Rolling-element bearing (REB) failure is one of the general damages in rotating machinery. In this manner, the correct prediction of remaining useful life (RUL) of REB is a crucial challenge to move forward the unwavering quality of the machines. One of the main difficulties in implementing data-driven methods for RUL prediction is to choose proper features that represent real damage progression. In this article, by using the outcomes of frequency analysis through the Envelope method, the initiated/existed defects on the ball bearings are identified. Also, new features based on developing faults of ball bearings is recommended to estimate RUL. Early-stage faults in ball bearings usually include inner race, outer race, ball and cage failing. These features represent the sharing of each failure mode in failure. By calculating the severity of any failure mode, the contribution of each mode can be considered as the input to an artificial neural network. Also, the wavelet transform is used to choose an appropriate frequency band for filtering the vibration signal. The laboratory data of the ball bearing accelerated life (PROGNOSTIA) are used to confirm the method. To random changes reduction in recorded vibration data, which is primary in real-life experiments, a pre-processing calculation is connected to the raw data. The results obtained by using new features shows a more accurate estimation of the bearings' RUL and enhanced prediction capability of the proposed method. Also, results indicate that if the contribution of each failure mode is considered as the input of the neural network, then RUL is predicted more precisely.