Introduction
About a hundred years ago, architect Le Corbusier’s Modulor Man established a harmonious relationship between human proportion and spatial design (Chaillou, 2019). Since then, humanity applied logic to blend function and beauty in living environments, inspiring generations to push the boundaries of creativity.
Today, in an age of technological advancement, architecture is experiencing a fascinating paradigm shift. Artificial Intelligence (AI) is redefining the design field, enabling the construction of state-of-the-art buildings with cutting-edge efficiency. The synergy between human intuition and AI unlocks new doors – from producing high-definition conceptual imagery to mimicking the living patterns of social insects. The dynamic solar-responsive facades of Al Bahr Towers or the reconstruction of Notre Dame Cathedral evidently portray how AI-driven tools are heralding the transformative era of modern architectural design.
The Phases of Modern Architectural Design
In the early twentieth century, grids and human-centric proportions were the basic principles of design. The era echoed the modernist mantra of the house as a machine. Visionaries like Buckminster Fuller and Kiesler experimented with dynamic and fluid forms, reimagining the spatial interaction between humans, nature and technology.
By the 1980s, computational design emerged, empowering architects like Frank Gehry to break free from rigid grids using early CAD-CAM software. (Chaillou, 2019, p. 13)The advent of tech-savvy applications like SketchPad marked a new era of parametricism – allowing architects to play with the geometry of forms by controlling the design parameters. Building Information Modeling (BIM) evolved in the early 2000s, enabling architects to explore complex designs with 3D modelling. (Chaillou, 2019, p. 11) Today, smart tools like generative AI, diffusion models and self-learning algorithms mark the pinnacle of AI-driven design. As Stanislas Chaillou notes, these four stages are not isolated but interwoven, paving the path of modern architectural design.
AI-driven Tools in Modern Architectural Design
AI and its subfields of Machine Learning (ML) and Deep Learning (DL) are significantly transforming modern architectural design. (Shema, Abdulmalik, 2024)
Machine Learning (ML) is revolutionising architectural design by enabling more data-driven, efficient, and innovative approaches. (Shema, Abdulmalik, 2024, p 164) It facilitates systems to learn, adapt and improve from vast datasets without being thoroughly programmed for each task. It has reduced the need for human intervention in operating the programmes. (Shema, Abdulmalik, 2024)
Neural networks (NN) are an integral subset of Machine Learning. They are trained in data collection, containing input features like building dimensions and material properties and desired outputs like optimised layouts and structural stability. By repeating patterns, they provide architects with multiple alternative design solutions. (Cudzik & Radziszewski, 2018). By acting as the brain of the system, Neural Networks enhance human-machine collaboration. (Shema, Abdulmalik, 2024).
Another fundamental subdivision of Machine Learning is Deep Learning (DL). It can develop innovative concepts, automate the design development process, and predict user response. (Shema, Abdulmalik, 2024).
Key Applications Reshaping Modern Architectural Design
● Generative design
Explores various design options using algorithms to produce solutions based on predetermined parameters and constraints. The predictive model studies existing designs to suggest new alternatives in architectural layouts, facades, and interior designs, and forecast performance outcomes. (Shema, Abdulmalik, 2024)
● Natural Language Processing (NLP)
NLP is a subfield of AI that allows algorithms to communicate intentions in human language allowing AI to convert those ideas into actionable solutions. (Matter & Gado, 2024)
● Image Recognition and Analysis
Used in Deep Learning models to recognise and analyse images. This can be applied in site analysis, identifying structural issues, analysing design trends and assisting architects in deciding on the desired style.(Shema, Abdulmalik, 2024)
● Predictive Analysis
Predictive models can forecast building performance, energy usage, and maintenance costs, (Shema, Abdulmalik, 2024) helping architects make the most economical design decisions.
● Virtual Reality (VR) and Augmented Reality (AR)
These technologies create immersive experiences where designers and users can experience new realities. DL models can enhance VR and AR by generating more realistic simulations and interactive environments. (Shema, Abdulmalik, 2024)
● Optimisation and Automation
Optimisation attempts the best solution from many probabilities. Automating repetitive tasks and optimising complex design processes, makes workflows more systematic.(Shema, Abdulmalik, 2024)
● Sustainability and Smart Buildings
Sustainable architecture aims to reduce the environmental impact of buildings. DL models can analyze and predict the performance of sustainable design features, contributing to smarter, more eco-friendly buildings.(Shema, Abdulmalik, 2024)
How AI is Molding Modern Architectural Design Process
According to Steenson, 2022 Architectural projects typically follow several development phases: conceptual or pre-design (PD), schematic design (SD), design development (DD), construction documents (CD), procurement (PR), construction administration (CA), and operations (OP) for building management. Artificial intelligence is turning the time-consuming design processes into a limitless canvas of creativity.
