Review of dynamic façade typologies, physical performance and control methods: Towards smarter and cleaner zero-energy buildings

• Published in: Journal of Building Engineering 2024-12, Vol.98, p.111310, Article 111310
• Main Authors: Wang, Mengmeng, Jia, Zhuoying, Tao, Lulu, Xiang, Changying
• Format: Article
• Language: English
• Subjects: Architecture design; Dynamic façade; Machine learning; PRISMA method; Zero-carbon buildings
• ISSN: 2352-7102; 2352-7102
• DOI: 10.1016/j.jobe.2024.111310

Dynamic façade (DF) technology is key to mitigating global climate change, benefiting significantly from machine learning (ML) advancements in interdisciplinary research. However, a systematic review and summary of the key roles and performance of intelligent control methods (ICM) in dynamic façade design are lacking, and there is a pressing need to delineate future research directions. In this context, we have conducted a comprehensive review of ICM applied in DF systems. We classified DF from 97 articles into four categories from the perspectives of expression, movement type, response to environmental conditions, and architectural performance objectives: adaptive, kinetic, biomimetic, and passive façade. Further, we subdivided these into transparent, semi-transparent and opaque façades, analyzing their energy and visual performance. Generally, depending on the climatic zone and environmental conditions, DF can reduce energy consumption by 10–50 % and achieve about 80 % visual comfort. Additionally, we explored control methods based on rules, models, and machine learning. Notably, control methods based on ML have achieved over 90 % accuracy in real-time adjustments of façades in response to predicted future climate and environmental changes. Despite this, the application of ML in DF is still in its infancy, with the development of advanced ML algorithms, human-centered control strategies, and fault detection technologies not yet widely explored. Therefore, the broader application of ML technologies in DF systems is set to significantly advance the transition towards smarter, cleaner zero-energy buildings. •Dynamic facades are classified into four types of facades.•Improving the energy and visual performance of buildings is currently the main research focus on DFs.•Complex and comprehensive ML algorithms need further application.•ML is currently mainly used for prediction in DFs.•Office buildings are the most researched type of building in DFs.