Human-machine cooperation in safety management: An intelligent decision-making framework for cooling water intake reliability

This paper develops an intelligent decision-making framework for urgent response in the water intake path before marine blockage. Using machine learning and decision-making techniques (heuristic approach and systematic approach), human-machine cooperation supports exploring the pattern of the decision-making process concerning multiple knowledge areas in the nuclear power plant (NPP) reliability of the large cooling water intake. The decision-making framework contains: (1) monitoring datasets with IoT-supported detection, (2) a dynamic expert system, and (3) lump-sum treatments for improving the reliability of NPPs' cooling water intake systems. Through dynamic data collection and analysis, the framework improves the decision quality through higher-level cooperation between human-driven and machine-driven data. Interviews were conducted to illustrate the operating value of the proposed framework in practice. The proposed framework responds to operational, tactical, and strategic requirements via human-machine cooperation design. It effectively finds a valid solution before incidents can happen. It also sheds light on the NPP's operation safety epistemologies from an intelligence-based decision-making viewpoint. Theoretically, the framework presents a human-machine cooperated method for projects involving various decision-makers and multiple datasets. It is expected to bring insights into other projects with similar decision-making processes, like NPP water intake issues.