● Conceptual or Pre-design Phase
Understanding the nature of the project, and client requirements, determining the project’s scale, and detailing all tasks until project completion is decided on this stage.
Artificial Intelligent Agents can generate parameters for a design with climate, topography, FAR, zoning, accessibility, and numeric (number of floors, rooms, etc) data. The system can perform site analysis, spatial distribution, and feasibility studies using the data and function layout automatically. (Wenjun & Malaeb, 2022)
AI’s ability to create conceptual 3D visualisations of the desired project early on is taking modern architectural design to a groundbreaking level. Patrik Schumacher, principal at Zaha Hadid Architects, revealed that the firm is utilising AI-driven text-to-image generators, such as DALL-E 2 and MidJourney, to explore design concepts for their projects. (Barker, 2023)
● Schematic Design Phase
The schematic design phase outlines a preliminary layout of the complete project, determining the roadmap of the latter phases. Platforms like Aino can accelerate the process by gathering GIS and zoning data, whereas Giraffe allows for scene testing with design, financial, and management features. (Brownell, 2024). Generative Adversarial Networks (GANs) can draft initial blueprints while style transfer models can trial with desired aesthetics. For example, Neural networks were used to reproduce the historical design of classical Corinthian column capitals. (Cudzik & Radziszewski, 2018).
Moreover, evolutionary algorithms can optimise materiality, structure, and sustainability – providing the big picture in less time and effort. All these advantages play out to reduce design risks and finalise the schematic planning with a clear idea of the project scale, programme, function and building code.
● Design Development phase
The design development phase transforms conceptual ideas and schematic blueprints into detailed layouts, forms, and functions. This includes architectural plans, sections, elevations and producing the final model of the project. Machine learning models can assist in generating optimised floor plans and predicting performance based on factors such as light, noise, and climate. Simulation tools allow architects to explore multiple design options. (Matter & Gado, 2024)
AI-driven parametric forms leverage smart algorithms to originate modern architectural design. (Wenjun & Malaeb, 2022). Zaha Hadid Architects’ design for the Heydar Aliyev Center in Azerbaijan reveals how advanced technology can be harnessed to achieve a futuristic, fluid structure that balances performance with aesthetic appeal. Building Information Modeling (BIM) takes the planning phase further by integrating 3D modelling with additional dimensions like time, cost, environmental analysis, and lifecycle management. (Wenjun & Malaeb, 2022)
●Construction Documentation Phase
AI-generated cost and time prediction models can prevent cost overruns on large-scale projects. AI can monitor on-site jobs and risks, allowing machine learning agents to evaluate subcontractors for lower risk. Robots can capture 3D images, conduct site inspections, and generate accurate reports that align with project plans. (Shema, Abdulmalik, 2024)
This saves time and labour and reduces plan lag in time and cost.
●Construction Administration and Operations
Machine Learning systems can monitor building systems and predict maintenance needs, reducing downtime and extending the lifespan of building components. Precast structures, including beams, columns, slabs, walls, claddings, and HVAC systems, can be assembled by robots in factories and cast by humans on-site. The Edge, an office building in Amsterdam incorporates Internet of Things (IoT)-driven technologies to provide BIM-like benefits such as automated energy performance visualisation and building usage monitoring. The Edge is also known as “a computer with a roof,” exemplifying smart building design and sustainable innovation. (Walters, 2018)
Conclusion:
With the advent of AI in modern architectural design, ethical and cultural considerations must remain central – ensuring inclusivity, and preserving cultural heritage. Architects, engineers and data scientists can play a pivotal role in producing more human-centric spaces using groundbreaking technology.
REFERENCE LIST:
Chaillou, S. (2019) Ai + architecture: Towards a new approach, Harvard University. Available at: https://www.academia.edu/39599650/AI_Architecture_Towards_a_New_Approach (Accessed: 14 December 2024).
Cudzik, J. and Radziszewski, K. (2018) (PDF) Artificial Intelligence Aided Architectural Design, researchgate.net. Available at: https://www.researchgate.net/publication/328018944_Artificial_Intelligence_Aided_Architectural_Design (Accessed: 14 December 2024).
