Machine tools generate substantial heat, resulting in elevated operating temperatures that cause deformation of machine elements and subsequent machining inaccuracies. This study investigates the thermal behavior of a machine tool ball screw under varying environmental conditions by integrating experimental measurements, numerical simulations, and machine learning (ML) predictions. A real‐time temperature monitoring system was established using sensors affixed to a machine tool to record temperature fluctuations under various experimental conditions. To enhance our numerical simulation analysis, ML techniques were employed to develop a predictive model for ball screw temperature using ambient and external conditions, along with machine tool temperature data. This research evaluated various ML models for predicting machine tool ball screw temperatures. The results indicate a strong correlation between environmental temperature fluctuations and ball screw temperature distribution. Environmental conditions must be considered in machine tools, as they significantly affect ball screw temperature. Multiple regression models, including support vector regression (SVR), multilayer perceptron (MLP), random forest (RF), XGBoost, and polynomial regression, were evaluated. Tree‐based methods achieved the highest prediction accuracy, yielding the following performance metrics: R 2 (0.9945), root mean square error (RMSE; 0.0319°C), MAE (0.0219°C), and mean absolute percentage error (MAPE; 0.0793%). The proposed framework provides a practical solution for thermal‐aware machine tool operation and offers potential support for improving machining precision under varying environmental conditions. Accordingly, this study underscores the critical role of environmental factors in thermal stability. By leveraging the proposed predictive framework to optimize these conditions, manufacturers can significantly mitigate temperature deviations and enhance overall machining precision.
Kusnandar et al. (2026) studied this question.