Artificial Intelligence (AI) is already being used by many industries. While AI is in its early stages, the design community is starting to embrace some of the uses of AI. AI has the potential to change architecture. Its ability to perform repetitive tasks and process information quickly makes AI a powerful tool for architects and design teams. AI can support both the creative and technical sides of architecture. For years computer-aided design (CAD) has undergone a dramatic evolution alongside the design community. Adding AI to the tools available to architects will change how projects are undertaken, how buildings are constructed, and how architecture firms conduct business.
AI tools are already being used in the design process to aid designers in decision-making and problem-solving. AI already excels at generating images and design options. Architects are able to work by selecting design options generated by AI. This allows architects to optimize designs for building-performance metrics, generate floor plans from simple inputs, reorganize floor plans, and use generative AI to rank design variations and optimize for energy efficiency and building performance. Where BIM is used to create a digital twin of a building, the digital twin can be tested by AI, enabling designers to explore changing variables and reassess designs. AI can even predict equipment failure and allow for predictive maintenance.
Architects subjectively analyze design suitability, quality, and variations. This is time-consuming and could be streamlined with AI, while also empowering design using an objective and quantitative approach. Repetitive design tasks can be performed by AI. AI can also enable smaller teams to work on bigger projects.
AI is ideal for finding areas for improvement and efficiency. The technology can provide projections for energy use. Smart homes can be improved with suggestions based on data generated from smart home systems. It also can be directed to find cost savings and identify areas to reduce design and construction costs.
One area of AI that has captured the attention of many is text-to-image software, wherein a text input is used to create incredible images. Also referred to as image-generative software, it has the potential to change concept design. Architects can quickly create visual representations of concepts. This software can also enhance creativity and experimentation. Text-to-image software requires user knowledge to produce useful results, but in the right hands, it can allow a client to see a designer’s vision. Examples of text-to-image software include Midjourney, DALL-E, and Stable Diffusion.
Tools are available for specific design processes such as urban development, bidding, and schematic development. Midjourney is an AI-powered image generation tool similar to DALL-E and Stable Diffusion. Tools like Finch and Maket can generate floor plans that can then be adjusted. Variables such as local codes, and number of units, are incorporated. cove.tool automates improving building design for reduced energy and carbon consumption, daylighting, costs, and more by preconstructing a digital twin and applying machine learning algorithms to refine building performance. ARCHITEChTURES is available for leveraging AI for residential planning. Kaedim generates 3D art based on 2D images. Autodesk Forma is an AI-planning tool aimed at designing sustainable buildings and cities. More tools are available than can be covered here, and many more tools are under development.
AI is poised to have long-term impacts on design and construction. Some AI-driven changes are likely to happen quite soon. The pre-design phase of construction projects requires a comparison of design ideas to building codes. Databases of designs and codes could be used to generate optimized design variations rapidly. During planning for BIM-enabled projects, AI software can take guesswork and trial and error out of the planning process. With growing amounts of data from smart technology, urban and infrastructure planning are poised to benefit from AI advancements. AI could use local codes, traffic patterns, and additional data available in smart cities to assist with Urban Planning. Infrastructure planning will benefit from the ability to identify optimal routes for infrastructure development, reducing the planning time required. Construction activities can benefit from AI-enabled monitoring of movement to detect both safety and productivity issues. AI has the potential to create intelligent building envelopes that respond to changing weather.
While the potential is vast, currently, AI lacks the ability to create buildable schematics from images and to create images from buildable schematics. This is likely to change in the near future. Work is underway for software to advance and go from text-to-image to text-to-3D. Other limitations include that AI cannot get understand how people move through a space and interact with objects. AI is limited by the data from which it has to learn.
The future of AI in architecture will require changes to education, establishing policies for the use of AI, and developing skillsets to allow architects to leverage AI. AI will need to integrate with existing systems, while also ensuring privacy and security.
AI is only as good as the data they are trained on. Biased data leads to biased algorithms.
Are you a building product manufacturer interested in finding ways to leverage AI for your business? Book a call with our experts to learn more about how CADdetails can help you get specified, and some of the latest tools we offer ensure our partnered manufacturers are on the cutting edge of design technology.
Cover Image by SevenStorm JUHASZIMRUS on Pexels