Visual comfort index (VCI): An interpretable fuzzy-logic model for human-centric indoor artificial lighting

Visual comfort is influenced by variables that are inherently uncertain and context-dependent. Glare varies with view, uniformity depends on spatial layout, correlated color temperature shifts with adaptation, and measures such as flicker and contrast rendering factor change with use conditions. Expecting precise outcomes from imprecise inputs is therefore unrealistic, which explains why single-threshold methods, such as illuminance targets or UGR limits, often fail to capture actual perception. This study introduces a Visual Comfort Index (VCI), a Mamdani fuzzy logic model that integrates eight determinants: illuminance, uniformity (average-to-minimum ratio), glare (UGR), correlated color temperature, task criticality, visual age, flicker (PstLM), and contrast rendering factor. The model produces both a continuous score and categorical outputs of Comfortable, Acceptable, or Disturbing. The framework is demonstrated on 20 representative scenarios, tested through local sensitivity analysis and validated with 3000 randomized cases, with unsupervised clustering used to reveal typical operating regimes. A standards-referenced simulation benchmark is also conducted against UGR and ISO/CIE 8995–1. A MATLAB-based VCI application is developed to support design reviews by enabling real-time exploration of measured or simulated inputs. The proposed approach provides a practical and interpretable bridge between physical lighting metrics and perceived visual comfort under uncertainty.