Generative AI is already finding a purpose in helping to design and craft the built environment, so is it time more of the construction industry took it seriously? Tim Clark finds out
Imagine if you can, pressing a single button, and watching an entire building be designed from scratch, in front of your eyes. A decade, or even a few years ago such notions would have been included as the plot of a Christopher Nolan film.
Today, it is arguably closer to reality than many people realise. Artificial Intelligence (AI) has been able to bring images to life, build film narratives out of prompts, and when it comes to construction, is revolutionising just how we design buildings. Or at least will do as soon as the sector adopts it.
When it comes to AI, all aspects of the real estate world are affected, from property management through to architecture, construction and engineering. Although the construction world has become aware of, and is getting to grips with AI, a new iteration of machine learning known as Generative AI is promising to take innovation to the next step.
Generative AI (GenAI) is a specific type of AI that goes beyond analysis and decision-making and is able to create new content. This can be text, images, 3D models, or designs – based on learned patterns. GenAI doesn’t just interpret data; it generates new outputs that didn’t exist before, making it particularly useful for design and planning.
“We started to use AI for ideation in the office to generate ideas and concepts. It was a real creativity boost, so we started to be more systematic about it,” Patrik Schumacher, chief executive of Zaha Hadid Architects’ told The Times in January this year.
“We’re now using the AI systems we’ve created nearly every time we start a project. They generate many more options than we would otherwise have been able to.”
Zaha Hadid Architects (ZHA) has a five-strong technology team devoted to AI, and has invested in its own AI platform after dabbling on commercial off-the-shelf apps such as Midjourney, which is able to produce visuals from prompts.
Sector snapshot.
As befits a sector that has suffered from productivity issues for decades, it may not come as a surprise to find that at least thus far, construction has been slow to adopt the applications that AI can allow.
“The AEC industry has been a relatively late adopter of AI compared to most other industries, and its applications have been mainly in construction, more rarely in architecture and virtually not at all in engineering,” according to a major study, titled Generative AI Applications in Architecture, Engineering, and Construction. This saw academics from Florida, South Africa, Nigeria and Australia analyse the impact and opportunities that AI presented.
The study, published by MDPI, showed that, as it stands, AI is able to help design and construction teams keep track of progress, manage resources more efficiently, and communicate better – all of which boost productivity and sustainability.
Even after a building is complete, AI plays a role in creating detailed documentation to support long-term performance. But while the benefits are clear, adopting Generative AI in architecture, engineering, and construction isn’t without its challenges. Issues like data privacy, bias, and accountability need to be carefully considered.
Alongside such issues, the industry faces hurdles like the high cost of new technology, the need for specialised training, and the complexity of gathering and using vast amounts of domain knowledge.
In the construction world, AI needs to be looked at in comparison to other innovations, such as big data analytics, IoT, and cloud computing, all of which are reshaping not just how projects are built but also how professionals are trained. It has, however, taken over a decade to establish Building Information Modelling in the UK, with progress still being made on wider adoption.
Charlie Sheridan, chief data and AI officer at Nemetschek Group, a software company specialising in providing digital solutions for architecture, engineering, construction, has urged construction firms to ‘think big’ when it comes to adopting AI.
“Put simply, AI’s burgeoning influence over the AEC/O space and across the building lifecycle has the potential to result in considerable economic, infrastructural, and environmental benefits, and fortuitously comes at a time when the industry needs it most,” Sheridan says.
“Perhaps most notably, the introduction of generative artificial intelligence in recent years has greatly enhanced capabilities and led to the development of new supportive technologies for all phases of the building lifecycle.”
Beyond Chat GPT
Large Language Models (LLMs), and programmes such as ChatGPT (which stands for Generative Pretrained Transformer) have helped shine a light on the capabilities of AI to be able to produce on-demand text, images, visuals and other documents.
With GenAI, the level goes up somewhat. In construction, generative AI tools like Autodesk’s Forma or OpenAI’s DALL·E can generate entire floor plans, 3D models, and façade designs from basic prompts.
AI is also able to create multiple variations of a single design based on prompts over cost, sustainability or materials.
Detailed project reports, feasibility studies and construction methodologies are also all able to be produced at speeds above and beyond what is possible from humans alone.
Nemetschek’s Sheridan predicts that AI will also have a large impact on the rise of digital twins, the use of which is predicted to rise 15-fold by 2032.
Coupled with AI-driven real-time monitoring, digital twins enhance construction safety, reduce material waste, and support sustainability goals, thereby redefining industry standards for efficiency and environmental impact – much like the one we recently built for our headquarters – are becoming more and more advanced every day,” he says.
“Rather than simply existing as a reflection of an environment, we are quickly approaching a future in which AI-powered digital twins become the very foundation on which building projects are designed, executed, and maintained.”
How far could AI take the construction industry? It seems that with GenAI the opportunities are almost limitless.
It isn’t only the realm of design and building management that AI may begin to change. When it comes to building more sustainable homes or offices, AI can be used to adapt natural systems to the construction process.
Florida-based professor Thomas Spiegelhalter, co-director of the structures and environmental technologies lab at Florida International University has, along with students at the university, used GenAI to conceptualise bio-inspired, carbon-neutral urban designs for the city of Miami. The professor has utilised BIM and machine learning.
These designs focused on hurricane-resilient structures and green-blue infrastructure, addressing environmental challenges anticipated by a changing climate through to the end of the century.
The research found that AI-assisted computational design approach with deep neural networks were able to “combine bottom-up agents-based geometry models for growth processes found in natural systems with town-down genetic algorithms to optimise construction and design”.
In essence, the use of AI allowed the research team to use the inherent advantages found in nature to design their structures.
Spiegelhalter believes that AI can play a crucial role in advancing sustainable architecture and construction. Commenting to CBS news Spiegelhalter said: “It’s just a matter of time, and it can happen quickly. To design efficient, self-sufficient, and resilient structures, we must learn from nature because it has been here long before us, and will remain after we are gone.”
Behavioural change
As with many new tech innovations, the main barrier may be implementation within the working culture of construction.
“The application of AI using large language models to automate the design creation process requires a profound change in the way all professions related to the built environment think and operate,” the MDPI study added.
“As AI continues to evolve, AEC professionals must be prepared to adapt and embrace new opportunities, ultimately transforming the way projects are designed, managed, and executed.”
When it comes to architecture, a new generation of architects may be trained on how to use AI to assist with design as a co-partner in the creative process.
“AI is transforming architecture by enhancing project conception, design, and execution, but its effectiveness still depends on the architect’s expertise,” says Autodesk’s Jon Holmes.
While AI-powered generative design tools rapidly explore possibilities and optimise layouts, their true potential is unlocked when architects ask the right questions and pose the right scenarios. Architects guide the AI, ensuring that designs are efficient, sustainable, and aligned with the project vision.
Although the world of GenAI is still new, clients in particular may catch on to the implications it can have in terms of speeding up and lowering the cost of design.
“Don’t pretend that AI wasn’t involved,” ZHA’s Schumacher warned in his Times interview. “If anything, the client shouldn’t be worried, as it should mean that you’ve selected from a larger pool of contenders and therefore theirs is more of a ‘winner proposal’.”
Image credit: Gerd Altman/Pixabay
Read next: Equilibrion SHyNEs light on nuclear-enabled hydrogen
Are you a building professional? Sign up for a FREE MEMBERSHIP to upload news stories, post job vacancies, and connect with colleagues on our secure social feed.
