Here at Make the Dot we’re not new to the world of AI, or fashion Ai, nor how AI could be potentially used in the fashion industry throughout the value chain. In fact, we’ve actually been testing how AI could be used in fashion design in a few different forms for almost 3 years now, using predominately GAN technology:
In summary we initially researched a few ways in which AI could be implemented in the fashion industry:
None of which got any real traction at the time 👎
Maybe we were too early because at the moment anyone can see, hear or talk about in the world of fashion tech is AI for fashion, and it's potential uses.
So, dusting off some old data, and digging through my g-drive I thought I’d revisit our past research on AI for fashion design and provide a bit of first-hand info on the topic.
Read on for primary research and use cases straight from the source, not LinkedIn influencers 😉
If 2022 was the year of NFT hype, then 2023 appears to be the year of Artificial Intelligence (AI). From mid-journey allowing users to create visualizes in seconds from just a few prompts to ChatGPT writing copy better than some humans, AI is everywhere we look 🦾
But how does that translate to AI in fashion? And how can AI actually be used in the fashion industry, during the whole process?
Well, scratch below the surface and move past the hype and you can see that AI has the potential to revolutionize the fashion industry.
The key word is “potential” 👀
From AI fashion design and manufacturing to retail and customer engagement to AI-powered virtual styling assistants and automated production, AI is changing the way fashion is inspired, visualized, created, marketed, and finally consumed.
This hyper-growth in technology over the past few months also raises more questions than we currently have answers for.
Ok. Let’s begin right at the start of the process. AI for fashion design 👚
Let’s walk through the role AI for fashion could play in the first part of the process, namely inspiration, design, and creating concepts.
As a start, AI-powered platforms have the potential of assisting fashion designers in creating new designs, prints, patterns, and eventual styles faster and easier than ever before. For example, the technology has the potential to analyze historical design trends, types of products that have sold well, colors, prices, and potential markdowns and make suggestions for future concepts.
Although in it's infancy at the time, this was a product the team at Make the Dot built and implemented for a number of large fashion brands, with the ability to speed up the design process and create products that are more likely to be purchased by end consumers, based on historical data 📈
Additionally, on the 3D product development side of things, AI technology can assist designers in creating digital sketches, as well as direct 3D CAD models of styles used for prototyping, sampling, and virtual try-on, saving a lot of time, effort, and resources that would otherwise be spent on many physical sample rounds. With the world of fashion moving from 2D physical products to 3D digital product development, the addition of AI into the process to further increase efficiency seems like a natural evolution.
Want to know which clothing brands are using AI, for trend analytics and forecasting, there’s no surprise that Nike, Adidas, H&M, and ASOS are leading the way, with a large amount of data they possess a huge asset to train their respective AI models on.
There’s also another great use case for using AI for fashion to speed up the inspiration process, but I’ll save that for a later blog, or maybe a stand a lone AI fashion technology product 🤫 #watchthisspace
Next up, we have production.
Dirty, important, ripe for disruption 👨🏾🏭
On the fashion production and manufacturing side of things, AI can assist in large parts of the process. From automated cutting and sewing to intelligent inventory management in factories and warehouses, AI technology for fashion has the potential to further streamline the production process and reduce costs.
Let me give you a simple example.
A factory producing womenswear wants to improve the efficiency of its it’s production, so it turns to automation as well as AI for fashion technology.
They automate the cutting of 2D patterns so that it’s more efficient than a human, but also layer on top AI technology that ensures the cut is 100% perfect. Thus reducing waste, while learning and getting better every time ♻️
Could AI for fashion technology be the silver bullet in fashion’s quest for sustainability?
Although absolutely not a sustainable brand by any measure, Shein is a fashion brand that uses AI in this way in order to optimize its production process.
Finally, supply chain and inventory management are controlled using AI for fashion technology, ensuring orders are never missed, and the exact SKU count is produced at the exact time, thus overproduction becomes a thing of the past and the supply-demand gap is closed.
All of this means, the quality of the end product improves, waste is reduced and products are created faster.
More accurate ✅
Not a bad improvement 👍
Last, but by no means least. Consumers 🤩
Because what’s the point in all that efficiency of using AI for fashion if there’s no one shopping for the product?
From virtual styling assistants to personalized product recommendations when browsing, AI for fashion can potentially every part of the value chain, helping retailers with engage customers in new exciting ways, while improving the overall customer experience.
Brands like Amazon using AI have already implemented this technology in their platform by recommending products based on past purchases, but this process could go into overdrive and transform the industry as AI for fashion technology matures 📊
A start could be AI-based styling assistants helping consumers to find the best clothes by analyzing their preferences, body shape, and styles, all of which will save consumers time and friction and should lead to an increase in their likelihood of making a purchase.
In particular, fashion brands like Zalando and Unspun are using AI in this way to improve the overall fit and sizing for customers.
Another way a brand could use AI is to remind customers of items in the basket (currently just under 70% of all items in the basket are never actually purchased), or when there are trending colors, styles, designers, and collabs they like 🗑️
From our experience with fashion and apparel brands, this is an area we’ve seen brands invest in the most in the past because of its direct and immediate impact on sales and I see no sign of this stopping, as the sales and efficiency gains will only compound by further implementing this AI technology in fashion.
Now we've spent some time walking through the value chain, I want to spend the final part of this blog talking about some of the other powers and questions machine learning tools, algorithms, and innovation pose to the fashion industry.
AI can be used to improve the sustainability of the fashion industry in a variety of ways.
We've spent the best part of this blog talking through the opportunities, but what about the potential risks associated with using AI in fashion?
Overall while it's obvious there are a lot of potential benefits to AI in the fashion industry, it also raises some ethical concerns that should be addressed before it can be fully adopted into the mainstream.
Well, so we say....
Now we’ve been through the whole AI for fashion value chain and answered the question of how AI could be used in the fashion industry, how do I feel AI fashion will develop in the short to medium term?
Well, rather than conclude with a nice, neat summary I’d like to suggest a potential near-future scenario, let’s say 5 years from now 🥸
I’ll add a couple of caveats, however:
With the fine print out the way, let's get into the story 👇
It’s around Valentine's Day and you’re lonely.
The AI in your health tracker knows this as well because it can tell your endorphins are low and you’ve been browsing your favourite fashion website more frequently than normal. Which of course, the fashion brand now knows 💕
What does your favourite fashion brand do about this?
Well obviously, it uses it’s own AI fashion technology to create a quick mockup of your favourite jacket using colors, design references, and styles you love, and suggests this to you at a 3D render at a price it knows you can afford given the time of the year and your last salary raise (You think big tech doesn’t have that data?) 🧥
Naturally, you give in to temptation given the data and AI models know you better than yourself these days.
Once you click that “buy” button a signal is sent to the factory to produce this product, on-demand, quickly, without waste, given the strides in AI technology in fashion manufacturing and production 💨
The AI books your cargo and the cheapest, faster and quickest mode of transport, and within a few days your new, personalized, tailored coat is at your door, and low and behold your endorphins are up again 😍
Just in time for Valentine's Day.
Your health tracker knows ❤️