Special Issue: Artificial Intelligence and Machine Learning for Smart, High-Performance and Resilient Buildings and Cities

Rapid urbanization, the intensification of climate-related risks, and the growing adoption of artificial intelligence (AI) and machine learning (ML) are reshaping how buildings and cities are designed, operated, and made resilient. This Special Issue brings together eight contributions that collectively advance the frontier of AI and ML applications across the full lifecycle of the built environment, from individual building thermal dynamics to neighborhood-scale urban hazards, reflecting both the breadth and the urgency of this emerging research domain.A foundational challenge in building performance is the accurate prediction of indoor thermal conditions and their cooling implications. Najafi and Beitelmal [1] address this through a data-driven, physics-informed grey-box framework that models free-floating indoor temperature evolution in hot-arid residential buildings, demonstrating a practical pathway for translating field measurements into simplified predictive models applicable to passive thermal assessment. Complementing this building-scale perspective, Senthilkumar and Ghosh [2] provide a systematic three-decade literature synthesis on the influence of built environment factors, including building height, density, and sky view factors, on thermal comfort, offering a structured evidence base for urban planners and architects seeking to configure the built environment to mitigate heat exposure. These two contributions provide insightful physical and contextual grounding for the development of intelligent control and design strategies.The challenge of real-time thermal regulation is addressed by Baltazar et al. [3], who develop and experimentally validate a fuzzy-rule-based control strategy for a subscaled multiroom building facility, demonstrating that fuzzy logic approaches can maintain room temperatures within tight tolerances under variable disturbance conditions. Extending the AI toolkit to the building envelope, AlBattat and Sharaf [4] investigate the integration of ML, specifically XGBoost, into adaptive facade systems for high-rise buildings, showing that AI-enhanced facades can meaningfully outperform both static and preprogramed systems under harsh climatic scenarios. These contributions highlight the growing maturity of AI-driven control and optimization strategies at the building scale, and their potential for deployment in real-world energy management applications.Scaling from individual buildings to urban systems, Montoya-Rincon and Gonzalez-Cruz [5] present an AI load forecasting framework for dense urban environments that uses a convolutional encoder–decoder architecture trained on high-resolution urban climate simulations to predict spatially distributed cooling loads directly from meteorological fields. This work addresses a critical gap in utility-scale load forecasting. Conventional territory-averaged approaches fail to resolve the neighborhood- and substation-scale demand variability that increasingly governs grid reliability during extreme heat events.In a novel contribution, Sailor et al. [6] present the first field measurements of neighborhood-scale air temperature impacts from data center waste heat, establishing data center anthropogenic heat emissions as an emerging and potentially underrecognized urban thermal hazard with direct implications for building energy demand, urban climate modeling, and future infrastructure planning, a finding of growing significance as data center capacity continues to expand rapidly across urban areas.The Special Issue also contributes to the growing methodological toolkit for AI-assisted design workflows and research synthesis. Mehta et al. [7] introduce VizArQA, a benchmark dataset for evaluating vision-language model performance on architectural environmental simulation outputs, establishing baseline metrics for AI-assisted interpretation of radiation, wind, and solar insolation analyses, a capability that could substantially accelerate early-stage sustainable design workflows. Rounding out the collection, Gupta et al. [8] provide a systematic review of 165 peer-reviewed studies on AI applications in construction safety and sustainability, identifying recurring themes of real-time sensing, multimodal data fusion, and user-centered system design, while highlighting persistent challenges around cross-project generalizability and long-term field adoption.Together, these eight contributions reflect a maturing field in which AI and ML applications are moving beyond proof-of-concept demonstrations toward physically grounded, operationally relevant frameworks for smart and resilient buildings and cities that respond to the compound pressures of climate-related risks and urban growth. Persistent challenges remain, including the generalizability of data-driven models across climates and building typologies and their impact on microclimate, the integration of AI outputs into real-world decision workflows, and the need for high-quality labeled datasets that capture rare but consequential extreme events. These challenges are not merely technical; they are also institutional, requiring closer collaboration between researchers, practitioners, policymakers, and the communities most exposed to building and urban energy risk.The Guest Editors thank all contributing authors for their rigorous and innovative work and extend sincere appreciation to the reviewers whose expertise and diligence ensured the quality of this Special Issue. We hope these contributions will stimulate further research at the intersection of artificial intelligence and the sustainable, resilient built environment.