Chung-Ang University Realizes Self-Operating "Physical AI" Architecture

by Lee Jiye

Published 24 Apr.2026 11:02(KST)

Successful Demonstration of Autonomous Environmental Sensing and Control in Architecture

A research team at Chung-Ang University has succeeded in demonstrating "Physical AI Architecture" technology, in which buildings autonomously perceive their environment and control themselves using artificial intelligence (AI).

Kinetic facade mockup operating in real-time with artificial intelligence autonomous control. Chung-Ang University

On April 24, Professor Chae Youngho's research team at the Virtual Environment Laboratory (VELAB) of the Graduate School of Advanced Imaging Science, Multimedia & Film at Chung-Ang University announced that they have developed a "Large-Scale Kinetic Facade Predictive Swarm Control Framework," a next-generation architectural exterior technology, and successfully implemented it as a full-scale kinetic facade mock-up.

The core of this research is the transplantation of a "Physical AI Brain" into buildings, allowing them to make autonomous decisions and movements. The team integrated a Graph Neural Network (GNN) and Reinforcement Learning (RL) into the "Kinetic Facade," which enables exterior materials to move autonomously in response to weather changes and regulate energy. By building an AI swarm intelligence system in which thousands of exterior modules communicate in real time, the researchers solved the bottleneck problem caused by increased computational loads and opened up the possibility of applying the technology to large buildings.

The team directly applied the trained AI policy to a 1:30 scale hardware prototype and conducted empirical tests. As a result, they confirmed a 10.3% reduction in solar heat gain in the seating area and a 25.4% decrease in unnecessary motor operations. Notably, by allowing the AI to autonomously coordinate conflicting objectives-"minimizing radiant heat for the seating area" and "maximizing radiant heat for the stadium grass"-the research also improved the mechanical stability of the dynamic exterior materials.

The research results have been accepted for publication in the AI Special Issue of the international journal "Journal of Computational Design and Engineering (JCDE)," which focuses on design and engineering. Selected as a "SW Star Lab" by the Ministry of Science and ICT, Chung-Ang University's VELAB plans to expand the application of this technology to high-value industries such as smart city and landmark design.