Autonomous ships developments have been driven by recent advances in smart and digital technologies. As autonomous systems will be responsible for the MASSs’ operation, their reliability is of paramount importance. This study aims to develop a Bayesian network (BN) for monitoring the reliability time variation considering subsystem and component levels. The case study of a cargo vessel for short sea shipping operations is employed and its power plant is investigated. The BN is developed based on the power plant’s critical components, whilst defining the interconnections between these components. Pertinent data for the component failure rates are derived from multiple sources, including reliability databases and scientific papers. The derived results demonstrate that the ship main engine is identified as the most critical subsystem. This study serves as a foundation for the development of a dynamic reliability tool for autonomous ships which can incorporate sensor measurements to update component reliability in real-time.