<p>This paper presents a reliability-centered maintenance (RCM) methodology that combines Failure Mode and Effects Analysis (FMEA) with Bayesian Network (BN) modeling to improve the reliability and safety of shipboard systems. The proposed approach supports both qualitative failure assessment and probabilistic reasoning under uncertainty, allowing dynamic updates of failure probabilities and improved prioritization of maintenance actions. The methodology is applied to a marine fire and gas detection system, where 19 failure modes and 59 associated failure causes are identified using expert input and failure data from the OREDA database. Bayesian analysis is used to quantify the contribution of individual failure causes to overall system risk. Based on this analysis, targeted redundancy strategies are implemented and evaluated in terms of their impact on system reliability and associated costs. Results show that introducing redundancy for the most critical failure modes significantly reduces failure probability with cost increases. A comparative analysis highlights the trade-offs between safety gains and redundancy investment, particularly relevant for autonomous vessels where high system reliability is essential. The findings demonstrate that the FMEA-BN framework is a valuable tool for maintenance planning and design optimization in maritime applications.</p>