Selection of alarm displays for time- and error-critical tasks: A multi-hierarchy decision framework integrating network structure and multi-criteria decision analysis

Effective alarm design and evaluation are crucial for time- and error-critical tasks. This study introduces a hierarchical network-synthesized multi-criteria decision analysis (HNS-MCDA) framework that balances multi-dimensional, conflicting criteria and reconcile the divergent optimal alarm displays identified by various indices. By decomposing the evaluation into three layers (primary metric, attribute-synthesized, and cross-task synthesis), the framework enables hierarchical decision-making for alarm display selection within specific task types and across task types. It also identifies the most robust alarm under dynamic task types and provides full traceability and interpretability of the decision process. The framework was validated on the OpenMATB platform with six alarm displays and six types of tasks. Results showed that the optimal alarm display differed across the three decision hierarchical layers, with the auditory-tactile alarm exhibiting the highest robustness in the cross-task synthesis layer. The findings demonstrated that the HNS-MCDA framework could enable comprehensive evaluation and optimal selection of alarm displays, thereby enhancing the safety, reliability, and efficiency of alarm systems.