- Mamou-Mani uses AI software to generate creative designs
- Automating the creation of floorplans
- Ai Build produces AI-powered 3D printers
- AI decides which homes are worth renovating
- Different ways small architecture firms adapt to AI
- Cutting-edge tech is the future of home building
The way people approach building projects hasn’t changed much over the last few centuries. Investors decide on the location, architects draw the design, and construction professionals bring that vision to life, while workers use machines to lift heavy objects and transport materials. But this concept comes with significant drawbacks. Architects have to spend too much time on tedious tasks instead of focusing on the design, while many innovative construction techniques are yet to hit the mainstream market. And deciding whether it’s even worth investing in real estate remains a challenging process, as the wrong decision might spell financial ruin for investors. Many of these problems, however, can be mitigated with the use of artificial intelligence (AI).
Architects can now use smart algorithms to generate thousands of structure designs and then select and improve the best ones. Advanced tech also automates the creation of floorplans and even adds value to the 3D printing industry, enabling robots to optimise the printing process and correct mistakes on the fly. Real estate investors benefit, too. Instead of trusting imperfect human judgment, they rely on algorithms to find renovation projects that will yield the most profit. And although AI-powered solutions might initially require a substantial amount of money and data, they’ll become more affordable over time. Smaller firms will then have access to advanced technologies that will enable them to take on large competitors.
Mamou-Mani uses AI software to generate creative designs
The architecture and design practice Mamou-Mani Ltd is one of the pioneers of digitally designed and fabricated architecture. Its architects use specialised software to create structural designs, but instead of telling algorithms what the final form should look like, the team merely provides input such as the amount of wind or sun, wall thickness, load figures, and the plot size. After that, the computer “come[s] up with solutions within that space”, says the company’s director and chief architect, Arthur Mamou-Mani. The architects can then either select the best design or generate new ones. The final product is made using laser-cutting, 3D printing, or CNC-milling machines.
The AI-generated designs often look like strange plants or fossilised bones. Their organic and biological appearance isn’t what one might expect will come out of a machine-led process. The cavernous and latticed design of Terminal 3 at the Shenzhen International Airport, for instance, was made by AI. The collaboration of software and architects, however, is a delicate matter. Mamou-Mani explains that his work on the Holocaust Memorial in London couldn’t be fully entrusted to machines. He took charge in determining the design of the monument and used the software to calculate the pattern made by the gap between the bricks. “Where we position those bricks to create that pattern was influenced by the computer, but the story is about human loss, and the gaps between those bricks will be used for visitors to listen to the voices of the survivors,” concludes Mamou-Mani.
Automating the creation of floorplans
Computer researcher Stanislas Chaillou has also demonstrated the potential of AI to improve architectural design processes. He developed an algorithm that generates floorplans according to specific architectural styles, including Baroque, Row house, and Victorian house. The research Chaillou carried out at Harvard University shows that each architectural style has a unique set of parameters that algorithms can spot and apply. This breakthrough can potentially enable architects to automatically generate coherent room layouts and assemble them into a floorplan. Furthermore, the software can convert floorplans from one style to another, allowing designers to explore many different visualisations before making the final decision.
The architecture studio Wallgren Arkitekter and the construction company BOX Bygg, both based in Sweden, have also developed an AI tool that generates building floorplans. Called Finch, the software provides various designs according to the apartment size. Instead of architects manually fitting floorplans to different apartment sizes, the AI does the entire job automatically. The program will be launched in 2020 as a plug-in for the visual programming tool Grasshopper. Users will be able to add the orientation of buildings, planning regulations, and other parameters that the algorithms will use to generate plans. “By minimising tedious and repetitive tasks we free up time for design work,” says Pamela Wallgren, a co-founder of Wallgren Arkitekter.
Ai Build produces AI-powered 3D printers
AI also benefits the construction industry by improving 3D printing machines. Currently, additive manufacturing is best suited for small-scale objects. One reason for this is that tiny errors made by 3D printers compound and eventually turn into dangerous structural problems in the case of large building projects. Slowing down the printing process can reduce the severity of this problem, but also increases the amount of time and money it takes to complete the project. That’s why the London-based robotics startup Ai Build has come up with an AI-driven solution.
