Rolling element bearings are machine components used to allow circular movement and hence deliver forces between components of machines used in diverse areas of in-dustry.The likelihood of failure has the propensity of increasing under prolonged operation and varying working conditions.Hence,the accurate fault severity categorization of bearings is vital in diagnosing faults that arise in rotating machinery.The variability and complexity of the recorded vibration signals pose a great hurdle to distinguishing unique characteristic fault fea-tures.In this paper,the efficacy and the leverage of a pre-trained convolutional neural network(CNN)is harnessed in the imple-mentation of a robust fault classification model.In the absence of sufficient data,this method has a high-performance rate.Ini-tially,a modified VGG16 architecture is used to extract discri-minating features from new samples and serves as input to a classifier.The raw vibration data are strategically segmented and transformed into two representations which are trained sepa-rately and jointly.The proposed approach is carried out on bear-ing vibration data and shows high-performance results.In addi-tion to successfully implementing a robust fault classification model,a prognostic framework is developed by constructing a health indicator(HI)under varying operating conditions for a given fault condition.