AI-DRIVEN SUSTAINABLE INFRASTRUCTURE: SMART MATERIALS FOR CLIMATE-RESILIENT URBAN DESIGN

As climate-related hazards such as heatwaves, floods, and seismic events grow in intensity and frequency, the demand for sustainable and adaptive infrastructure in urban environments has become increasingly urgent. Recent technological advancements have introduced new opportunities to enhance infrastructure resilience through the convergence of Artificial Intelligence (AI) and smart materials. This study aims to systematically examine the combined role of AI technologies and smart material innovations in developing climate-resilient urban infrastructure. Employing a meta-analytic methodology grounded in the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) framework, the study critically reviewed 86 peer-reviewed articles published between 2005 and 2024. These articles were sourced from major scientific databases and selected based on relevance, methodological rigor, and availability of quantifiable outcomes related to environmental performance, adaptive capacity, energy efficiency, and operational cost-effectiveness. The research utilized statistical synthesis techniques, including standardized mean difference (Cohen’s d), random-effects modeling, subgroup analysis, and sensitivity testing, to evaluate the performance metrics and outcomes of infrastructure projects that integrated AI techniques—such as machine learning, deep learning, reinforcement learning, and fuzzy logic—with smart materials like self-healing concrete, phase-change materials, piezoelectric composites, and thermochromic façades. Key findings indicate that such integrations consistently improve infrastructure functionality by enabling real-time responsiveness, predictive maintenance, structural durability, and climate adaptability. Projects that employed AI-regulated smart materials reported performance improvements in energy savings ranging from 18% to 35%, alongside significant reductions in maintenance frequency and life cycle carbon emissions. Moreover, cost-benefit analyses revealed favorable economic outcomes, with most systems achieving a return on investment within 5 to 10 years due to reduced operational and environmental costs. In addition to affirming the value of mature technologies such as PCMs and self-healing materials, the study also uncovered notable knowledge gaps. These include limited empirical testing of novel and bio-based smart materials, inadequate integration of AI-material systems into digital urban design workflows, and a geographic imbalance in research outputs, with minimal contributions from climate-vulnerable regions in the Global South. The findings underscore the transformative potential of AI-driven smart material systems in addressing sustainability challenges, while also calling attention to the need for interdisciplinary frameworks that bridge urban planning, materials science, data analytics, and engineering.