Pedestrian Safety and Level of Service at Uncontrolled Crosswalks - Advanced Data to Enhance Evaluations
Loading...
Authors
Guan, Fei
Issue Date
2025
Type
Dissertation
Language
en_US
Keywords
Alternative Title
Abstract
Pedestrian safety and mobility assessment at uncontrolled crossings is often hindered by the limitations of crash-based evaluation and the inconsistent quality of high-resolution all-traffic trajectory data collected from emerging sensing technologies. As agencies increasingly rely on behavioral and surrogate measures to diagnose risk, the need for reliable, continuous pedestrian–vehicle movement data becomes critical. This dissertation develops a unified methodology structured around two major components. The first component introduces a robust trajectory data enhancement framework that addresses noise, temporal discontinuities, and structural errors inherent in raw multimodal trajectories. The proposed post-processing pipeline integrates trend-based smoothing, iterative outlier detection, and gap reconstruction to produce accurate and behaviorally consistent movement traces. The second component establishes an automated pedestrian–vehicle interaction analysis framework capable of extracting yielding behavior, compliance patterns, and pedestrian service performance directly from enhanced trajectories. This framework formalizes interaction zones, temporal alignment rules, and decision logic to ensure consistent and reproducible behavior classification. To demonstrate the applicability and scalability of the methodology, three case studies were conducted using multimodal trajectory datasets collected along a commercial corridor in midtown Reno, Nevada. These studies evaluate safety and operational performance under varying geometric designs and local traffic conditions. Results show that trajectory enhancement substantially improves data quality, enabling fine-grained, data-driven assessment of pedestrian–vehicle interactions. The integrated methodology provides transportation agencies with a rigorous analytical foundation for proactive safety evaluation and evidence-based design decisions at uncontrolled crossing.