| Title |
Development of a Database Using Infrastructure Maintenance Records and a Repair Cost Estimation Model |
| Authors |
Yeonwoo Nam ; Sugyeom Lee ; Ku-Hyuk Kwon ; Seokho Chi ; Tae-Hyung Lee ; Byoung-Gil Shin |
| DOI |
https://dx.doi.org/10.6106/KJCEM.2025.26.6.057 |
| Keywords |
Infrastructure Maintenance Database; Repair Cost Estimation Model |
| Abstract |
As infrastructure continues to age, the accurate prediction of maintenance costs and the efficient allocation of limited budgets have become increasingly critical. However, in Korea, infrastructure assets are managed by various agencies, and inspection and maintenance records are fragmented across institutions, making integrated data utilization difficult. To address this challenge, this study established a standardized facility breakdown structure at the element level, based on national infrastructure maintenance regulations. Using this structure, historical maintenance records were systematically collected and organized into a cost database. Then, prediction models for unit repair costs were developed for major elements such as decks, piers, and drainage systems?according to their repair methods, using the Extreme Gradient Boosting (XGBoost) algorithm. These models aim to reflect the unique characteristics of element-level damage and maintenance practices. By providing accurate and reliable cost estimates, the proposed approach supports maintenance managers and policymakers in improving the precision and efficiency of planning processes. Furthermore, the results are expected to contribute to the development of a data-driven decision-making framework for mid- to long-term maintenance strategy and budget formulation. |