Bonabeau, E., Dorigo, M. and Theraulaz, G. (2003) (PDF) Swarm intelligence : From natural to Artificial Systems / E. Bonabeau, M. Dorigo, G. Theraulaz. Available at: https://www.researchgate.net/publication/31729435_Swarm_Intelligence_From_Natural_to_Artificial_Systems_E_Bonabeau_M_Dorigo_G_Theraulaz (Accessed: 14 December 2024).
Wenjun, M. and Malaeb, J. (2022) AIA Artificial Intelligence in architecture general understanding and prospective studies, Artificial Intelligence in Architecture. Available at: https://www.academia.edu/40398871/AIA_Artificial_Intelligence_in_Architecture_GENERAL_UNDERSTANDING_AND_PROSPECTIVE_STUDIES (Accessed: 15 December 2024).
Matter, N.M. and Gado, N. (2024) (PDF) Artificial Intelligence in Architecture: Integration Into Architectural Design Process, researchgate.net. Available at: https://www.researchgate.net/publication/378855052_Artificial_Intelligence_in_Architecture_Integration_into_Architectural_Design_Process (Accessed: 14 December 2024).
Brownell, B. (2024) Ai is Shaping Architecture’s new reality faster than we expected. | architect magazine, architectmagazine.com. [online] Available at: https://www.architectmagazine.com/technology/ai-is-shaping-architectures-new-reality-faster-than-we-expected_o (Accessed: 15 December 2024).
Barker, N. (2023) Zaha Hadid architects developing ‘most’ projects using AI images says Patrik Schumacher, Dezeen. [online] Available at: https://www.dezeen.com/2023/04/26/zaha-hadid-architects-patrik-schumacher-ai-dalle-midjourney/ (Accessed: 15 December 2024).
Walters, A. (2018) The edge Amsterdam – showcasing an exemplary IOT building, Centre for Digital Built Britain completed its five-year mission and closed its doors at the end of September 2022. Available at: https://www.cdbb.cam.ac.uk/news/2018CaseTheEdge (Accessed: 15 December 2024).
Shema, A. I., & Abdulmalik, H. (2024). Artificial intelligence (AI) in architecture and design. In H. R. Husain (Ed.), AI-driven architecture: Pioneering the digital frontier (pp. 145-186). Alanya University.
Prada, A., Gasparella, A., & Baggio, P. (2018) On the performance of Meta-models in building design optimization, Applied Energy. Available at: https://www.sciencedirect.com/science/article/abs/pii/S0306261918306615 (Accessed: 16 December 2024).
Images:
Graft Architects (2023) Expo 2030 German Pavillion created with Midjourney. [Photograph] (Berlin: Artificial intelligence and architecture)
George B. 2015. Frederick Kiesler’s Endless House, from 1958. [Photograph]. (New York: The Museum of Modern Art)
icd.uni-stuttgart.de. 2015. Photographs by ICD/ITKE University Stuttgart. [Photograph]. (ICD/ITKE Research Pavilion 2014-15:https://www.icd.uni-stuttgart.de/projects/icditke-research-pavilion-2014-15/)
icd.uni-stuttgart.de. 2015. Development Process images by ICD/ITKE. [Photograph]. (ICD/ITKE Research Pavilion 2014-15:https://www.icd.uni-stuttgart.de/projects/icditke-research-pavilion-2014-15/)
archdaily.com. (2015). Diagram integrated design criteria. [Photograph]. (ICD/ITKE Research Pavilion 2014-15 / ICD/ITKE University of Stuttgart:https://www.archdaily.com/770516/icd-itke-research-pavilion-2014-15-icd-itke-university-of-stuttgart?ad_medium=gallery)
dezeen.com. (26 April 2023). Schumacher said he felt “empowered” by the technology. [Photograph]. (ZHA developing “most” projects using AI-generated images says Patrik Schumacher:https://www.dezeen.com/2023/04/26/zaha-hadid-architects-patrik-schumacher-ai-dalle-midjourney/)
Karanouh, A. and Kerber, E. (2015) Innovations in dynamic architecture, Journal of Facade Design and Engineering Available at: https://www.researchgate.net/publication/283683836_Innovations_in_dynamic_architecture (Accessed: 17 December 2024).
neom.com. [Photograph] (https://www.neom.com/en-us/regions/thelin
