A risk comparison framework for autonomous ships navigation

Maritime autonomous surface ships (MASS) may operate in three predefined operational modes (OM): manual, remote, or autonomous control. Determining the appropriate OM for MASS is important for operators and competent authorities that monitor and regulate maritime traffic in given areas. However, a science-based approach to this respect is currently unavailable. To assist the selection of the proper OM, this study presents a risk-based framework to compare risks in a given situation. To determine the risk level for a given OM, this framework utilizes expected failure modes (FM) related to people, organization, vessel, environment, and technology. FMs and associated accident scenarios (AS) were identified from conventional ship accidents, operating in manual control, in a coastal area in China, based on an extended 24Model. To expand these FMs to other OMs, experts' knowledge elicitation sessions were carried out. Subsequently, a metric for navigation risk of MASS in given OMs was introduced and estimated for the expected AS, using interval-based risk prioritization numbers to convey inherent uncertainty. Finally, by ranking interval-valued metrics in the three OMs, a risk picture was obtained. The feasibility of the proposed framework for risk comparison was verified using grounding in coastal areas where accident data were collected.

Automated summary

This study proposes a risk-based framework to compare the risks of Maritime Autonomous Surface Ships (MASS) in different operational modes. By studying conventional ship accidents, failure modes related to people, organization, vessel, environment, and technology were identified. These failure modes were then extended and applied to other operational modes through expert knowledge elicitation. Finally, a risk picture was obtained by ranking the risks in different operational modes.