So alien, yet so human.
Will generative design transform the job of a designer (and what it means to be a designer)?
Will generative design transform the job of a designer (and what it means to be a designer)?
I do not consider myself a designer, at least not in the traditional sense. But I do believe that design, perhaps supported by expert designers, is everyone’s job — inspired by the words IBM’s GM of Design, Phil Gilbert. So many of us are, in one way or another, preparing blueprints, implementation plans, designs of future products, services or market strategies. In our daily lives, we plan holidays, furnish our homes, create gardens. Sometimes we hire professional designers to help us, sometimes — where our skills allow for it — we do it ourselves.
Professional designers, especially in the component or product design space, use computer software to help them be more productive. They come up with design ideas and express them using software tools. This approach makes the design process more efficient and helps in interactions with both the end users (easier to communicate the designs) and in handing over to stages following the design (production or other forms of implementation).
It is about to change.
Since late last century, computer scientists, together with designers, have been exploring automating the generation of designs. In principle, the challenge of automating design has been considered in a similar way to any other optimisation, or search, task. If you think about it, the automatic design is not that different from using your GPS:
identify the current situation,
specify the end goal of the process,
find or generate all possible ways of getting from step 1 to step 2 (routes or new designs),
assess and prioritise the outputs from step 3.
Sometimes step 1 is very easy (when a new entity is created), and sometimes it needs a bit more work (for instance when redesigning an existing object — this information needs to be provided to the system).
The major difference between the “traditional approach” and generative design is in who is responsible for step 3 — finding or generating all possible new designs. Traditionally, humans would apply their creative skills to generate these. In generative design, it is computer algorithms that produce a large number of designs that are merely reviewed by humans. In a way, the computer algorithm is coached by a human. It becomes an apprentice, while the master (human designer) is only providing directions and feedback.
Imagine a designer wants to create a new chair. They will need to specify the end criteria. The chair might need to be of the right size for a human to sit on it. There might need to be a specific area of support. It might need to be able to withstand a certain weight, not topple, and so on. When this information is fed to a generative design system, it will come up with several options (a set of example outputs for this particular problem can be viewed here). Then, the role of a designer is to review the outputs. In some cases, the initial criteria need to be modified (if the chairs do not look any close to what the designer was hoping to achieve) and the process restarted. Once completed, the process results with a chair design that fulfils all the specified criteria, and hopefully also looks excellent. Just like the one below (I am not kidding, I love it!).
How about this for your next office? Source: Emmanuel Touraine CC BY-SA 4.0
Is generative design viable?
As we observe the shift from the economy of corporations toward the economy of people, more and more businesses look into mass personalisation as their dominant business model (as opposed to mass scale). Generative design can help scale up design such that it is possible to create vast numbers of highly unique, yet functional designs.
We are yet to see whether products created with the help of generative design are sustainable in the market. And we will soon be able to assess it based on real cases. The first chairs designed using the process will be mass produced soon. Adidas and Reebok now produce and sell shoes with soles and uppers created using generative design. If you are in an industry where the ability to quickly generate new designs, and designs that stand out, is essential, you should pay a lot of attention to generative design.
Is this a chair? Samba stool created using generative design. Source: Estudio Guto Requena [CC BY-SA 4.0]
Where to from here?
In this article, I focus on the applications of generative design for new products. But possibilities are much broader than that. Just like design should be everyone’s job, the approach of generative design could be applied in practically any area. Generative design is currently used mostly for product design and art. But why not use generative design to imagine new business models, organisational structures, or perhaps even gardens and holidays? As long as it can be expressed in a set of criteria, it should be possible to apply generative design to imagine or reimagine it!