A new methodology for accident analysis with human and system interaction based on FRAM: Case studies in maritime domain

In the maritime domain, a crew network is characterized by complex collaborative interactions for distributed decision making and rapid transactions of correct information to ensure maritime safety within a limited area and with limited resources. The interaction within the crew network determines the system performance, and the loss of interaction leads to maritime accidents with significant human, economic, and environmental losses. The interactions that occur within the collaborative crew network need to be reflected for system safety. Among systematic safety analysis methods, the safety-II-based functional resonance analysis method (FRAM) is very effective at evaluating the interaction of system functions under normal conditions. We propose a new methodology for accident analysis with human-system interaction (HSI) based on FRAM. To improve the resolution of interaction with crew network level, we defined the relationship between the system function analyzed by FRAM and the crew network as the link type between the function node and the crew node. HSI is analyzed descriptively using the defined HSI link type and graphically expressed. Through semi-quantification of the analysis, we tried to quantify the effect of HSI variability. As a result, this methodology enables to analyze the HSI variability that arises from the relationship between system and human, and to suggest critical part at the HSI level to support the discussion of strategy to manage variability. Two case studies were conducted to illuminate the importance of HSI analysis in the maritime domain.