Enhancing operational reliability in particleboard production through dynamic human reliability modeling using integrated SPAR-H and fuzzy cognitive maps

Human reliability is central to process safety and operational performance in industrial manufacturing, where human failures raise risk and vulnerability. In human reliability analysis (HRA), the standardized plant analysis risk-human (SPAR-H) technique captures this by modeling how performance shaping factors (PSFs), organizational, human, and task conditions, affect human-error likelihood. Building on this PSF view, this study proposed an integrated multicriteria decision-support framework combining expert judgment, Pythagorean fuzzy (PF) decision-making, and cognitive modeling for risk-informed decisions. PSFs are identified through literature review and expert evaluation, then weighted with PF-SWARA to obtain robust importance values under uncertainty. Critical tasks are prioritized with an improved PF-EDAS (IPF-EDAS), which ranks tasks by deviation from a system-representative average risk level. This average is computed in a dependency-aware manner using a PF-based fuzzy cognitive map (P-FCM) that represents interdependence and feedback among PSFs and supports scenario analysis. The framework is demonstrated in a particleboard production process in the wood industry. Results identify key risk drivers, distinguish high-risk tasks, and support targeted mitigation planning. Sensitivity and comparative analyses confirm robust rankings and show better discrimination when PSF dependencies are explicitly modeled. The framework offers a structured, dynamic basis for prioritizing interventions and allocating safety resources across tasks.