News/Research
03 Nov, 2023

Conference Grant Reports: Xinwei Zhuang at ACADIA

We are pleased to support our students sharing their work at the premiere conferences in their field. Xinwei Zhuang presented their "Encoding Urban Ecologies: Automated Building Archetype Generation through Self-Supervised Learning for Energy Modeling" paper at the Association for Computer-Aided Design in Architecture (ACADIA). From XinWei:

It was an honor to attend the 41st annual conference of the Association for Computer-Aided Design in Architecture (ACADIA), which explored the theme of "Habits Of The Anthropocene: Scarcity And Abundance In A Post-Material Economy." This year's conference centered on amplifying marginalized perspectives, including indigenous and feminist ideologies, as well as queer theory. The goal was to integrate these voices into a new vision for utopianism, capitalism, and consumerism.

During the conference, I had the opportunity to present my recent research on "Encoding Urban Ecologies: Automated Building Archetype Generation through Self-Supervised Learning for Energy Modeling". This innovative tool uses self-supervised learning to simplify intricate geometric data into archetypes that are unique to local environments. The ultimate goal is to transform the way we engage with constructed spaces by incorporating local parameters into customized energy simulations at the community level. This approach enhances the accuracy of energy consumption models and can be applied to multiple building types, facilitating the investigation of emerging local ecologies. To access the full paper, please follow the link below.

In addition, I had the opportunity to participate in a noteworthy panel discussion titled "Learned Ecologies". Esteemed professors and practitioners from renowned institutions, such as University of Florida, University of Toronto, ETH Zurich, and Perkins&Will, were present. The discussion centered around the application of artificial intelligence and machine learning in architecture, with a focus on how these technologies can help solve critical issues like sustainable development and preservation of historical heritage.


During the conference, I not only shared my research but also had the pleasure of connecting with other researchers, including Prof. Stouffs from the National University of Singapore. I was particularly impressed by his work on "A Synthetic Digital Method of Building Material Stock Representation Based on Distance Measurement," which represents a significant advancement in the precision of building archetypes with material considerations. This exciting development opens up potential opportunities for future collaboration and underscores the conference's pivotal role as a hub of innovative ideas and interdisciplinary exchange.

To Read the full paper, Click Here!