Muthanna Journal of Engineering and Technology
Volume (14), Issue (2), Year (2026), Pages (91-99)
DOI:10.52113/3/eng/mjet/2026-14-02-/91-99
Research Article By:
| Raniah Harith Khudhair, Rawnaq Arif Mohsin and Meena Muataz Abd |
Corresponding author E-mail: rania.h.khudair@uotechnology.edu.iq
ABSTRACT
This study investigates the performance of an AI-driven kinetic façade to reduce cooling demand and improve daylight conditions in administrative office archetypes in Baghdad, Iraq, where extreme summer temperatures and frequent dust events limit the effectiveness of static envelope systems. The research addresses a regional gap in the application of predictive façade control that simultaneously responds to solar exposure, indoor daylight requirements, and dust-shielding needs. A parametric building model was developed in Rhino and Grasshopper and evaluated using Ladybug, Honeybee, and EnergyPlus-based environmental simulations with Baghdad EPW climate data. A predictive control model was trained on simulation-generated data to predict façade opening angles based on solar geometry, outdoor temperature, and operational conditions, while maintaining indoor illuminance at acceptable workstation levels. The results indicate that the proposed system reduced the cooling energy use intensity from 185 to 121.7 kWh/m²·yr, representing a 34.2% reduction, and lowered July peak cooling demand by 41%. The system also improved useful daylight illuminance from 45% to 78% of occupied hours, reduced discomfort glare by 62%, and maintained an effective solar heat gain coefficient below 0.15 during peak hours. These findings indicate that predictive kinetic façades can provide a viable envelope-level strategy for improving thermal and visual performance in hot-arid administrative buildings, provided that control logic and mechanical operation are explicitly integrated into the evaluation framework.
Keywords:
AI-Driven Kinetic Facade; Dust Shielding; Energy Use Intensity (EUI); Machine Learning (ML) Model; Predictive Simulation.