<p>This study presents a vibration-based framework for identifying unbalanced masses in front-loading washing machines by integrating multibody dynamics (MBD) simulation and artificial neural networks (ANN). Training data were generated from an MBD model simulating vibration responses under various unbalanced mass conditions. Vibration signals were collected using three triaxial accelerometers, segmented per cycle, and converted into 72-dimensional feature vectors. A fully connected ANN with three hidden layers, each comprising 128 neurons, was trained on 1000 mass configurations generated via Latin hypercube sampling. The model exhibited strong predictive performance, achieving normalized mean squared errors below 0.001 and correlation coefficients exceeding 0.99 in the test set. Experimental validation using 15 representative mass cases resulted in low average relative errors of 4.56% for front masses and 3.60% for rear masses. The predicted mass distributions were re-applied to the MBD model to reconstruct vibration responses, which closely aligned with measured signals. The average root mean square errors of the time-domain responses ranged from 11.17% to 15.09%, and scatter plots showed high agreement between predicted and experimental displacements. These results demonstrate that the proposed ANN-based approach enables accurate estimation of unbalanced mass distributions while preserving the system’s dynamic behavior. By effectively integrating simulation and experimental data, the framework offers a robust and real-time diagnostic tool for addressing unbalanced mass conditions in front-loading washing machines. .</p>