A multi-method approach to navigational risk assessment for maritime autonomous surface ships

Maritime Autonomous Surface Ships (MASS) are an important innovation in the maritime transportation sector. However, the operation of these ships will involve new navigation risks and challenges. It is important to adequately assess and manage these risks to remedy this gap. This study aims to assess the risks of MASS operations using a multi-method approach that combines Failure Modes and Effects Criticality Analysis (FMECA), Dempster-Shafer evidence theory, and a rule based Bayesian network (RBN). Uncertainty is handled via reliability-weighted fusion of independent expert judgements, and the approach supports risk-management strategies by modelling complex navigational hazards more faithfully. Failure modes are prioritised by occurrence, severity, and detectability. The risk assessment is supported by expert opinions and a literature review. This study integrates FMECA–Dempster-Shafer–RBN, reports sensitivity factors (SF), and tests structural robustness by reclassifying over-reliance on automation (ORA) under Human-Related failures. The top hazards remain stable (Sensor malfunctions, ORA, Cybersecurity), and an uncertainty report clarifies limits of expert aggregation and rule design, supporting actionable prioritisation for industry and policy stakeholders.