MODEL REGRESI LINEAR UNTUK PREDIKSI BIAYA KONSTRUKSI AKIBAT PERUBAHAN GAMBAR AS BUILT DRAWING

Abstract

Construction projects are highly vulnerable to cost changes caused by discrepancies in work volumes across technical documents such as Detail Engineering Design (DED), Shop Drawing (SD), and As Built Drawing (AD). These volume changes often lead to budget adjustments and reduced efficiency, particularly in structural works such as reinforcement and concrete. This study aims to quantitatively estimate the effect of volume changes on cost changes using simple linear regression analysis. Data were collected from nine building samples with complete DED, SD, and AD documentation. The analysis was focused on structural works measured in kilograms (kg) and cubic meters (m³). Regression models were evaluated based on R², p-value, intercept, and regression coefficients. The results indicate that every additional 1 kg increases costs by Rp19,303.00, while every additional 1 m³ increases costs by Rp1,321,597.14. Both models achieved R² = 1.00, indicating perfect predictive accuracy of cost changes due to volume variations. Manual verification confirmed that the estimated results were identical to the actual data with a very small error rate. These findings demonstrate that simple linear regression can be reliably applied to accurately estimate construction costs and is practically useful for budget planning and project risk management. Furthermore, this study contributes academically by proving the direct relationship between volume changes and cost changes, which can be further developed in future research on various types of construction works.