Computational Design: Taking the Next Step with Generative Tools

Welcome to the second episode of Archi-Tales, the Interview Series. This interview series offers you a firsthand experience by featuring conversations with students and professionals from the field of computational design and architecture. 

Today, I am interviewing Mahak, a Master’s student in Computational Design and Fabrication at CEPT University. In Episode 1, we explored the essence of computational design, and in this episode, we delve into…

1. Do You Need a Tech Background for Computational Design?

2. What is Generative design 

3. Conventional design v/s or with Computational tools

4. Resources to begin your computational design Journey. 

So if you are looking for answers to any of this, this post got you covered. Talking to her cleared up so many of my personal doubts. She was incredibly kind and explained everything right from the basics, making it easy to understand. I highly recommend checking it out!

So here is a little background of how she got introduced to such an unconventional field. 

Why did she chose to do Masters in an Unconventional Field – Computational Design

She did her Bachelors in Environment and Space Design, where she got introduced to computational design by her faculty. Her thesis was also based on computational design, where she created a product to revive the Crafts of Gujarat. 

Initially it was very non intuitive for her to go for such an unconventional field because mostly computational design was seen mostly at an architectural scale. But after her thesis project, she saw a lot of relevance about computational tools in today’s time, hence she is here at CEPT.  

Do You Need a Tech Background for Computational Design?

Answer is Yes and No. You don’t have to be from a tech background but definitely having an interest towards coding or problem solving would greatly help. She emphasized that it is a very technical and data-driven field. 

So here is what you should be interested in to thrive in this field, which include –

1. Optimising, regenerating and making forms more functional. While saying this, she also clarified the most common preconceived notion – that computational design is all about creating fancy forms. Instead, she explained that form is a result of the process, not the primary aim. 
2. Learning and exploring the maximum potential of any material or product or the use of a product. 
3. Diving deep into the technicalities and base of things. How can I make this better? 

If any of the above resonates with you, computation design is for you. 

Does Computational Design = Abstracting Natural Intelligence?

She also mentioned that there are a lot of systems like cellular automata, swarm intelligence, fractals in nature that we wouldn’t notice otherwise, once you get into this field you’ll understand that even these forms are very complicated. 

From her perspective, it goes 2 ways: 

There is a lot of data available, and computational design is not a new field, but it is coming into recognition only now. There is a lot of exploration done, maybe not under the term computation design but it is a very very old field. 

There are some of the mathematical formula, some derivations which are already done, but there is a lot of information that is still to be decoded after that, because these are the formulas that work for generic cases, but as she explained that your design would have certain aims and parameters which won’t fit what is available, so you would have to design the workflow accordingly

• So if there is already some information there, it has to be taken forward in a more condensed manner.  
• and if there is not enough information, you would have to decode it yourself by making formulas and relationships between parameters, testing and how it would work. 

She gives another example of stomata on leaves to explain how computational tools work. Stomata on the leaves are located in those parts where it needs more oxygen or where there is a best possible way to get sunlight. She explained like there are a lot of layers to even a basic plant breathing, so if you reduce this scenario to a bunch of relationships and function, you can apply these to your design. In her words,” I feel like a lot of work has been done that is already in the books but a lot has to be done yet. Translation of this knowledge to digital mediums has become a challenge, so there are a lot of research labs like ADRL, Oxman, MIT, IAAC working on that.”

What is Generative Design ?

Generative design also falls under the umbrella of computational design. She explained that generative design involves creating through an output survey, where the process is made faster by automatically reducing and eliminating outputs that don’t align with the parameters provided as input. 

The algorithm generated design in a simpler manner, is narrowing it down to have a mathematical equation that will have all the parameters and the constraints which would work like a fitness criteria, according to which it will generate a lot of outputs. Her response highlighted that generative design makes the job of a designer or an architect very quick – basically you can get a summarised version of what is needed for your models. 

She uses Galapagos as a generative software, a plugin available in grasshopper. It allows her to generate a lot of visual outputs and can automatically rank the one that has the maximum compatibility to the inputs. There are options of minimise and maximise which allows more control over the parameters and hence the fitness criteria. 

She also mentions some examples where generative design is being used 

1. Creating real-time renders for designs 
2. AI tools – text to image 
3. Games where you can create thousands of permutation and combination of elements in a 3d Model (such a game, ZHA Code group is working on)