National road infrastructure is a strategic asset requiring effective, data-driven maintenance management systems. This study develops a predictive model for road maintenance management through correlation analysis of International Roughness Index (IRI), Pavement Condition Index (PCI), and traffic parameters on the 155 km Makassar-Parepare corridor. A quantitative descriptive-explanatory approach was applied to 26,818 observation segments spanning 2020-2024 from the BBPJN South Sulawesi Pavement Management System. Multiple regression models were developed with 70:30 validation split, and Earned Value Management (EVM) was implemented on 50 maintenance projects totaling IDR 65.08 billion. Results show a significant negative IRI-PCI correlation (r = -0.2710, p < 0.001) and weak positive IRI-AADT correlation (r = +0.1898, p < 0.001). The regression model yielded R² = 0.1110 for IRI prediction (RMSE = 1.03 m/km) and R² = 0.1053 for PCI (RMSE = 12.55), with pavement age (AGE) as the strongest predictor. Validation shows the Moderate Scenario 2024 prediction (IRI 5.71 m/km) nearly identical to actual data (5.70 m/km). EVM analysis produced mean CPI of 0.9758 and mean SPI of 0.9435. Integration of predictive models, EVM, and IRMS unit costs yields a data-driven maintenance strategy with 15-20% potential budget savings through timely preventive maintenance.
Copyrights © 2026