Generative AI-Driven Architectural Design: Automating the Transition from Concept to Building Energy Simulation
Time
July 22, 2026, at 3:00 PM (Beijing Time)July 22, 2026, at 5:00 PM (Sydney Time)
Contact Us
Email: jbdejournal@sciexplor.comSpeaker
Prof. Qingyan Chen
Department of Building Environment and Energy Engineering, The Hong Kong Polytechnic University, Hong Kong, China.
Qingyan Chen is the Director of the PolyU Academy for Interdisciplinary Research (PAIR) and Chair Professor of Building Thermal Science and Global STEM Professor at The Hong Kong Polytechnic University (PolyU). He is also Professor Emeritus of Mechanical Engineering at Purdue University, USA, and served as Editor-in-Chief of Building and Environment from 2008 to 2024. He received numerous awards from different international organisations, including recent election as a Fellow of the Royal Society (U.K. National Academy of Sciences).
Introduction
Early-stage architectural design heavily influences building performance, yet translating creative conceptual workflows into structured building energy modeling (BEM) remains a bottleneck. To bridge this gap, this study presents generative artificial intelligence (AI) workflows that automatically translate design intent into EnergyPlus-compatible inputs. In one workflow, ChatGPT-image and Hunyuan3D-2.5 convert 2D conceptual designs into 3D meshes for BEM, while in another, Claude 3.7 Sonnet generates Python-coded geometries as an automated architect. Both approaches evaluate these designs directly in EnergyPlus. Tested on a Hong Kong office building, shape and HVAC optimization reduced energy consumption significantly while eliminating traditional modeling time. This research demonstrates AI’s potential to seamlessly integrate creative design exploration with automated energy performance optimization.
