Answers are grounded in this paper and cite their sources.
Autonomy in maritime operations has increased, especially through advancements in Maritime Autonomous Surface Ships (MASS). Despite technical progress, there’s a gap in understanding how human involvement affects safety during transitions between different levels of autonomy. Such transitions are crucial for MASS’s effective operation in mixed fleets of marine vessels across both near-port and open ocean environments. To address this, a hazard analysis framework using System-Theoretic Process Analysis (STPA) is proposed to systematically identify and prioritize critical factors influencing these transitions between human and autonomous agents in marine navigation under different levels of autonomy. The method involves: (1) assessing hazards and operation modes at different autonomy levels; (2) modeling the control structure involving human and autonomous agents; (3) identifying unsafe control actions (UCAs), their causes scenarios and critical factors; and (4) prioritizing critical factors by analyzing triggering events and systemic vulnerabilities from maritime incident reports. We apply this framework to Autopilot (an example of a MASS navigation function) between two ports, reviewing incident reports to uncover UCAs like delayed transition activations, alarm failures, and incorrect initiations by humans or machines. Our findings support the development of targeted safety recommendations to improve MASS navigation operations and are adaptable to other maritime functions.
No takes yet. Share an insight, caveat, or question.
Saaiby et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: