Tools powered by artificial intelligence (AI) are having a significant impact on many activities in today’s world, including architecture. Many of these solutions are already transforming design processes thanks to their great computing power, which improves efficiency and expands creative horizons.
While AI-driven systems can generate multiple design options based on given parameters and criteria to help professionals explore a much broader range of possibilities, machine learning models make it possible to analyze large data sets from previous designs. In this way, it is possible to identify patterns, trends and preferences, which leads to generating proposals that are better informed and more relevant. On the other hand, AI can also help solve design problems that would otherwise require much more time, energy and resources, or that would be impossible to address in the traditional way.
However, the use of information technology is not new in architecture; it arrived along with the systems that automate representation processes. These drawing software make the process more efficient, but they do not lead to fundamental changes in the design method. For example CAD (Computer Aided Design) systems are digital tools and technologies that have been used for decades to graphically represent a design through drawings and models.
The evolution of these instruments gave rise to BIM (Building Informational Modeling), a work methodology that seeks efficiency by automating the design process. The system uses a digital model of the project which makes it possible to predict the behavior of the building and generate geometric, structural, lighting and bioclimatic studies with a single tridimensional model. The flow of information is no longer linear; rather, it is centralized so that all stakeholders have access to every detail of the project. This also leads to more effective communication than in the traditional system.
Going one step further, the exponential growth of technology and the proliferation of a wide variety of sensors that make it possible to obtain very precise information about the conditions of physical space have enabled the use of digital twins to create a virtual and dynamic representation of an existing structure. This is done with a software model using AI and machine learning based on the data it receives in real time. It allows different scenarios to be reproduced in order to evaluate the impact of changes in the design and use of space, the optimization of energy systems, air conditioning, etc.
Today, the possibilities offered by AI and machine learning are expanding the boundaries of tools in the service of creativity. We are witnessing the emergence of a new scenario for architecture, a practice that, for centuries, has relied on the experience and intuition of professionals to find new design solutions. Technology is changing this process completely.
Computational Design
Computational design is a method that uses the computing power of computers to achieve a given project, following instructions and design rules previously defined by the user.
Thanks to advances in AI and machine learning, designers can use large data sets as a starting point to find efficient solutions. They can also add considerations of behavior in specific conditions and analyses of the life cycle, the economic performance and the environmental impact of the project, among others.
However, while computational design offers great possibilities and is playing an increasingly relevant role in today’s architectural design, it can require specialized knowledge, forcing designers to acquire skills in other areas.
Computational design may be classified into several categories, as defined below:
→ Parametric design. These are models created using parameters, variables and restrictions that, through an algorithm, define the relationship between the design requirements and the resulting solutions. Changing a parameter automatically affects other elements of the design. For example, a parametric design might include a set of parameters that define the height, width and slope of a structure; adjusting any of these values automatically changes the shape of the building.
Parametric design makes it possible to create geometries that were previously more difficult to define, providing the possibility of developing more formally complex projects. This process became very popular thanks to avant‑garde architects such as Frank Gehry and Zaha Hadid.
→ Generative design. Generative models use algorithms and rules to create design options more or less autonomously, in order to provide solutions that meet the formal and performance requirements posed by the user. Generative design-based methods can create complex designs, even starting with simple algorithmic descriptions.
This methodology facilitates decision-making, as it allows a large number of options to be assessed and analyzed in a short time. And although aesthetic and geometric considerations are usually present, these are not its main purpose: the form arises out of the objectives pursued.
→ Algorithmic design. It is a paradigm that uses mathematical and logical algorithms to create and optimize specific design forms, structures and details. It can be more deterministic, more rigid and less adaptable to changes than generative design, since the results are directly linked to the application of certain defined algorithms. For example, such a model could be used to create a façade with specific patterns based on geometric criteria and contextual data.
It should be noted that these models are not mutually exclusive and they are often used in a complementary manner. Parametric and generative designs are often implemented through algorithmic techniques and, together, enable a more profound and efficient exploration of the design, providing flexibility, variability, and optimization.
