Architecture Inn: Room for (two) more?

One startup has gone so far as to claim that they can ‘build you the house of your dreams without an architect.’ So, where are we now?

As cities grow, new tools emerge to transform how architects, contractors, and users envision and create the built environment. The ancestral house in your village would be greener and more open than the 3 BHK you live in now (without a garden).

The building data and codes are becoming more accessible, thanks to technology. As a result, developers and we (the users) can feed information into the software and generate an optimised design. One startup has gone so far as to claim that they can ‘build you the house of your dreams without an architect.’ So, where are we now?

Currently, architects define design objectives and parameters. Meanwhile, artificial intelligence (AI) and machine learning algorithms create options for consideration. Though these options are already quite holistic, the bar is being raised every day.

Machine learning is extendable to socio-economic, aesthetic, and ideological patterns. It works with functional as well as programming models. AI and machine training now have a profound influence on how we shape the built environment. This comes as a turning point in architectural history.

Let us discuss these ideas and new building technology trends in this article.

Vladislav Doronin, CEO of Aman Resorts, recalls, ‘I told her [Zaha Hadid], “I want to wake up in the morning and I just want to see the blue sky. I don’t want to see any neighbours and I want to feel free.” She told me, “Do you realise you have to be above the trees?” and she just took a napkin and drew the sketch. I looked at it and I said I liked it, and this is how we started The Capital Hill project.’

Whether on a napkin, on a tracing paper, or on a black CAD background, a lot of the work of an architect goes into creating and reproducing lines. Lines of forms, objects, images — you name it. Run, start over, repeat — then repeat some more. It is a long and winding road from the initial idea, to the final project.

It is all about choices. Even one change influences countless other elements in the process of decision-making. Thus, this is an exercise in choosing the advantages and concessions.

These choices may take several forms. From determining the area to cover, to fitting as many tables into an office as possible without loosing good circulation — it is all a choice. So is minimising environmental impact.

Think of the position of the window in your bedroom. Its placement is decisive for the placement of the bed as well as the energy consumption of the building — no matter the aesthetics.

Of course, tight deadlines and budgets exist in every project. Customers devote hurried time to thinking about every combination. The precise adequacy of every decision is weighed. Enter, generative design.

Generative design who?

When multimedia and headers join forces, a phenomenon is born. Every function and method of design is interpreted along with these two ideas. This makes entities ripe for digitisation and integration with existing information.

So, architects can import existing designs or design components from libraries — line by line — instead of creating designs from scratch. Why stop here? Let us crank it up a notch.

Since design represents objective, stored information (and information has parameters), it is possible to achieve designs by reverse engineering those parameters.

This, is generative design.

The computer generates design potential based on machine learning. How? By a process wherein large quantities of data are fed into the computer, along with the requirements of the building.

Machine learning is about breaking a million examples into bits. Next, these bits are assigned, and algorithms are created. These algorithms can predict the suitability of the design parameter according to a user’s needs.

Automation has finally made its way to our workstations. If a few years ago, we believed that technology will replace everything, save design and creative tasks — we could not have been more wrong.

The algorithm, Finch, produces various spatial configurations in accordance with default parameters. As the entire space area changes, so does the spatial design. This helps to determine areas at the initial project stages. It then can be further refined in accordance with the special task requirements. BOX Bygg and Wallgren Arkitekter have developed this algorithm, and it is available in Grasshopper, for now!

When parameters such as climate or the number of rooms or materials or colours are set, the computer uses the algorithm to combine design elements best adapted to your specifications. The more data you feed, the more you detail that data with parameters and relationships, and the more exact your statistical knowledge, the better the set of designs the computer develops. Of course with appealing renderings and animations, which you can show on the VR Headset, by a click of a button!

Still, an algorithm is just an algorithm. The human decides which problems to solve, which goals to achieve and which factors are crucial. Computers can help organise and prioritise these decisions, but they cannot actually make them. Only people can.

Generative design gives architects, engineers, and builders new freedom to design and create a better world. Sounds reasonable, does it not? How much power do you imagine a person or a company, that has the code and the software to do so, will have?

We are incentivised to answer these questions by being a ‘digital architect.’ Newer permutations and process combinations provide more than just ease. They provide us with the potential to create new content from existing information packages.

Being digital does not only mean using the software and producing high-quality renders. Just as we imagine the processes and functions we perform every day as digital, we must also imagine the processes of design, construction, management, and communication, digitally.

Peter Eisenman rightly says,

‘The problem with digital architecture is that an algorithm can produce endless variations, so an architect has many choices.’

About the Writer
Sana Paul is an undergraduate architecture student and writer at the Jamia Millia Islamia, Delhi, hailing from the cozy streets of Punjab. She has experience working at the India Lost and Found (ILF) by Amit Pasricha, and Rethinking The Future (RTF).

About the Editor
Nishtha Singh is an editor, writer and researcher in the fields of Philosophy of Language, Ethics and Artificial Intelligence (AI). She has trained as an editor at the Seagull School of Publishing, Calcutta and is a graduate of the Department of Philosophy, and the Hansraj College, University of Delhi (DU), India.

About the Illustrator
Diksha Garg is an undergraduate architecture student at the School of Planning and Architecture Bhopal, hailing from Chandigarh. She is an illustrator, graphic designer and writer. She has received a citation for G-Sen Trophy and a Juror’s Choice Award for Journalism Trophy by the National Association of Students of Architecture (NASA), India.