A novel approach to risk analysis of automooring operations on autonomous vessels

To compensate for the lack of historical incident data available for the risk assessment of unmanned ships, this study adopted failure mode and effects analysis (FMEA) to analyse potential failures, andbased on experts’ experience, ranked them in terms of the likelihood of failure occurrence, severity of consequences and difficulty of detection. First, fuzzy linguistic variables were introduced to indicate experts’ preferences when addressing the inherent uncertainty of unmanned cargo ships' failure modes. Second, in the expert survey, the indicator of hesitance was introduced to reflect experts' degree of familiarity with the evaluated failure modes. Third, to overcome the inability of FMEA to sort failure modes of the same risk value, this study adopted a technique for order of preference by similarity to ideal solution (TOPSIS) to further rank the identified failure modes. Finally, 19 failure modes for the automooring system were selected to demonstrate how the assessment was conducted. The empirical analysis results indicated that the developed model can effectively screen out failure modes with high-level risks. This study also provides risk control options (RCOs) and offers a new and practical approach to the risk assessment of unmanned cargo ships.