- The AI for engineering space is getting pretty crowded, with many companies developing ways to automate time-consuming processes.
- For urban development, Norway’s Spacemaker analyzes noise, wind, traffic, and other data points to make cities more livable.
- AI tools are also being created to improve bidding for modular construction, interior layout planning, and real-estate development.
As more buildings generate founts of data on how they’re built and operated, a new question emerges: Just who, or what, is best suited to sort through it all? The answer, especially for the early planning and design phases, is artificial intelligence (AI).
A new crop of innovations empowered by AI and machine learning is changing the architecture, engineering, and construction (AEC) industry. Before designers even begin creating iterations, using automated tools to organize site and contextual data can sweep away ambiguity and, hopefully, risk. These tools can make very technical, programming-heavy tasks more accessible to noncoders such as designers or developers.
From research projects to commercial products, the following examples show how AI in architecture can create opportunities to improve the design process so human creativity can take center stage.
1. AI for Urban Development
New AI tools can apply generative and iterative power to urban-scale sites, looking beyond individual building requirements. This concept is exemplified by Spacemaker, the Norwegian technology company acquired by Autodesk, which offers cloud-based AI and generative-design software that helps planning and design teams make more informed decisions faster and enables improved sustainability opportunities from the start.
Applied in the early stages of real-estate development, Spacemaker can analyze up to 100 criteria across city blocks: zoning, views, daylight, noise, wind, roads, traffic, heat islands, parking, and more. Its wind-modeling features analyze how buildings channel wind, using computational fluid dynamics to refine designs for human comfort. Its noise features can predict sound levels from traffic or other sources, then compare that data with local regulations. The platform can suggest alternative compositions to, for example, mitigate noise pollution, an often-overlooked component of environmental health.
For Økern Sentrum, a 1-million-square-foot, mixed-use development containing 1,500 apartments in Oslo, Norway, developer Steen & Strøm and Storebrand refined noise and daylight levels using Spacemaker. After plugging their plan into Spacemaker with architects at A-lab and the city-planner client, they reduced the noisiest residential facades by 10% and decreased low-light residential areas by more than 50%. Even with these revisions, the team squeezed out more sellable real estate, a rarity for retroactive regulatory adjustments.
“We can adapt the project with many different parameters, such as noise and daylight, and test different hypotheses by manually changing the design, then view the results in just a few minutes,” says Peter Fossum, a developer at Steen & Strøm. He adds that workshops conducted by Spacemaker have been a boon to architectural master-plan development, improving both process and outcome.
Spacemaker also works for planning landscape elements such as streams and terrain, as well as smaller-scale projects. Building geometry is one of its design parameters; at a granular scale, Spacemaker can, for example, automate the design of floor-plan programs such as apartment layouts. Valode and Pistre Architects report that using Spacemaker increased its productivity by 35% in the design phase, resulting in lower project costs and a wider range of design variations.
2. AI for Better Bidding
ConXtech, a Bay Area–based modular-construction company, is using AI to gain control of one of the most unpredictable steps in construction: the bidding process.
ConXtech, like many construction companies, is solicited by owners and developers during the project-development phase. At that time, the viability of the project is not yet secured, and multiple options are still on the table. This forces companies like ConXtech to go through multiple iterations for projects that may never be built. In the end, millions of dollars can be spent on unsuccessful projects or unsuccessful bids. At the same time, owners and developers expect quick answers, as they seek a path to a viable and cost-effective solution for their business.
To shorten the bidding cycle and reduce the bidding costs, ConXtech worked with Autodesk Research to develop a prototype bidding platform that uses AI to find the most cost-efficient structural-steel design based on the costs of material procurement, fabrication, and construction. These costs are influenced by the vendors and subcontractors selected for the project and vary depending on the project’s location.
After the project-management team identifies a list of potential vendors and subcontractors, the prototype notifies the project’s structural engineer to design the most cost-competitive structure, with three AI agents. The first AI agent, HyperGrid, places columns and sets the structural grid for a given site using a combination of structural-engineering knowledge and reinforcement learning. It takes into account the requirements and constraints imposed by the owners and architects. The second AI agent, the Approximator, predicts the size of beams and columns and the location of ConXtech connectors (the system’s fixed connections) using graph neural networks trained on more than 4,000 building-simulation data points. The third AI agent is the Optimizer. It refines structures to decrease construction costs, factoring in local building codes.
