--Reliable condition monitoring and timely fault diagnosis of industrial electric motors play a vital role in ensuring operational efficiency and reducing unplanned downtime in manufacturing systems. In this study, a data-driven intelligent framework is proposed for detecting bearing faults in an industrial mixer motor utilized in a tire production line. The approach employs a Multi-Layer Perceptron (MLP) neural network trained on a hybrid dataset obtained from both vibration sensors and Programmable Logic Controller (PLC) records. Comprehensive preprocessing was performed, including denoising, signal normalization, and extraction of statistical features such as root mean square (RMS), kurtosis, and crest factor to enhance feature representation. The MLP model, implemented in Python using the backpropagation algorithm and ReLU activation function, was trained and validated under four operational conditions: healthy, inner race fault, outer race fault, and ball fault. Model performance was assessed using confusion matrix, accuracy, F1-score, and receiver operating characteristic (ROC) metrics. The experimental results revealed that the MLP network achieved a classification accuracy of 90.33%, demonstrating high robustness and stability in noisy industrial environments. The proposed framework can be effectively adopted as a practical tool for predictive maintenance and intelligent condition monitoring of industrial rotating machinery