Optimizing and quantifying risk assessment in socio-technical complex systems using the FRAM method and fuzzy logic

Conventional safety risk analysis methods, based on linear cause–and–effect relationships, are inadequate for capturing the dynamic and nonlinear nature of socio-technical systems. This study proposes a novel hybrid fuzzy logic-FRAM framework to quantitatively assess performance variability and emerging risks in complex socio-technical systems. The framework is applied to an air separation unit (ASU), a critical component of process industries with significant safety and socio-economic implications. Ten key functions of the process were identified through semi-structured expert interviews. Their variability was assessed using a Common Performance Conditions (CPCs) checklist and expert judgment. Fuzzy logic was integrated with FRAM to address uncertainty and transform qualitative evaluations into a numerical Internal Variability Function (IVF). MATLAB was employed for fuzzification, transforming binary judgments (0 or 1) into a continuous scale (0–1) to enable more precise numerical analysis. Results indicated that compression, distribution, and storage (human functions) and purification and fractional distillation (technical functions) exhibited the highest variability and potential risk. Fuzzification results showed these functions had the lowest IVF, highlighting their need for more careful risk management. The proposed framework enables transparent prioritization of risk-prone functions, supports resilience-based decision-making, and helps reduce operational disruptions, accident probability, and associated economic losses in complex industrial environments.