Fuzzy Logic-Based Aggregated Risk Values in Occupational Safety Risk Assessment: A Systematic Review

Occupational safety risk assessment combines measured data, qualitative observations, and expert judgment to prioritize preventive action. Fuzzy logic can structure linguistic and uncertain information, but poorly justified variables, membership functions, rules, or weights can embed bias, while aggregation can obscure hazard-specific differences. Following a PRISMA-aligned selection process, this systematic review examined 30 fuzzy-based occupational safety studies, focusing not only on technical design but also on output types, intended decision users and purposes, traceability, temporal operation, and evidence of safety outcomes. Fifteen models primarily produced a score, index, level, or class; nine produced a ranking or priority; and six preserved a relational or causal representation. Only two directly supported workers, one operated in real time, and none demonstrated recalibration driven by observed safety outcomes. Although 28 studies reported positive evaluations of their own solutions, only one reported improvement in accident outcomes observed over time, and one provided partial follow-up. Aggregated fuzzy values are therefore most defensible as traceable, context-bound decision-support indicators. They should complement rather than replace hazard-specific assessment and should be linked to worker participation, documented control actions, assigned responsibilities, and verification of implemented measures.