The Development of Safety Surrogate Measures for Vulnerable Road Users through Computer Vision Trajectory Tracking

Status

In Progress

Project Timeline

January 1, 2026 - December 31, 2026

Principal Investigator

Tatsuya Fukushige

Project Team

Campus(es)

UC Davis

Project Summary

In this study, the research team proposes to evaluate several SSMs, their threshold parameters, and develop new, valid metrics for monitoring unsafe travel behavior. To achieve this, researchers will conduct field observations at multiple locations to record incidents such as unsafe interactions, near crashes, and crashes of active travelers. The team will categorize incidents by mode type (pedestrian, bike, e-bike, and e-scooter), maneuver type (head-on collisions, turning movements, merging, and platooning), and observer subjective risk level (negligible, minimal, moderate, high and severe). The project will also use a computer-vision sensor from Viva City Labs (Viva) in the field-based observation to record trajectories at the field location. The sensor will track all interactions by recording mode type (bike, e-bike, e-scooter, pedestrian) and movement trajectories (position, speed, acceleration). Using these trajectories and subjective assessments, the team will statistically examine their relationships to evaluate existing SSMs with various thresholds and develop new SSMs based on the kinetic potential of interactions.