Parametric design uses algorithms to come up with solutions in terms of architectural forms and functions. They can use relationships and established rules to refine the analysis of designs. Generative architecture, as the name implies, uses software to create design options. Artificial Intelligence generates hundreds of iterations and samples based on a prompt. Regardless, there are many applications that both these tools can provide for architects.

Drafting with Rules and Guidelines

A parameter defines a condition or system using set numerical values or measurable factors such as sunlight, temperature, or humidity. The collected data is then inputted to figure out the best design iteration to respond to these factors. 

In relation to the design process, parametric design is a tool that can greatly increase the efficiency and accuracy of an architect’s design intention. From measurements constraints to shape restrictions, parametric design can adjust the structure’s aesthetic. Using this technique results in designs with more uniform characteristics and geometric shapes. It has a tendency to produce repeated patterns in a symmetrical fashion while still remaining within the confines of the architect’s intention.  

Parametric design overcomes former limitations to introduce innovative structures. Designs can be quickly modified with unerring accuracy and consistency. The automation of repetitive patterns can make adding these properties more feasible. Complexity in forms are generally simplified and better documented for a more streamlined design process. 

Photo Credits:  Ar. Kern Parangan, Kaleej Times, Killa Designs, CNN, and MOTF©

The “Museum of the Future” in Dubai utilizes parametric design techniques to reflect their cultural heritage and visions of the future. Across the facade, you’ll notice Arabic calligraphy inscribed across the stainless steel cladding. These are actually windows that provide natural lighting throughout the structure despite its large size. The use of parametric modeling made the execution of this design possible. 

Architects tested the complexity and engineering of the letters through software that indicated weights and constraints. Furthermore, Building Information Modeling (BIM) helps ensure the building’s sound and sustainable construction. The intersection of different construction technologies allowed the team to navigate through the complexity of the facade. 

Generating Plans and Models

Generative design creates different versions of the same output to meet goals or conditions. These iterations allow architects to prospect based on what the AI or algorithm can create.

Being able to generate a large quantity of possibilities in a short amount of time and effort allows designers and stakeholders to narrow down options to the best ones. The implications and feedback of a design is made more apparent. The generated models are automatically documented to reflect data sets across iterations. Furthermore, designers retain feasibility and performance throughout, and it is up to the computer to figure out what potentially works best. 

Generating iterations involves a complex process of evaluation that allows the computer to learn. From there, the computer continues to simulate more designs as the requirements become more specific until only a few iterations remain. The designer component then comes in to harmonize the digital foundation made with a human touch. 

International firm Gensler and local partner Aidea Inc, utilized generative design for “The Finance Centre,” a 42-story building in Bonifacio Global City. The algorithm allowed the firms to find a form that would make the building taller while still compliant with our local standards. Using this basis, the developer maximized the lot further with additional floor areas.

Limitations of Computational Design

A major criticism against using computational technologies is in how they subvert human creativity in favor of a computer’s algorithms. Professionals may develop a dependence on these tools patterned after human skill in the first place. Furthermore, the data given is the only reference of computational design. One wrong input and the entire output becomes completely erroneous. 

In the context of skill sets among architects, learning how to work with algorithms can be a daunting task. Taking the effort to learn how to use computational tools can be difficult for those with no background in using digital software.

The intersection between technology and the design process continues to evolve. Both parametric and generative design aim to enhance how architects work. Parametric adjusts based on factors while generative creates multiple design iterations. A widespread adoption of computational design should come with an awareness of the implications of integrating such technology in practice.

Download this month's BLUPRINT magazine digital copy from:
Subscribe via [email protected]