AI-POWERED SMART HOME AUTOMATION: ENHANCING SECURITY, ENERGY EFFICIENCY, AND USER EXPERIENCE IN MODERN HOUSING

Abstract

This quantitative study examined how AI-powered smart home automation capability influenced three outcome domains in modern housing—security performance, energy efficiency, and user experience—within one integrated empirical framework. The study’s theoretical grounding was developed through a structured review of 112 peer-reviewed studies covering AI-enabled home security analytics, residential energy management, and smart-home UX and trust measurement. Empirically, a non-experimental, explanatory cross-sectional time-series (panel) design was applied to a balanced household dataset. The final panel retained 420 smart-home households tracked over 12 monthly windows, yielding 5,040 household–time observations across apartments and detached houses, multiple climate zones, and varying interoperability conditions. Descriptive results indicated meaningful heterogeneity in AI capability (M = 0.61, SD = 0.17), high average security performance (M = 82.4/100, SD = 7.9), moderate energy-efficiency gains (M = 14.8%, SD = 6.6), favorable user-experience scores (M = 3.89/5, SD = 0.54), and low-to-moderate manual override behavior (M = 3.6 per month). Zero-order correlations showed coherent bivariate alignment: AI capability correlated positively with security (r = 0.41), energy efficiency (r = 0.49), and UX (r = 0.44), and negatively with overrides (r = −0.38). Fixed-effects panel regressions confirmed statistically significant direct effects of AI capability on security performance (β = 0.287, p < .001), energy-efficiency gain (β = 0.352, p < .001), and UX (β = 0.318, p < .001), controlling for household heterogeneity, seasonality, and weather. Moderation analyses indicated that interoperability amplified AI benefits across domains, while baseline energy intensity and hot–humid climates strengthened the efficiency pathway, and digital literacy strengthened the UX pathway. Mediation tests showed partial transmission through occupancy inference for energy outcomes and personalization accuracy for UX. Overall, the findings provided robust multi-domain evidence that higher AI automation capability was associated with safer, more energy-efficient, and more satisfying residential performance in heterogeneous modern housing contexts.