Influence of Interruptions on the Diagnosis Task Performance

ABSTRACT

Digital nuclear power systems increasingly expose main control room (MCR) operators to multitasking and interruptions that disrupt cognitive resource allocation and threaten safety, yet the underlying mechanisms remain underexplored. This study conducted behavioral experiments on a nuclear power simulation platform, manipulating interruption type (parameter‐query, audio, alarm‐confirmation) and interruption complexity (simple vs. complex), with the three interruption types inherently differing in their similarity to the primary fault‐diagnosis task (high vs. low). Results showed that higher interruption complexity consistently reduced diagnostic correctness and increased diagnostic time, interruption recovery latency, and workload across all interruption types. Interruption type did not affect correctness or workload, but higher similarity to the primary task lengthened diagnostic time and interruption recovery latency. Working memory capacity consistently predicted faster recovery, especially under complex interruptions. These findings, obtained under controlled experimental conditions, suggest that high‐similarity auxiliary tasks should be minimized or time‐limited during critical diagnosis, low‐similarity tasks prioritized, and complex interruption‐handling assigned to operators with stronger working memory. These preliminary implications require further validation with licensed operators in high‐fidelity control room simulations before informing differentiated interruption management and personalized task allocation in actual MCR operations.