Introduction and Key Terms - Virtual Reality Applications In Archiecture - AI Art in Architecture
7-8 - Practical Uses of AI in Architecture
9-12 - Applications of AI in Other Design/Construction Fields
Key Terms:
AI – software used to perform tasks or produce output previously thought to require human intelligence.
Machine
Learning – The capacity of computers to learn and adapt without following explicit instructions, by using algorithms and statistical models.
Text-to-image generators – uses neural networks to process text input and generate images.
Conceptual Design – the first stage of the design process, where designers establish the vision and goals for a project. Relates to an idea or artwork that expresses the creators idea of how a completed project may look.
Extended Reality – umbrella term that encapsulates Augmented Reality, Virtual Reality, and Mixed Reality
BIM – Building Information Model, process that used 3D models to plan, design, and construct buildings.
Architecture is a fluid profession. Meaning that the resources and tools we use are constantly changing with every decade. It was not too long ago that architecture was done strictly with paper, pencils, and rulers to create our built environment. It was even sooner that BIM software such as Revit and SketchUp started taking the forefront in architectural engineering. And now after 20 years, AI and mixed reality have come to aid architects in their everyday practice. AI is an innovative tool in architecture that can be used to explore many options during the early design process. The key to keeping the integrity of the architecture field, however, is to confine the use of this tool to the early conceptual stages of the project.
Virtual Reality
The architectural landscape is evolving with the integration of digital and virtual dimensions, creating challenges for interaction between virtual and real spaces. Traditional design methods often fall short in accommodating real-time interactions, workflow integration, and other modifications. To tackle some of these issues, the article Enhancing Interaction in Virtual-Real Architectural Environments by Xinxing Chen introduces a generative design method that highlights the symbiotic relationship between virtual and real environments through Mixed Reality (MR) and Generative Artificial Intelligence (AI) technologies.
Two approaches were developed, one using the Rhino modeling platform and another based on the Unity3D game engine, aimed at different applications. Users experience evaluations in exhibition, leisure, and residential contexts demonstrated improvements in design flexibility, interactive efficiency, and user satisfaction. The Unity3D approach excelled in allowing quick design changes and immersive experiences. MR was found to provide superior visual comfort and immersion compared to Virtual Reality (VR), aiding architects in interface and scale design. Analysis of participant behavior in virtual-physical spaces yielded valuable insights for optimizing spatial layouts and interactions. The findings highlight how this generative AI-driven MR method streamlines traditional design processes by enabling real-time modifications through simple verbal and motion interactions, paving the way for innovative architectural design and exhibition experiences.
AI ART IN ARCHITECTURE
Recent advancements in diffusion-based AI art platforms enable the generation of impressive images from text prompts, making them valuable tools for concept design across various creative fields, including the conceptual stage of architectural design, which involves ideation, sketching, and modeling. The paper AI Art in Architecture by Joern Ploenninigs explores the applicability of AI platforms like Midjourney, DALLE 2, and Stable Diffusion for common architectural tasks. The authors present several contributions: first, they compare the capabilities of these public AI art platforms; second, they outline the requirements for these platforms to effectively support civil engineering and architectural
use cases; third, they analyze 85 million queries from Midjourney using Natural Language Processing (NLP) to identify prevalent usage patterns. This analysis leads to the development of two specific workflows, one for generating images suited for interior design and another for producing views for exterior design. These workflows aim to leverage the strengths of each platform, facilitating a more efficient and effective design process in architecture. Overall, the findings suggest that diffusion-based AI models have significant potential to enhance creativity and streamline tasks in architectural design, bridging the gap between technology and traditional design practices.
While these approaches are to be taken with a grain of salt in the design process, it can still generate form of inspiration quickly for the designer. This makes the work flow of the conceptual and post processing stages of the design go much smoother. I have used this method a couple times when appropriate. However, the class above mine used this in a study for their research studio. I talked to one of the students about the process and he described that AI was used to drive a lot of their initial design phases. Due to the technology being new, there were some hiccups when trying to do a direct implementation from the AI into the actual concept. However, that was to be expected. Fully implementing an AI generated façade or interior is frowned upon. As there is no design process in that case. As mentioned earlier, AI can be used as a tool and not a 1:1 translation of the program into the design.
There are no current examples of fully constructed buildings that were designed with AI. However, this technology is being used to push new ideas of what the future could look like.
PRACTICALITY
Not only is AI being used to design the physical aspects of architecture, but it is being used to help with the overall experience inside buildings.
One case study that stood out to me, which ties into the recent accessibility and inclusion topics discussed the Fall 2024 building studio is Wayfindr, which can be used to help the visually impaired.
This works through a smartphone by using beacons placed in buildings. The AI processes the spatial data that is received by these beacons to alert the user by audibly guiding them through a space confidently
Another example is located in the National Museum of African American History and Culture in Washington D.C. AI was used to provide feedback to the designers ensuring that the placement of ramps, elevators, and seating arrangements were sensitive to users in wheelchairs and other mobility devices.
“Wayfindr is a groundbreaking AI-powered navigation tool to assist visually impaired individuals in navigating complex public spaces.”
Another way AI and mixed reality can be used is in other professions as well. The world is seeing a giant leap towards new ways to create works of art. In the graphic design world, AI is being used to create fast designs for logos and other works. One of the most recent features was presented by Adobe Illustrator in their recent expo showcasing their AI design capabilities.
Not only can it be used to create artwork but change existing pieces inside of a composition. The example they showed were two characters in a scene. One of those being a cartoon knight and the other a dragon. With AI, a person could change the direction the character was facing ultimately creating a 3D character out of a 2D image.
Another profession that is making great strides in this new technology is construction, HVAC, electrical engineering, and plumbing. With a virtual reality headset, a project manager could theoretically see a 3D BIM model placed on top of the site in the real world. This allows the person to see the entire vision of the project before anything is constructed.
I have had one personal experience with using this technology in architecture that proved to be very valuable. One of the important aspects of using this technology is interaction. A client can seamlessly walk through the design and get a feel for all the space. Once there is some changes to be made, the critiques can be implemented in real time, driven by AI technology. My experience with this comes from the Brookings Library project that took place in the Spring of 2024.
My group and I were designing the entrance space into the building. The building has a cantilevered room over the top of the entrance, and it seemed that the height of the room was going to be too close to the patrons heads as they walk in. Sure enough, once we put ourselves into the model using VR, the height of the room was too close to our heads and would have been a problem if we did not use this method.
HVAC
Plumbing
Works Cited
Architecturecourses (n.d.). AI in Architectural Design - Real-World Examples. Architecture Courses.org. Retrieved November 24, 2024, from https://www. architecturecourses.org/ai/ai-architectural-design-real-world-examples
Chen, X., & Gao, W. Enhancing interaction in virtual-real architectural environments: A comparative analysis of generative AI-driven reality approaches (266th ed.). Science Direct.
Ploennigs, J., Berger, M. AI art in architecture. AI Civ. Eng. 2, 8 (2023).
