A Mamdani Fuzzy Inference System-Based Model for Time and Cost Estimation of Structural Works in Construction Projects

Abstract

This study addresses the persistent problem of inaccurate time and cost estimation in construction projects caused by uncertainty during the planning and execution phases. Structural works are particularly vulnerable to delays and budget deviations due to factors such as rainfall conditions, labor discipline, planning quality, design changes, and material price fluctuations. Therefore, this study aims to develop a Mamdani Fuzzy Inference System (FIS)-based model for estimating the time and cost of structural works under uncertain project conditions. The research employed a case-based model development design focusing on structural works in a villa construction project. Data were collected from deterministic schedules and budgets, actual project records, expert questionnaires, Focus Group Discussions (FGD), MATLAB simulations, and Microsoft Project integration. The fuzzy model generated Duration Factor and Cost Factor outputs to adjust deterministic baseline values into uncertainty-adjusted fuzzy baselines. The findings showed that the fuzzy model improved estimation accuracy compared to the deterministic baseline. The project duration estimate increased from 346 days to 369.12 days, approaching the actual duration of 377 days, while the cost estimate improved from IDR 3.74 billion to IDR 3.93 billion compared to the actual cost of IDR 4.01 billion. In conclusion, the Mamdani FIS-based model provides a more realistic and adaptive baseline estimation approach for structural construction works under uncertain conditions.