In the first of a new series of articles exploring the evolution and application of AI in the world of architecture, LA Associate and Architect Miruna Stroe looks at AI’s impact and capabilities.
Architect Miruna Stroe
As an industry, architecture has a history of quickly and successfully incorporating new technology. Just look at CAD (computer-aided design), without which iconic structures like the Sydney Opera House, the Gherkin and the Bilbao Guggenheim would never have been built.
Specialist architecture software led to Building Information Modelling (BIM) systems that integrated information management with design work, setting the stage for the comprehensive data handling that AI would require.
It’s no surprise, then, that the industry is embracing AI with both foresight and enthusiasm. A recent survey by RIBA revealed current adoption levels at around 41%, with the likes of Foster + Partners and Zaha Hadid Architects exploring the use of AI by training models on their previous projects.
What can AI do?
RIBA’s survey identified four main areas where AI is currently being used:
Design ideations (basically visualisation iterations)
Concept designs (image generation from text prompts)
Text generation (to improve professional communication)
Practice archiving (training local models on a specific practice’s architectural style and design approach)
While AI is already being used to produce visuals and improve communication, it has the potential to transform the traditional processes of architectural design, planning, and construction. This would open up a new era where data-driven and algorithmically generated designs offer unprecedented innovation and efficiency in the built environment.
AI’s capacity to analyse vast amounts of data and generate design iterations at speed means it can improve the efficiency of design, sustainability and environmental integration. It also has many potential uses in urban planning and construction.
How are we using it?
At Lees Associates, we have long embraced technology to make us more efficient and to de-risk our designs. This is why we were early adopters of BIM, and we strongly advocate its use. Our principal design software, ArchiCAD, has already implemented an AI-based visualisation tool, which we have started exploring in the early stages of our projects.
Future AI Applications in Architecture
As models become more sophisticated, there are several ways in which AI might be applied to architectural and design processes:
1. Design Assistance
Exploring multiple design solutions is time-consuming, but AI can quickly analyse design proposals for structural feasibility, environmental impact, and functionality, offering architects a range of solutions that align with specific project criteria. This would allow for a rapid exploration of design alternatives, providing architects with the means to visualise potential outcomes and refine their designs in real time to balance aesthetics with practicality.
So far, models like Midjourney, Dall-e and Stable Diffusion have become part of the architect's toolkit for trying new things quickly. And while they still have their limitations, they will only become more proficient at responding to prompts.
Miruna working on the BIM model of a current Prime Central London project
2. Construction Optimisation
AI will enhance project management through predictive analytics, which can forecast project timelines and identify potential disruptions. It will allow for real-time decision-making, saving time and money. Robotic automation will bring precision and efficiency to construction sites, handling repetitive tasks, reducing human error, and ensuring safety.
3. Sustainable Solutions
AI is likely to change how we research and specify sustainable solutions for particular projects by analysing environmental data to pinpoint an energy-efficient design. It will also allow for the optimisation of materials and construction methods, reducing waste and carbon footprints.
The capacity to simulate the performance of a specific project for the duration of its lifespan, under various conditions, will also aid in designing truly green buildings in the future.
4. Facilities Management
“The acceleration of cloud-based AI machine learning means homes and buildings of the future literally thinking for themselves.”
We are already seeing the use of AI in the post-construction or post-occupancy phase, relating to ongoing building and facilities management. Chris Thorne from home technology specialists Imperium Building Systems reveals just how sophisticated and sensitive this technology is: “Integrated Building Management Systems (BMS) that continuously monitor and collect data relating to internal conditions such as temperature, air quality and lighting, as well as mechanical and electrical infrastructure performance, external weather conditions and even occupation levels, mean that control parameters can be automatically adjusted to optimise energy performance, reduce plant wear and tear, and improve occupant comfort and wellbeing. Identifying the smallest variances in the ‘Big Data’ collected, such as having to run a pump or fan just 1% faster on average to achieve the same system pressure, can indicate a maintenance requirement and automatically alert the maintenance team.”
The possibilities are almost endless, says Chris: “The acceleration of cloud-based AI machine learning means homes and buildings of the future literally thinking for, and regulating, themselves to maximise comfort, while simultaneously protecting their infrastructure and limiting their carbon footprint.”
Look out for the next instalment in this series of articles, which will look at some case study examples of the use of AI in architecture. Part 3 will consider the ethical arguments.