Reducing unsafe behaviors in aviation maintenance: a study on formation mechanism and intervention strategies based on system simulation

Aviation safety problems lead to casualties and property damage, with unsafe behaviors of aviation maintenance personnel being a critical factor. This study firstly constructed a four-stage cognitive model (information acquisition, information processing, response selection, and action execution) to build a cognitive model about unsafe behaviors. In the first two stages, an information processing model was established to analyze personnel cognitions, while in the latter two stages, the Theory of Planned Behavior (TPB) was used to explain operational decision-making. Subsequently, an Agent-Based Modeling (ABM) framework was developed to simulate multiagent interactions in aviation maintenance environments. By synthesizing safety responsibilities across managerial hierarchies, interaction rules between operators and managers were formalized, which was rigorously described by ODD (Overview, Design concepts, Details) protocol to ensure clarity and generalizability. Finally, the ABM was visualized on NetLogo platform and validated through a case study of a maintenance operation. Then simulation analysis of different intervention strategies was conducted to quantify the efficacy in reducing non-compliant operations, providing actionable recommendations. This study innovatively integrated perspectives from social psychology and cognitive psychology to investigate the cognitive model of unsafe behaviors among aviation maintenance personnel. The findings provided a foundational reference for developing safety management strategies in aviation maintenance.