Research

ChatGPT-based digital human for analysis and value discovery of the Avant-garde housing projects

• Principle Investigator Hongju Yang (Master thesis project, Inha University)

• Advisor Professor Seung Wan Hong, Inha University

• Related Publication  Yang, H., Hong, S.W. (2023). Developing a Conversational Artificial Intelligence-Equipped Digital Human Model for Analysing Performances of Housing Experimented by the Avant-Garde Architects. In Proceedings of 2023 Autumn Annual Conference of Architectural Institute of Korea.

• Funding details  This thesis was supported by the National Research Foundation of Korea (NRF) grants funded by the Ministry of Education (Grant Number NRF-2021R1A2C1004608), and the Korea Agency for Infrastructure Technology Advancement(KAIA) grant funded by the Ministry of Land, Infrastructure and Transport (Grant Number RS-2021-KA163269). 

The avant-garde housing projects, while significant in their creativity, often pose challenges in analyzing resident-focused housing performance due to their unbuilt and imaginary nature. This project endeavors to develop a novel simulation model utilizing Chat-GPT, a conversational artificial intelligence that generates schedules for digital human behavior, reflecting residents' social and cultural contexts of a specific era. In this simulation, Chat-GPT, integrated with a prompt module, captures user demographics and generates responses based on learning data about resident details. These responses guide the simulator's input for digital human schedules and behavior rules. Finally, the simulator autonomously computes geometric and semantic properties of a space, generating digital human spatial usage and behavioral patterns. It analyzes avant-garde housing performance in relation to user demographics, occupations, age, daily routines, and behavioral rules. This development is expected to enhance the understanding of architectural design and opens avenues for rediscovering the visionary proposals of Avant-Garde architects within human life. 

Digital humans' responses reveal the unexpected functionality of experimental features, aligning with the avant-garde architects' radical innovations for practical use. This data-driven approach prompts experts to consider wider applications for these projects. Experts recognized the perceptual mechanism for assessing independence and privacy as valid, highlighting that independence should account for both active behaviors like visual interference and passive behaviors such as "being watched".