Design of a system-related risk management chain for an intelligent decision support system prototype in Korean nuclear power plants under normal and abnormal conditions

다국어 초록 (Multilingual Abstract)

An Intelligent Decision Support System (IDSS) is being developed for the Korean APR1400 nuclear power plant (NPP) to enhance decision-making and reduce human error under normal and abnormal conditions. As a non-safety monitoring prototype, the IDSS is currently in development. While traditional Man-Machine Interface Systems in NPPs face risks from software and hardware failures, the integration of deep learning introduces further challenges, particularly those related to open-source software and adversarial AI threats. This study presented a comprehensive system-related risk management chain to address these issues. Risk factors in the IDSS’s development and operation were identified using the NUREG/CR-6430 risk assessment methodology. Based on this, a management plan was established, including administrative and technical controls. Administrative measures addressed licensing and cybersecurity for open-source components, while technical controls aimed to protect against cyber intrusions and maintain IDSS performance.
The proposed risk management chain is expected to reduce operational losses and improve efficiency in NPP applications. It also supports the safe integration of intelligent systems into nuclear operations. Future efforts will apply the management chain across the IDSS development lifecycle to validate its effectiveness and expand its application to broader intelligent system deployment in NPPs.