The architect’s guide to generative design

Posted on: 15 July, 2024

Generative design is transforming the way architects and built environment professionals approach the design phase of construction projects. Here’s how.

From building information modelling (BIM) to augmented reality and the Internet of Things (IoT), it’s no secret that technology has transformed the way architects work and revolutionised the industry. These tools allow architects to realise their greatest visions and offer new and exciting ways to design and create structures within the built environment. They can also transform existing buildings, communities and even whole cities.

Generative design is a new innovation shaping the architectural industry. But what is it, and how exactly does it work?

What is generative design?

Generative design is a strategy where the architect or user inputs data into their chosen software. This data might include design goals, time or cost constraints, materials and other relevant information.

Once the data has been inputted, the generative design program then uses algorithms to analyse and explore this data to provide all design possibilities and highlight which solutions could be the best. This allows designers, architects and others in the construction industry to work faster and smarter, making more informed decisions based on real-world data.

Why generative design is important for architecture

The traditional method of design is very labour and time-intensive, and there’s often a limited time and budget to dedicate to this process.  Every initial idea is a long way from completion – there’s a lot of back and forth, trial and error and plenty of changes that need to be made before a project can even begin.

Generative design can enhance the existing design process and to make it more efficient and effective for architects and all those involved. By allowing them to explore all possibilities, reduce design flaws and test ideas, designers can propose the best possible solution for the project. Not only this, but generative design can increase their creative abilities, generating ideas based on specific criteria, constraints and goals.

By decreasing the time and money spent on the process and making it possible to generate the most informed project proposal or design, this software supports the design process from start to finish and is transforming the industry. This leads to the best results for both architects and their clients.

The 3 stages of the generative design process

1. Pre-generative design

The preliminary stage defines the goals, parameters and constraints for the project. This will provide the data set entered into the design software during the main phase.

This information should be set out by the relevant architects and stakeholders. The more precise this data is before the generative design process begins, the better the overall results. Some of the details may include:

• Regulations like construction boundaries or maximum heights
• Commercial requirements like the number of dwellings needed
• Structural performance and energy consumption
• Liveability factors like occupancy size, views and daylight hours

2. Generative design

This is the main stage where data is entered into the software and the computer algorithms begin their work. Once the data has been provided, the cyclical process of generative design can start and this itself can be broken down into a further three stages:

• Generate: The human operator will enter a seed into the software so the algorithm can begin creating a variety of different, feasible designs.
• Evaluate: The computer system can then analyse the designs to determine their compatibility based on the requirements, constraints and parameters outlined in the first stage.
• Evolve: Based on the findings, the algorithm will determine which direction the new design needs to go to make sure it is meeting all those requirements. Based on the best design of the current options, a new seed can be introduced and the process will begin again to generate a set of final designs. This could take several attempts.

3. Post-generative design

This is the point where the architect, designers and relevant stakeholders can review the options and choose from the final set of designs. The architect can then take this design and refine it manually to ensure that all requirements or constraints are being catered for and add any finishing touches.

10 benefits of generative design in architecture

Here are some of the key benefits of incorporating generative design into the architecture industry:

1. More creative and technical freedom
2. Quickly and easily evaluate whether a design is suitable and adheres to the requirements or constraints
3. Optimise a building’s performance using parameters like energy usage, materials and occupant comfort
4. Generate designs that are more accurately tailored to the specific needs of the users
5. Analyse parameters such as energy efficiency and material usage that can make building designs more sustainable and environmentally-friendly
6. Produce a number of unique designs in a much smaller timeframe, saving time and money on each project
7. Liberate designers from repetitive tasks through the automation of tedious tasks
8. Offer enhanced flexibility, allowing the architect to consider changing environments and how their designs will fit in an ever-changing world
9. Generate 3D models that respond to advances in technology and other developments in the area, ensuring that their design and building stay relevant for years to come
10. Provide a shared space where architects and other stakeholders can all access the designs at the same time, avoiding delays in the project and further reducing costs

The challenges of implementing generative design

Despite its many benefits, generative design is not without its problems:

• Quality of data: As generative design tools rely on data from a real-world setting, this data needs to be as accurate as possible. Inaccurate, outdated or incorrect information can lead to flawed designs that don’t meet the project requirements.
• Legal and ethical issues: Generative design relies on AI, automation and the use of potentially sensitive data. This can raise questions about ownership, responsibility and liability. There could also be issues around privacy, security and consent of the data sources.
• Innovation and uniqueness: While these tools may produce different design possibilities, relying on AI could lead to the design of homogenised or unoriginal ideas.
• Skills gaps: Generative design tools are complex and sophisticated and architects can find this daunting if they’ve not used them before. This means that many would need to retrained, otherwise the industry could face an ongoing skills shortage.
• Software compatibility: It can be tough to find the right generative design software that is compatible with existing systems and workflows, as this often requires specialised platforms. As such, architects will need to evaluate the compatibility of the software to make sure it will support their design goals and existing
• Collaboration and communication: While it’s great that architects and other members of the team can view the designs on a shared platform, without a clear and transparent framework for decision-making and feedback, this can lead to confusion or frustration.

Learn more: What is the green skills gap (and why does it matter)?

5 real-world examples of generative design

Here are some examples of how generative design is already being used in the architecture industry:

The Heydar Aliyev Centre

The Heydar Aliyev Centre in Baku, Azerbaijan is one of the most renowned creations of the late Zaha Hadid. The curvy, eye-catching building somewhat resembles a spaceship, yet it perfectly meets all design parameters thanks to its use of generative design.

The Shanghai Tower

The Shanghai Tower in China is one of the tallest skyscrapers in the world and one of the finest examples of generative design. The designer, Gensler, used these tools to optimise the building’s shape, making it more wind-resistant and energy-efficient as well as visually appealing and structurally sound.

Daiwa House Industry

With a staggering 92% of the Japanese population living in cities, the need for urban housing is on the rise. Generative design has made high-density urban housing manageable and the Daiwa House Industry is a company that utilises these platforms to create attractive, functional living spaces in small spaces like cities.

San Francisco Transbay Transit Center

Designed by Pelli Clarke, the San Francisco Transbay Transit Center was created to manage the city’s growing transit needs. Generative design tools were used to explore a number of options to ensure the structural efficiency and form of the building.

These tools also helped to create the visually appealing rooftop park on the iconic roof. The rooftop park includes an amphitheatre, restaurant and water features.

Autodesk Technology Centre Toronto

Autodesk’s Toronto offices are a wonderful example of generative design as it was the first AI-designed office of that magnitude, at three floors and 60,000 square feet in size.

As part of the pre-generative design process, the designers gathered feedback from 250 employees on their office preferences and input the parameters into the relevant software to create the most suitable design. As a result, this is a much-celebrated blend of human ideas and machine learning.

Final thoughts

Generative design is transforming the architecture industry and built environment. From saving time and money to saving the planet through sustainable, energy-efficient buildings, both architects and others in the industry are reaping the rewards.

That being said, these platforms are not without their challenges, and architects must overcome these key issues in order to achieve the full potential of generative design. Digitisation and generative design are the future of the industry and upskilling, data management, legalities and implementation must all be addressed in a clear and focused way to get the most from these platforms.

Architectural technologists are the digital experts that bridge the gap between an aesthetic vision and a practical reality. If you want to be at the forefront of the built environment’s digital future, UCEM’s BSc (Hons) Architectural Design Technology will give you the technical expertise and literacy you need to become a valuable contributor to the sector.

Find out more: BSc (Hons) Architectural Design Technology – University College of Estate Management