The traditional Wiener-process-based models often overlook historical data and suffer from missing data issues, thereby resulting in inaccurate RUL prediction. This chapter proposes a novel RUL re-prediction method based on the Wiener process, incorporating both current monitoring data and historical degradation information to improve prediction accuracy. In the initial prediction phase, the Wiener process is used to model system degradation, with the drift and diffusion coefficients estimated using the Expectation Maximization algorithm. To account for uncertainty due to missing data, a DBN model is established. In the re-prediction phase, multiple sets of performance degradation data and historical predictions are combined to refine the degradation stages in the Wiener process, with DBNs used for modeling. The RUL is calculated as the time difference between the detection point and the predicted fault point, determined by a failure threshold. The method is demonstrated on a subsea Christmas tree system, showing its effectiveness in improving RUL prediction by integrating historical data and addressing data uncertainties.