Empirical Modeling of Pavement Condition Index (PCI) Based on Traffic Volume and Heavy-Vehicle Composition

Road networks represent essential municipal infrastructure assets that require timely preservation to maintain serviceability and traffic safety. However, premature pavement degradation driven by operational traffic loading presents a continual challenge for municipal pavement management. This study presents a locally calibrated empirical framework to evaluate the degradation of the pavement condition index (PCI) under different traffic compositions. Standardized visual distress surveys across 406 sample units (ASTM D6433) were combined with directional peak-hour classified traffic monitoring across primary urban corridors in Burdur, Turkey. An empirical multiple linear regression (MLR) model was developed using hourly traffic volume (V) and heavy vehicle ratio (HVR) across 12 representative non-rehabilitated pavement sections. The calibrated model demonstrated high explanatory power (R2 = 0.925, SE = 3.56, p < 0.001) and robust internal validity, as indicated by Leave-One-Out Cross-Validation (Q2 = 0.884) and a 1000-iteration non-parametric pairs bootstrap analysis. Key findings show that the standardized regression coefficients highlight heavy-vehicle exposure (β* = −0.712) as exerting an approximately 1.66-fold stronger relative impact on surface distress than gross traffic volume (β* = −0.428). Sensitivity simulations further reveal that increasing the heavy-vehicle ratio from 4% to 12% results in a 21.0-point absolute reduction in PCI, substantially compressing the operational margin before reaching critical rehabilitation thresholds (PCI ≤ 40). Ultimately, this cross-sectional framework provides municipal agencies with an indicative screening reference for understanding traffic-related deterioration patterns without requiring high-cost data collection platforms.