A Fuzzy Risk Assessment Approach Based on <i>Z</i> ‐Numbers for Enhancing Safety and Human‐Robot Collaboration in Automotive Sector

Industry 5.0 (I5.0) technologies introduce new workplace safety challenges, particularly in human–robot collaboration environments. While robots handle nonergonomic, repetitive, and hazardous tasks, existing risk assessment methods often fail to address the uncertainties in dynamic human–robot interactions. To bridge this gap, this study proposes a novel risk evaluation framework integrating fuzzy set theory with Z ‐numbers. The methodology integrates Delphi method for expert consensus on risk factors, decision‐making trial and evaluation laboratory for causal relationships, analytic network process for importance considering interdependencies, and VIseKriterijumska Optimizacija I Kompromisno Resenje for ranking risks to prioritize mitigation actions. The methodology uniquely addresses hesitancy of experts’ judgments and data imprecision through a systematic approach validated through a case study in one of Turkey's leading commercial vehicle manufacturers on a bus production line utilizing gantry‐type industrial robots. A sensitivity analysis using VIKOR parameters further validates robustness. The Z ‐number framework overcomes traditional risk scores by differentiating scenarios yielding similar scores but distinct profiles, distinguishing low‐probability/high‐severity hazards from high‐probability/low‐severity ones, leading to nuanced prioritization.