Towards an online risk model for autonomous marine systems (AMS)

Considering that few or no human operators are directly involved in the operation of Autonomous Marine Systems (AMS), an online risk model is necessary to enhance the intelligence of the AMS, its situation awareness, and decision-making. The current study identifies the criteria for an online risk model for AMS, which can be used to assess its validity and effectiveness. Taking an under-ice Autonomous Underwater Vehicle (AUV) operation as an example, the current work in-vestigates how different risk analysis methods, namely the Preliminary Hazard Analysis (PHA), the Systems Theoretic Process Analysis (STPA), and Procedural Hazard and Operability Analysis (HAZOP), contribute to fulfilling the different criteria for online risk modeling of AMS. The analysis results show that STPA can be considered a good basis for developing an online risk model due to its relatively good coverage of the identified evaluation criteria, especially its ability to handle the interaction between system and software failure. In addition, considering some shortcomings of using STPA and the changing role of human operators in the AMS operation, PHA and Procedural HAZOP can be used as complementary tools. It is expected that the analysis results and conclusions can be adapted to other AMS as well.

Automated summary

Considering the lack of direct human involvement in the operation of Autonomous Marine Systems (AMS), an online risk model is necessary to enhance the intelligence, situation awareness, and decision-making capabilities of AMS. This study identifies the criteria for evaluating an online risk model for AMS and examines the role of different risk analysis methods in fulfilling these criteria. The results suggest that Systems Theoretic Process Analysis (STPA) is a good basis for developing an online risk model, while Preliminary Hazard Analysis (PHA) and Procedural Hazard and Operability Analysis (HAZOP) can be used as complementary tools.