AI-Driven Multi-City Optimization of Glazing and Shading Systems for Building Energy Use and Operational Carbon Reduction Across Global Climate Zones

Buildings account for a substantial share of global energy consumption and operational carbon emissions, and a significantly high value of this impact is formed long before mechanical systems are specified. The building envelope, particularly glazing and shading design, controls how solar radiation and heat exchange, along with daylight, are managed across different climates. Despite this, façade optimization research has largely been conducted within isolated climatic contexts, producing results that are difficult to transfer beyond their original location. This fragmentation leaves designers without robust and comparative guidance for climate-responsive façade decisions.

Rather than focusing on a single city or environmental condition, this study aimed to develop and evaluate an AI-driven, multi-city optimization methodology for glazing and shading systems capable of operating across diverse global climate zones.

A standardized prototype-based simulation methodology was adopted to ensure consistency across all analyses. The building model was developed in DesignBuilder, with EnergyPlus performing the underlying energy calculations. The prototype was tested in representative cities corresponding to major Köppen climate classifications, including hot–arid, hot–humid, Mediterranean, temperate, and cold climates. An AI-driven optimization workflow was then used to change the ratios of glazing to walls, the fixed external shading configurations, and the building's orientation. Performance was assessed using annual energy use intensity, heating and cooling demand, and operational carbon emissions, enabling direct comparison of façade strategies under contrasting climatic conditions.

The analysis showed that façade performance varied significantly across different climatic conditions. AI-optimized solutions consistently outperformed baseline configurations and achieved measurable reductions in energy use and associated operational carbon emissions. Climates that were mostly cool were the most sensitive to shading depth and glazing proportion. Climates that were mixed or mostly warm had more complex interactions between solar gains and thermal losses. No single glazing or shading configuration emerged as optimal across all climates, highlighting the limitations of uniform façade design practices.

In conclusion, these findings support a shift away from standardized façade solutions toward façade possibilities that respond directly to the climate in which they are built. This is achieved by framing façade design as an adaptive problem rather than a fixed solution; this study contributes a transferable, AI-assisted methodology suitable for early stage decisions. The approach provides a scalable foundation for future research on more complex building typologies and advanced façade systems. It also provides climate-responsive pathways toward operational carbon reduction.