The practice of architecture, its traditional design processes, methods, and technical know-how are currently being changed by emerging technologies. This includes the use of creative tools such as 3D printing, robotics, cloud computing, and artificial intelligence (AI). These integrations have sparked many debates about the effectiveness of emerging technologies when applied to architecture design. This article will explore the effects of AI on architectural design, construction, and urban planning.
Mixed Viewpoints on AI in Architecture
Within the architecture community, some practitioners are worried about adopting AI. Some even fear it, as can be deduced from Natasha Luthra’s words at the AIA Conference in 2018. In response to questions about the integration of AI in architecture, the chair of Technology in Architectural Practice stated that “The entire industry could change on us on a dime, we have to find a way to get ahead of the revolution before we get flattened by it.”
On the other hand, some commentators view the integration of AI as something not worth worrying about. In fact, approximately 81% of designers believe AI will never match human creativity. While both beliefs may be viewed as extreme, recent case studies indicate that the answer lies somewhere in the middle.
Automating Architecture Design with AI
Traditionally, the design of architectural floor plans has always been an important aspect of architecture. The art of communicating the specifications of such plans has itself gone through many evolutions since the humble pencil and drafting sheet. One evolution was introducing computer-aided design (CAD) to make the process more intuitive and accurate, and now AI is being applied to automate the CAD process.
Automating Floor Designs – Today, designing floor plans requires extensive research into the intent, features, and designated parameters before deciding how the optimum design can be accomplished. Traditional research involved perusing perhaps dozens of past projects and ideas before choosing to stick with a particular process or combining multiple design processes. But AI can change all that. AI’s ability to adopt large data sets and analyze them to produce multiple end results brings a new perspective to the process of architecture design. It achieves this by working within set constraints – which could be dimensional, a chosen architectural style, or topological in nature. With these constraints and an understanding of core architectural design procedures, AI frameworks can produce multiple viable design options in a matter of seconds.
An example of this was the effort put into applying AI to architecture design by students of the Harvard Graduate School of Design. In this case study, Generative Adversarial Neural Network (GAN) models were used to automate the creation of building plans and the design of space across diverse scales and constraints. The AI framework used was built and provisioned using:
· historical design data
· Space orientation data
· Floor connectivity information
· Dimensioning templates
The AI framework produced viable building plans that integrated the key metrics and parameters it was fed with. More importantly, AI delivered real-time sequential updates or changes to shop floor designs as new parameters and constraints were introduced into the framework.
Enabling Generative Design – A key benefit AI brings into design processes is generative design which the automotive, aviation, and other manufacturing-based industries currently exploit. In architecture, generative design can be viewed as a more advanced form of parametricism. Generative design takes the architectural style of parametricism to new heights by adding the ability to handle complexities within the architectural design in real-time.
For example, while parametricism links the elements, architecture, spatial relationships, and urbanism when designing complete building systems, the integration of AI speeds enhances the process in different ways. These ways include the real-time analytics of how the elements, human movement and urbanism should be reflected or integrated into building designs.
An example of AI’s application in generative design involves the use of BASILICA, a design system reliant on AI and machine learning to handle architectural and environmental design. BASILICA was used to design complex buildings in Taiwan. The design system has also been used to redesign medieval towns that existed centuries ago in Italy using the constraints and resources with which it was fed. Its pattern recognition and machine vision developed building designs that took into consideration the topography and environmental relationships around them.
Remote Site Analysis with AI – Other examples of using AI to simplify the extensive workloads and research that goes into architecture design and urban planning exist. These include the use of AI to conduct an extensive analysis of spatial networks without having to actually visit physical sites. An example of this is the use of depthmapX to analyze sites and produce a map of spatial elements that are connected via featured relationships. The results can then be used to inform the building of urban areas where certain considerations such as visibility levels must be considered.
Thus, an architect can design fluid urban areas with visual hotspots that can be seen and accessed by residents as they move within the created space. It can also be used to model the visual areas an individual sees when walking through an area and these results can provide a spatial map used to determine the location of high-rise buildings.
A couple of design firms are currently taking advantage of AI’s impressive ability to perform real-time analytics, generative design, and the automation of architecture design and urban planning. One such example is Japan’s Daiwa House Group. The enterprise currently makes use of AI and generative design to enhance its ability to solve architectural issues within its urban enclaves. The integration of AI led to streamlining and speeding up the process of architectural design for its clients.
Optimizing the Design of Smart Cities – Smart cities produce a rich array of data that AI can thrive with when designing, planning, and building smart cities. According to Gartner, there will be approximately 20 billion sensors deployed across the world by 2020 and the data they capture provide the information needed to allocate resources in smart cities. If the captured data is properly harnessed, AI can be used to design and manage public amenities relating to transportation, healthcare, and road networks within smart cities.
Achieving Your Design Goals with AI
While AI remains a powerful tool that can drive the next paradigm shift in architectural design, it is still a tool. This makes it an extension of the traditional pencil as it requires experienced architects to build the frameworks and provide the resources it needs to automate, streamline, or speed up architecture design processes. Thus, its main objective is to assist architects in creating viable design options in less than ideal situations. Learn more about integrating AI in your architectural designs by contacting us today.