Machine Learning Applications in Pile Load Capacity Prediction: Advanced Analysis of Pile Driving Forces and Depths in Urban Ho Chi Minh City Construction Sites

In urban environments with complex geotechnical conditions, accurately estimating the load-bearing capacity of driven piles is crucial for structural safety and economic efficiency. This study introduces an innovative approach using an artificial neural network (ANN) model to estimate the load-bearing capacity of driven piles in Ho Chi Minh City. Data were collected from static load tests on 16 precast high-strength concrete piles and enhanced with empirical formula-based calculations incorporating soil properties and standard penetration test results. The proposed ANN model demonstrated high accuracy in performance metrics, namely mean absolute error, mean squared error, root mean squared error, and R-squared (R²), which reached the values of 0.3230, 0.2396, 0.4895, and 0.9180, respectively, on the testing set. The model’s robustness was confirmed across various pile diameters and construction sites. These findings highlight the potential of machine learning to enhance predictive accuracy in pile foundation design, suggesting broader applications in geotechnical engineering.