Spatial object and its information-enabled modeling automation and demonstration based on individual physical needs

This research proposes a detailed design automation system optimized by incorporating users’ physical characteristics into the design process. Based on previous studies analyzing the relationship between users and spatial components, we developed a mapping system that configures objects within spaces and developed a building information modeling (BIM) automation system called ‘spatial object and spatial BIM-enabled high-level of development modeling automation system (SOMA).’ The system provides specialized design knowledge through a large language model and generates design alternative images using image-generative artificial intelligence, extracting information on required spatial objects from the selected images. The extracted text/image information is filtered based on a mapping system to select relevant object design information, which is then transmitted to the automation module for 3D BIM modeling. This process was evaluated through demonstration tests using three elderly-user scenarios validated the system’s ability to provide design information and automatically generate BIM models based on established design knowledge.HIGHLIGHTS A design automation system that reflects users’ physical characteristics.Developed a mapping system based on user–space relationship analysis.Extracted spatial object requirements using LLMs and image-generative AI.Automated BIM modeling using extracted text and image data.Verified system performance through elderly scenarios and expert evaluation.