Can automation in fashion transform the design process?

Imagine if a machine could design, cut, and create several garments in minutes with limited human intervention. Sounds like a reality we may not see for some time, but if you’ve been paying attention, it’s already happening.

Automation and AI have transformed several industries, and now the robots are coming for fashion.

Take the German fashion marketplace Zalando, for instance, which spearheaded AI-powered fashion design based on customers’ favorite colors, textures, and other style preferences in collaboration with Google. Or Synflux, which, in 2019, collaborated on a project called Algorithmic Couture that uses machine learning to create optimal fashion pattern modules using computer-aided design tools.

Not only do these projects help make things easier for brands, but they also have the ability to create more revenue streams alongside business models and help retailers reevaluate their product-related processes.

Plus, it can have a “dramatic impact” on the design process, Erin Schmidt, an independent apparel consultant, told Retail Brew, pointing to the example of OpenAI, which developed CALA, a design image generative-AI platform which pops out a design in a few seconds based on a word prompt. “It would fast-track sketching and ideas into one step and would reduce design teams,” she said. “This could also impact the design process over the long term away from mood boards and collection themes to curated pieces.”

Cut, copy, automate

Beyond that, an integral part of the design process is cutting and sewing—which has also been disrupted by automation thanks to companies like Ditto, a sewing tech firm, that combines algorithmic intelligence with digital projection to make patterns paperless, customizable, and adaptable to specific body measurements—all in real time.

The tech has been over three years in the making and took 10 rounds of consumer research. With the ability to be set up within minutes, the platform features original patterns alongside options from major pattern producers.

It allows sewists to create customizable fits with accurate precision, according to the company.

“It [targets] somebody who is going to be looking to sew to get rid of the pain points of using paper patterns,” Carolyn McNeeley, design project manager at Nottingham Spirk Design Associates, a business innovation and product design firm that partnered with Singer and Joann to create the tech, told Retail Brew.

She added that it is aimed at those who want the “versatility” of being able to have access to their patterns anywhere.

While it is not being adapted by retailers yet, innovations like Ditto have the potential to transform small businesses and push the future of made-to-order fashion forward, which is an area of the industry where automation seems to be having a direct impact on presently.

“If you're talking about made-to-order types of products, that's a huge opportunity for a small business,” Brian Ehrig, a partner in the consumer practice of management consulting firm Kearney, told Retail Brew. “Because the No. 1 expense that you have is the product costs, and inventory has to be financed somehow. If you can move to a model that has no inventory, or limited inventory, that would be a game changer for a small business.”

Future motives

Moreover, Ehrig said, “practical AI” can lessen repetitive tasks like taking measurements and making garment specifications. “AI can help a lot in terms of speeding up the amount of time or reducing the amount of time that it takes to do a lot of manual entry,” he said. “Companies spend tons of time on that, and they’re actually very talented and highly paid people doing that kind of thing. So there’s absolutely a business case to do more functional and operational AI to take mundane tasks off people’s plates and let them focus on the most important ones.”

Of course, this would also require wide-scale adoption of the tech, for which most companies, like Ditto, are in the nascent stages. “There are some limitations there,” McNeeley said, adding that for the company to be able to grow, it would need to offer more versatility and options in terms of designs, which is on its radar, but in the future.

Even more “futuristic” types of tech—like DALL-E, which creates realistic images and art from a description—is a long way from being able to generate an archetype image and translate it to the technical specifications required to actually make it, per Ehrig.

“Digitizing product creation has been going on for quite a long time now, and it’s really had slow adoption,” he said. “To effectively design things in 3D, for instance, someone still has to make them and there’s not really a lot of great 3D printing that can be done at scale just yet. You can 3D knit clothing, but it’s very expensive and not too accessible yet.”—JS