Advantages and Disadvantages
New design tools powered by AI and machine learning are transforming creative processes and improving the efficiency of built space. These advances have a positive impact on many aspects. Here are some of the possibilities:
→ Generating innovative ideas. AI can generate a wide range of design options, exploring innovative ideas that may not have been considered by designers.
→ Efficiency in the design process. AI algorithms can generate proposals quickly, accelerating the creative process and enabling greater iteration in less time. Automation saves time, money and resources.
→ Generating complex geometries. These design tools make it possible to go beyond the boundaries of simple geometries. Intelligent algorithms make it possible to develop, calculate, optimize and represent complex structures and surfaces that would be impossible using traditional instruments alone. They also facilitate detailed and accurate analysis of the behavior of performance factors such as structural strength.
→ Simulation and analysis. AI can perform analyses and simulations of the behavior of environmental variables and energy performance, as well as carry out the structural assessment of a project. It can also use real-time data on weather conditions, space usage and other factors to dynamically inform and adjust the design over time.
→ Customization and adaptability. These tools can customize the layout based on individual preferences, such as ergonomics, work style, and other specific needs. This helps create spaces that are better adapted to the needs of users. AI can also contribute to the design of adaptable buildings that are able to evolve over time to satisfy the changing needs of their users. This reduces obsolescence and promotes durability, both key aspects of long-term sustainability.
→ Sustainable solutions. Design algorithms can incorporate sustainability data, which contributes to create more efficient and sustainable buildings that minimize environmental impact throughout their life cycle.
AI makes it possible to perform detailed energy efficiency analyses, optimize waste management by identifying strategies for material reuse and waste reduction, and optimize the consumption of vital resources, such as water.
→ Challenges. AI-assisted design offers great advantages in terms of efficiency, optimization, and creative exploration. However, it also poses challenges in connection with human creativity, the interpretation of cultural and social contexts, a dependence on precise data, ethical issues, and initial costs. The most effective approach might be a balanced combination of the capacities of AI with human experience and sensitivity.
The Implementation of AI in Design Processes at Contract Workplaces
by Leandro Boggiatto*, Architecture and Design Manager for Argentina, Chile, Paraguay and Mexico at Contract Workplaces.
Over the years, the company has maintained its focus on the constant search for innovation, while committing to explore and foster new ideas that allow us to expand our boundaries and contribute to business growth. In this context, we have incorporated the use of several AI-driven apps and programs into the design processes to streamline and optimize our proposals.
On the one hand, the use of AI to improve the performance of test fits—the process by which the possibilities of a property are assessed to satisfy the programmatic needs of the future office—has become a first-rate tool. This fit test, which used to take days or weeks with the traditional methods, can now be completed in a matter of minutes and with high-quality, tangible results.
Likewise, we have incorporated AI driven apps to create high-definition renders, resulting in some early visual representations that can be used as a starting point to communicate our ideas to the client. Without a doubt, this is a great advantage, as we were able to considerably reduce the time spent on the preliminary stages of the design process, streamlining decision-making regarding shape, layout, materials, colors, textures and lighting, as well as the project’s general perception.
On the other hand, we are also beginning to use immersive reality so that customers can explore a digital representation of the design proposal at full scale. Our current goal is to be able to make changes while on this virtual tour (e.g., in materials, colors, etc.), for customers to have the possibility of appreciating different options more realistically. This tool would also make it possible to detect and solve potential construction problems.
Based on the advances we have made by integrating AI into our design processes, we can conclude that this technology has proven to be an invaluable resource in improving the efficiency and quality of our work and, above all, in meeting the expectations of our customers at all stages of the project.
References:
CAETANO, I. et al. (2020): “Computational design in architecture: Defining parametric, generative, and algorithmic design”.
GARCÍA TORIJA, A. I. (2021): “Diseño generativo. Algoritmos como método de diseño”.
ROYAL INSTITUTE OF BRITISH ARCHITECTS (2024): “RIBA AI Report 2024”.
RUSU, A. M. (2015): “Geometry and complexity in architecture”.
KAICKER, A. et al. (2019): “Enhancing Workplace Design through Advanced Floor Plate Analytics”.