The company has developed proprietary algorithms that help 3D printers fix mistakes and prevent them from escalating into critical failures. “In this case, rather than the system becoming an architect, it’s becoming a builder with greater flexibility to handle unexpected problems,” says the company’s CEO and co-founder, Dağhan Çam. Ai Build mostly works on installation-sized projects made with thermoplastic composites. In the future, however, Çam plans to enable manufacturers working with metals and concrete to use the technology as well.
He also envisions a future in which advanced machines will build entire cities. Robotic builders led by AI architects could potentially make the construction process faster and more efficient, but getting to that point will require years of hard work. Authorities around the world are yet to be convinced that 3D printing is safe even for bungalows, let alone skyscrapers. Nonetheless, AI-powered additive manufacturing will add value to a range of industries, including construction, aerospace, and automotive, says Çam.
AI decides which homes are worth renovating
AI also helps investors to find lucrative business opportunities. Main Street Renewal, a US-based real estate brokerage owned by the investment firm Amherst Holdings, uses data analytics to find homes that could be renovated and rented for profit. The business is booming because of a falling homeownership rate across the country, with millions of people unable to purchase homes due to tight lending standards, college-debt loads, lagging wage growth, and lack of savings. This translates into five million households that rent single-family homes, which is the market targeted by Main Street Renewal.
The company’s core business is renovating houses in areas with a strong middle-income class and affordable rents. The first step is for human teams to visit cities and find respectable neighbourhoods. Then, Amherst’s proprietary algorithmic tool called Explorer picks individual homes that meet the specified price range and geographic criteria. The software runs several calculations to provide buying specialists with qualified leads.
The cost of renovation, for instance, is an important criterion. Explorer’s machine learning algorithms determine how much money it would take to repair a house by taking into account the cost of previously renovated buildings of similar age, size, and location. If a home is older, the system is programmed to include the cost of replacing the HVAC system, while for structures with visible wear and tear damage it will incorporate the cost of adding a new roof. “Explorer has become so precise that the actual renovation costs average within 5% of the estimates,” says Amherst’s managing director, Joe Negri. The AI also looks into homes rented in a three-kilometre radius to assess what the purchased homes could rent for. In the end, the software provides the team with an estimated rental yield for a prospective house.
Explorer delivers around 1,400 leads each day. Sales agents then analyse each listing, eventually making all-cash offers to 20 per cent of homeowners selected by the algorithm. One in ten offers get accepted, and Main Street Renewal acquires roughly 30 homes a day. Once the renovation is completed, the buildings are rented to families. The company currently manages some 16,000 single-family homes, with plans to expand its portfolio further.
Different ways small architecture firms adapt to AI
The ability of AI to automate various procedures in the construction industry excites architects. But algorithms are complex tools, and architecture bureaus may need a lot of money and data to fully benefit from the latest technologies. This could leave smaller companies at a disadvantage. According to The American Institute of Architects's 2018 Firm Survey Report, three-quarters of firms have only one to nine employees. Despite this, many businesses are optimistic. They believe that AI tools will become cheaper over time, enabling smaller players to access sophisticated technologies.
Natasha Luthra, the director of the innovation program at the engineering and construction giant Jacobs, says that “small firms can be nimble in their adoption of it, such as by renting software on a monthly basis to see if it works for them before looking to make a big investment”. Software won’t become cheaper overnight, though, and it will initially be used by large firms before trickling down to mid-market users. But smaller bureaus have other strategies at their disposal. David Bell, the founder and president of BELL Architects, an engineering and architecture firm, says that he doesn’t have the funds to invest in AI technologies but has clients who might be willing to do that. “Hence we’d be using it more on a project-by-project basis,” says Bell. He also points out that algorithms could be used in the design of net-zero energy buildings to generate “previously unrecognized forms that are optimized for passive strategies”.
Cutting-edge tech is the future of home building
Advances in AI are transforming the way we approach construction projects. Algorithms could eventually not only design homes, but also power smart 3D-printing robots that erect structures with minimal guidance from humans. Smart tech is already deciding which neighbourhoods are to be renovated, and architects rely on AI to come up with increasingly innovative designs, pushing the boundaries of aesthetics. Although large companies will initially be the major beneficiaries of cutting-edge tech, innovation will eventually benefit the entire market and improve people’s lives.