A Data-Driven Decision Support System for Urban Heat Resilience: Comfort Optimization during Extreme Events
Metadata
Show full item recordAuthor
Date
2025Subject/s
Unesco Subject/s
1203.04 Inteligencia Artificial
3305 Tecnología de la Construcción
Abstract
This article investigates the critical interplay between energy consumption, climate change, and thermal comfort highlighting the implications on the built environment and disadvantaged population. The building sector faces escalating demand for cooling system due to urban heat island effect and increasingly frequent heatwaves. In this scenario, ensuring optimal thermal comfort becomes essential not only to mitigate energy consumption and emissions but also to prevent heat-related emergencies for vulnerable populations. In response to these open challenges, the study proposes an AIdriven decision support system (DSS) that integrates real-time environmental monitoring, predictive models, and machine learning algorithms to dynamically forecast indoor and outdoor thermal comfort in terms of Universal Thermal Comfort Index (UTCI) and Predicted Mean Vote (PMV). The framework incorporates AI-based estimation of mean radiant temperature (MRT) and integrates Heat Vulnerability Index (HVI) and Social Vulnerability Index (SVI) to tailor early warning thresholds and adaptation strategies for fragile populations. A pilot project in Barcelona (Besós district), characterized by high thermal vulnerability, is showcased to demonstrate the practical application of this scalable, data-driven approach to enhance urban climate resilience while ensuring thermal comfort and health to the population. © 2025 IEEE.
This article investigates the critical interplay between energy consumption, climate change, and thermal comfort highlighting the implications on the built environment and disadvantaged population. The building sector faces escalating demand for cooling system due to urban heat island effect and increasingly frequent heatwaves. In this scenario, ensuring optimal thermal comfort becomes essential not only to mitigate energy consumption and emissions but also to prevent heat-related emergencies for vulnerable populations. In response to these open challenges, the study proposes an AIdriven decision support system (DSS) that integrates real-time environmental monitoring, predictive models, and machine learning algorithms to dynamically forecast indoor and outdoor thermal comfort in terms of Universal Thermal Comfort Index (UTCI) and Predicted Mean Vote (PMV). The framework incorporates AI-based estimation of mean radiant temperature (MRT) and integrates Heat Vulnerability Index (HVI) and Social Vulnerability Index (SVI) to tailor early warning thresholds and adaptation strategies for fragile populations. A pilot project in Barcelona (Besós district), characterized by high thermal vulnerability, is showcased to demonstrate the practical application of this scalable, data-driven approach to enhance urban climate resilience while ensuring thermal comfort and health to the population. © 2025 IEEE.
Collections
