Architects are feeding natural language prompts into AI image generators like Midjourney and DALL-E, but these tools cannot yet plug into the detailed digital models that actually govern how buildings are designed and constructed. This project tackles that gap: it aims to build a practical toolkit that lets generative AI work directly with Building Information Modelling (BIM) software—the industry-standard system architects use to coordinate everything from structural loads to energy performance. Currently, BIM data is locked inside proprietary project files, and AI applications require bespoke coding for each task. The secondee, Wang, will embed with the architecture firm Hawkins\Brown to co-develop a deployable toolkit that any architecture, engineering, or construction firm can use. If successful, the toolkit would let architects use plain-language prompts to automate labour-intensive tasks such as room classification, daylight glare assessments, and building performance evaluations—work that today requires hours of manual modelling. The broader aim is to reduce the stress and confusion architects face from the flood of new AI tools, and to give the industry confidence that generative AI can genuinely boost efficiency and creativity without compromising professional standards.
View original technical description
The rise of Artificial Intelligence (AI) has evoked a transformative journey in many disciplines of science, technology, engineering, manufacturing, medicine, and art, promising unlimited possibilities for innovation, efficiency and creativity. In the field of architecture, the growing interest in AI-empowered tools, such as Midjourney and DALL-E, have sparked new lines of enquiry into disruptive architectural design technologies and methods: they herald the revolution of contemporary practices to come and contribute significantly to (re)shape our built environment. This project aims to delve into the burgeoning realm of generative AI (GAI) in architecture and focus on GAI for transforming architectural practice through innovative Building Information Modelling (BIM) processes. The advent of GAI in architecture marks a shift towards more integrated, adaptive, and sustainable design processes, but more work is still needed to ensure a seamless integration of GAI. A significant gap exists between the technological capabilities of GAI and its practical implementation with inter-connected, domain-specific BIM models in everyday architectural practice. The proprietary nature of individual building projects limits public access to BIM data for direct AI use. Existing AI applications with BIM are often task-driven projects with bespoke software development. Therefore, the broader use of GAI with natural language prompts would require further development with BIM to fully understand the extent to which GAI technologies can support the transformation of the built environment. Wang has investigated AI-enabled technologies to be integrated with building information modelling (BIM) processes in contemporary Architecture practice. The central inquiry revolves around understanding how AI can boost efficiency and creativity in BIM processes within practices, ultimately impacting sustainability and innovation in built environments. Wang's recent work on automating room classification using 2D images was to reduce labour-intensive modelling processes for easy-access building performance evaluation (BPE) in early residential building planning (Zhao et al., 2022). Another example investigated a streamlined parametric modelling workflow to automate daylight glare assessments for the Chinese green building standard (Wang et al., 2022) and afford optimised design solutions, rendering opportunities to address intricate architectural challenges with unprecedented precision and creativity. Research conducted in this secondment aims to bridge the gap between the theoretical potential of GAI and its practical implementation in architecture. By examining the current state of GAI use in architectural practices, the project will identify key challenges that hinder its widespread adoption, such as technical limitations, knowledge gaps, and resistance to change. Furthermore, this project will explore the ethical considerations and implications of AI-generated designs, ensuring that the GAI integration with BIM upholds the creative integrity of the architectural profession. In this project, Hawkins\Brown (HB), one of the leading Architecture firms investigating future affordable, inclusive, sustainable built environments, will host the secondee (Wang) to co-develop a deployable GAI for BIM toolkit for the broader architecture, engineering, and construction (AEC) industry uptake. The outcomes from this secondment will alleviate the stress and anxiety of inundating GAI tools and boost the confidence of stakeholders in leveraging GAI for enhanced efficiency and creativity across various stages in architecture design and construction. With the prowess to analyse vast datasets, GAI is envisaged to drive the architecture revolution forward, from streamlining BIM processes to fostering innovative design ideas that respond to ever-evolving environmental and societal demands.
Plain English summaries and category classifications on this site are generated by AI and may not perfectly reflect the original research.
Is something wrong? Let us know