Unsettled surrogate-safety thresholds and conflict-to-crash ratios

Determine appropriate threshold values for surrogate safety measures and characterize the conflict-to-crash ratio for simulated mixed-traffic safety analysis.

Background

The paper evaluates traffic safety using surrogate safety measures (SSMs), including time-to-collision and post-encroachment time, because simulated crashes are rare and unreliable. It notes that the interpretation of these measures depends on selecting suitable thresholds for classifying conflicts.

The paper identifies two unresolved methodological issues: the appropriate SSM threshold selection remains unsettled, and the relationship between observed surrogate conflicts and actual crashes is not fixed. These uncertainties motivate the paper's use of attack-versus-baseline differences and monotonicity-based retention of SSMs, but the underlying issues remain unresolved.