“This proposed AI-assisted technology could help owners and developers at the front end of a project obtain structural designs and estimations of materials required for their buildings without hiring professional engineers,” says Adam Browne, ConXtech’s chief engineering officer. “The product envisioned could be to the structural engineering profession what LegalZoom is to the legal profession: an online analytical technology that helps its customers create material estimates, plans, and calculation documents without having to hire professionals.” This AI technology will not replace the structural-engineering mission and the role of the engineer of record, which is still mandatory during the execution of a project.
3. AI for Volumetric Design and Planning
Japanese construction, engineering, and real-estate development company Obayashi also worked with Autodesk Research to envision an AI solution—one that lets architects plug in basic parameters for buildings and, with minimal guidance, get volumetric estimates and interior programming layouts. Used mostly for office spaces, the AI for this application was trained with a subset of Obayashi’s portfolio of more than 2,800 Autodesk Revit files.
The AI tool understands abstract relationships between programs and the desired connectivity, size, and proportion expressed in a building’s volume. To generate interior programming layouts, the designer and client work through a series of lexical parameters: simple sentences that specify building elements and their location and show how they relate to each other. This might be, “Meeting rooms should be placed close to windows,” or, “Lunchroom should be placed away from the lab for security.”
Architects can demonstrate to the AI agent the meaning of vague concepts such as “close to” or “away from.” Once those are learned, the AI agent can quickly place design objects in their perfect position in the current project and reuse these high-level design principles in future projects with different geometric layouts.
This process is the opposite of an architect’s freehand napkin sketch to win a client on the fly. In the time it takes to make a quick drawing, designers or builders can give prospective clients a concise draft of what their building might look like. With Obayashi’s research prototype, these prospective designs exist in real times and places, defined by what’s actually buildable.
“The AI-assisted design prototype developed through our long collaborative journey with Autodesk Research reflects how architects think about the what, why, and how of the design process,” says Yoshito Tsuji, general manager of Obayashi’s architectural design & engineering division. “The collaboration between AI and architects enables us to communicate the design faster and get clients’ buy-in in a timely manner.”
4. AI for Real-Estate Developers
Parametric design is usually reserved for formal extravagance and dramatic architectural swoops, curves, and cantilevers. Instead, Parafin uses parametric-iteration AI to balance program, cost, and commercial viability. Developed by architect Brian Ahmes and developer Adam Hengels, a Chicago-and-Miami-based duo who are residents in the Autodesk Technology Centers’ Outsight Network, the program generates near-infinite derivations for objective profitability and performance.
Parafin is a cloud-based generative-design platform that’s currently used for hotel developments. Aimed primarily at real-estate developers, it helps quickly evaluate the financial viability of potential building sites in early-stage planning. It asks for just a few parameters (number of rooms, parking, site, height, and brand guidelines for hoteliers) and then can generate millions of iterations fulfilling these guidelines—all searchable by financial performance, cost, and more. It works through a map- and menu-based interface in a web browser; highly detailed floor plans, 3D views, and Revit files are generated for each design.
“A developer can quickly understand, ‘What can I build on the site,’ and ‘Does it make money?’ in a matter of minutes instead of weeks or months,” Hengels says.
It’s typical for developers to sort through dozens of sites and development opportunities for a single project: This phase can prove overwhelming before even acquiring a property. The platform saves critical labor time by forestalling the need to pull team members from existing projects to assess new sites and determine feasibility. It also saves time, with initial programmatic estimates available in a few minutes. This lets designers do what they do best and enjoy most: spending more time on the richer, formal qualities of buildings.
Parafin sets projects on the right track—a digital one—as early as possible. “Today, design projects are often started outside of Revit and put into Revit later,” Ahmes says. “But when you run Parafin, the design is born in Revit at first conception.”
All of these AI applications share this benefit: beginning construction projects as digital natives in order to exert greater control over time, resources, viability, and performance throughout. From this stronger starting point, designers can take their skills further, with more confidence, no matter which industry they serve.
Like Obayashi and ConXtech, you can collaborate with Autodesk Research to investigate how to apply AI/ML in your daily work. If you are interested in collaborating with Autodesk Research’s AI in AEC team, please contact Mehdi Nourbakhsh, a research manager and principal research scientist at Autodesk.