Revolutionizing Design: The Power Of Generative AI

Generative Artificial Intelligence (Gen-AI) has been making waves in various industries, and architectural design is no exception. Gen-AI has the potential to revolutionize how architects approach design challenges and create innovative solutions. With more than 1,000 Gen-AI platforms and models available, and more to come, our team has tested and explored more than 50 in past-and-current live projects to compare the time and resources gained versus traditional methods. The results are in, and Gen-AI is significantly advantageous.

Gen-AI is a machine learning (ML) technology that transforms new ideas or designs based on a set of input parameters and constraints. In architectural design, this means that Gen-AI can be used to explore design options quickly and efficiently, allowing architects to think outside of the box, considering more possibilities than they would be able to with traditional methods – without being limited by time or resources.

One of the key benefits of Generative AI in architectural design is its ability to optimize designs for specific criteria or constraints. For example, an architect could use Gen-AI to explore different options for a building’s energy efficiency or structural stability. By inputting specific parameters such as materials, site conditions, and budget constraints into the algorithm, Gen-AI can generate multiple design options that meet those requirements (e.g. establishing the column numbers in a parking garage structure)..

Fig. 1: Spacio – Buildings that are measurable, comparable, and self-aware.

Gen-AI also has the potential to improve collaboration between architects and other stakeholders involved in a project. By generating multiple options based on specific criteria or constraints, architects can present their clients with a range of possibilities that meet their needs while also considering other factors such as sustainability goals or community impact.

However, there are also some challenges associated with using Generative AI in architectural design. One challenge is ensuring that the generated designs are feasible from an engineering perspective. While Gen-AI can generate many different design options quickly, it may not always consider practical considerations such as structural integrity or construction feasibility.

Another challenge is ensuring that the generated designs meet aesthetic standards and align with an architect’s vision for a project. While Gen-AI can help generate new ideas quickly, it may not always produce designs that align with an architect’s artistic vision.

These challenges mean while we may tap Generative AI for assistance, we cannot rely on it wholly for outcomes. In partnership with architects and engineers, Gen-AI is an excellent tool to spur creativity and unlock the art of the possible.

As technology continues to advance at a rapid pace, it’s likely that we will see even more applications of Generative AI in architectural design in the future. From optimizing energy efficiency to improving collaboration between stakeholders on complex projects – there are endless possibilities when it comes to using this powerful technology in architecture.

Digital art is another field being transformed by Gen-AI. One of the most exciting applications of Gen-AI in digital art is text-to-image generation, which allows computers to generate images based on textual descriptions.

Fig. 2: Digital Blue Foam (DBF Engine)

ChatGPT took the world by storm in the fall of 2022 – is there anyone left who has never heard of it? Most of those aware of ChatGPT, whether they have tested it out or just heard about it in passing, understand it as a “text-to-text” tool. A user asks ChatGPT a question by typing it into a field, and ChatGPT replies with a text answer.

Text-to-image generation is different and involves training a machine learning model on large datasets of images and their corresponding textual descriptions. The model then uses this information to generate new images based on textual input.

One example of text-to-image generation is the BigGAN model developed by researchers at Google. This model can generate high-resolution images from textual descriptions with remarkable accuracy, producing realistic-looking images that are difficult to distinguish from real photographs.

Another example is DALL-E, a project by OpenAI that can generate highly detailed and complex images from textual prompts. For example, if given the prompt “an armchair in the shape of an avocado,” DALL-E can generate an image of exactly that – an armchair shaped like an avocado.

Fig 3: An avocado armchair generated by DALL-E. Image generated by OpenAI

Fig 4: Building variations using Text-to-Image with detailed prompt: “The public building is a modern architectural masterpiece in the heart of the city. It features clean lines and glass facades, designed by Zaha Hadid. The building is surrounded by bustling city centers, vibrant squares, and a lively urban landscape. The streets are filled with cars and people, creating a dynamic and realistic atmosphere.” Images generated by Parsons Corp.

One potential benefit of text-to-image generation within the field of architecture design is its ability to democratize digital art creation by reducing barriers to entry. With text-to-image generation tools becoming more accessible and user-friendly, artists without extensive technical skills can create higher-quality digital art with ease. This breakthrough has the potential to streamline the design process for clients and architects, allowing both to bring concepts to life faster than ever before.

We are embracing this technology and is using the LookX.AI platform that specializes in developing Generative AI applications for visual content creation. One of their most exciting applications is text-to-image generation, which allows users to generate high-quality images from textual descriptions.

Fig 5. Frank Gehry beach house variations from LookX.ai – images generated by Parsons Corp.

The LookX.AI text-to-image application uses deep learning algorithms to analyze textual descriptions and generate corresponding images. The application can be used for a wide range of purposes, including creating product images, illustrations for books or magazines, or even generating photorealistic artwork.

One of the key benefits of the LookX.AI text-to-image application is its ability to generate high-quality images quickly and efficiently. This allows users to create visual content at a much faster pace than traditional methods while also reducing costs.

Fig. 6: Aircraft hangar using a detailed prompt from LookX.ai. Images generated by Parsons Corp.

Another benefit of the LookX.AI text-to-image application is its ability to customize generated images based on specific requirements or preferences. Users can input specific parameters such as color schemes, image size, or style preferences into the application, which then generates images that meet those criteria.

The LookX.AI text-to-image application represents a significant step forward for visual content creation by enabling users to create high-quality imagery quickly and efficiently while also pushing boundaries beyond what was previously possible using traditional methods.

For 3D applications, our team is evaluating a plugin within Autodesk Revit that allow this time text-to-3D Model using prompts and styles to generate high-definition images. This tool has developed the possibility to select a portion of the image and with a new prompt render that selection only. We can then side-by-side offer different shape, materials, and styles in the context of the project.

The next big thing in architecture design will be to generate architectural details based on prompts of building specifications that will allow us to gain a significant amount of time in our digital delivery process.

Generative AI represents a significant step forward for architectural design by enabling faster exploration of more complex possibilities while optimizing designs for specific performance metrics. As technology continues to evolve at an unprecedented pace within the architecture industry, Gen-AI will continue revolutionizing how we approach designing buildings.

About The Author

Kal Houhou, Director, Digital Delivery is a technology strategist with a focus on building information modelling and digital transformation for the Architecture, Engineering and Construction Industry. Kal has over 35 years of expertise in technology implementation for the AEC industry. He has a strong business acumen coupled with a clear vision of emerging technology trends that deliver clear results in a complex and ever-changing World.

Be the first to receive updates about Parsons news, events, and innovations. Subscribe Today!

Back to top
facebook-pixel linkedin-pixel linkedin pixel focused image