Transforming Maritime SAR Operations: Towards a Theoretical Framework for Human-AI Collaboration
TL;DR
It is argued that AI should function as a complementary reasoning partner, improving outcomes without replacing human judgment in critical MSAR operations, though at the cost of greater cognitive effort.
Abstract
This paper examines how Generative AI and Large Language Models (LLMs) can support decision-making in Maritime Search and Rescue (MSAR) operations through a human-centered collaborative framework. The authors propose a theoretical Human-AI Deliberative Framework (HADF) based on the ExtendAI approach, into Maritime Rescue Coordination Centers. In this framework, the human operator first develops and explains an operational plan, after which the AI extends the reasoning by providing structured feedback, identifying cognitive gaps, and supporting reflection while keeping the final decision under human control. The model combines LLM interaction, optimization methods, and supporting datasets to enhance planning quality. Reported benefits include increased decision confidence and satisfaction, as well as improved reasoning depth, though at the cost of greater cognitive effort (the results are drawn from the main study Reicherts et al., not from primary data). Overall, the study argues that AI should function as a complementary reasoning partner, improving outcomes without replacing human judgment in critical MSAR operations.