An Integrated HFACS–FMEA Model Under an Interval-Valued Fermatean Fuzzy Environment for Human-Factor Risk Assessment in Construction Engineering

Construction accidents are predominantly rooted in human factors that cut across organizational, supervisory, and operational levels, yet traditional risk-assessment tools struggle to capture the multi-level causation and cognitive uncertainty involved. This study integrates the Human Factors Analysis and Classification System (HFACS) with an improved Failure Mode and Effects Analysis (FMEA) that adds Expected Cost as a fourth risk factor alongside Severity, Occurrence, and Detection. Interval-Valued Fermatean Fuzzy Sets (IVFFS) is used to capture the hesitancy of expert judgment; the Best–Worst Method (IVFFN-BWM) determines factor weights, and the Measurement of Alternatives and Ranking according to Compromise Solution (IVFFN-MARCOS) ranks fourteen failure modes. Severity (0.575) dominates the weighting structure, followed by Occurrence (0.218), Detection (0.141), and Expected Cost (0.066). Organizational Culture and Safety Climate, Personnel Physical and Mental Condition, and Environmental Factors emerge as the three most critical failure modes, while sensitivity analysis confirms the robustness of this ranking across a wide range of severity-weight scenarios. The proposed framework offers a theoretically grounded and practically actionable tool for prioritizing safety interventions under conditions of incomplete information.