled to other roadways and wildfire events
- How confidently abandoned vehicles could be distinguished from short-term stops
These questions are in line with our project’s next research steps, which include replicating the analysis across the Eaton Fire and developing predictive models to identify where vehicle abandonment is most likely to occur. More broadly, the discussion highlighted a growing need for mobility research for data-driven tools that can support real-time emergency response and evacuation planning rather than only post-event analysis.
Looking Ahead
Vehicle abandonment may involve only a small share of evacuees, but its impacts ripple across entire evacuation systems. By combining transportation engineering, urban planning, and high-resolution mobility data, this project demonstrates how research can help communities prepare and respond more effectively to wildfire emergencies.
Learn more and about the 2026 INRIX x MetroLab Challenge.