A hybrid approach to enhance decision-making in marine structures: Combining sensor data with human perception

• Probabilistic framework merging sensor data & human input for real-time assessment. • Bayesian updating refines failure risk using human observation after the damage. • Results show enhanced decision-making for damage management in naval applications. In modern ships, safety procedures during emergency responses to incidents often rely on a combination of manual human assessments and communication among personnel. However, this approach can introduce inefficiencies, delays, and miscommunication, ultimately compromising the accuracy and timeliness of decision-making. Structural Health Monitoring (SHM) systems provide real-time damage data, but their effectiveness is limited by the complexity and scale of the ship’s structure. Moreover, human observation remains crucial, particularly when sensor coverage is incomplete or uncertain. In this paper, we propose a novel probabilistic framework that systematically integrates sensor data, numerical simulations, and human input to enhance real-time decision-making during missions. Unlike traditional approaches, this framework provides a structured method to incorporate human observations into the assessment of structural integrity, allowing a mathematical reduction of uncertainty in the ship’s damage state after an incident. By leveraging offline databases of damage morphology and an onboard pattern recognition system, the novel approach improves the accuracy of failure probability (PoF) predictions, demonstrating that human input—when integrated in a structured and quantitative manner—can significantly refine reliability assessments. The results show a substantial reduction in the uncertainty associated with PoF estimation, particularly in high-risk conditions where human assessments play a key role. This integrated approach provides a novel and practical strategy to enhance both operational efficiency and safety in maritime operations.