Back to Blog
One area artificial intelligence is seeing growth in is the textile industry. AI is finding a home with textile manufacturers, helping with visual inspection jobs like color matching and pattern making. And some companies are using artificial intelligence to assist with quality control, supply chain management, and an overall improved customer experience.
One company using artificial intelligence to help fashion businesses identify their links to China’s Xinjiang region received a public vote of confidence last week when it struck a multi-year deal with US Customs and Border Protection.
Fashion’s days of operating with little to no government scrutiny are over. Now, the industry has to figure out how to comply with all the rules.
That’s not always simple. Brands and retailers operating in the US and Europe face regulations covering forced labour to greenwashing, with more likely to come.
“It’s no longer acceptable from a regulatory standpoint, but even desirable from an economic standpoint, to simply manage your supply chain as a direct buyer-and-supplier relationship,” said Evan Smith, co-founder and chief executive of Altana, a tech-centric supply chain mapping company. “You need to be managing your upstream network as an actual system.”
The solution, Altana believes, is artificial intelligence. It compiles billions of data points on how goods move through global supply chains and applies AI to make sense of it all. The result, in Smith’s words, is “a shared source of truth that is used by manufacturers, retailers, brands, logistics providers and government agencies as a common operating picture.” Like Google Maps for supply chains, he said.
Altana said it already has a number of customers in fashion and textiles, though it couldn’t disclose any names. But its platform received a public vote of confidence last week when US Customs and Border Protection struck a multi-year deal with the company as it works to block goods it believes to be made by Uighur forced labour in China’s Xinjiang.
The issue has been of particular interest to fashion since Xinjiang is the epicentre of cotton production in China, the world’s biggest producer of textiles and clothing. CBP has proved it’s willing to block goods under the Uyghur Forced Labor Prevention Act. Official statistics show it has denied 812 shipments of apparel, footwear and textiles totaling nearly $34 million in value from entering the US since June 2022.
For brands and retailers, it can already be a struggle to determine the origins and impacts of the materials they use for their products. Many are turning to technologies such as blockchain and AI as an answer.
“The bulk of our prospects who are trying to solve the forced-labour problem are from the apparel and textile industry,” said Amy Morgan, Altana’s head of trade compliance. “If it’s cotton, it’s tracing back to a farm, which is almost in some cases easier than synthetics that trace back to oil.”
While CBP agents will use Altana, the purpose of the platform isn’t to flag shipments. It’s to provide customers with a clear, up-to-date picture of the full value chain for different goods as well as insights for businesses, like how to structure your supply chain to take advantage of free-trade agreements, minimise tariffs and manage risk. A research brief Altana created on tracking ties to Xinjiang includes an example of a real apparel seller whose name has been changed. It gets into granular information about its supply chain down to details like shipments of specific fibre blends from yarn suppliers.
But it’s hard enough for fashion businesses to gather this information on their own products, so how does Altana do it?
Smith said they collect a mix of public and non-public information. The public information includes any data they can scrape from the internet as well as commercial data sets they purchase. (They’ve spent millions on those, Smith noted.) But that’s not enough to create the full picture they need, so they also gather non-public data siloed in customs authorities, logistics providers, financial firms and more through what they call “federated learning.”
Altana’s customers can’t see raw data from other companies, but they have access to the insights Altana gains from it — for instance, knowing how a yarn provider connects to a specific textile mill that supplies numerous big brands.
To create a useful picture from the information takes more than just gathering records, however, which is where AI comes in.
The sheer volume of data is overwhelming. For a fashion company to sift through and manually decipher which raw materials went into a yarn that was woven into a fabric they used for a product would be impossible, according to Smith. Altana uses AI to extract the information it needs from billions of records in different languages and formats, and then to analyse them, establishing connections between companies, determining what’s relevant and deriving useful insights for businesses.
Smith ultimately envisions Altana helping companies make international trade easier in other ways, too, such as retailers being able to fast-track clearance of their imports through customs if they share their data with CBP on the platform, similar to the options that let pre-checked travellers speed through security in airports. While they’ve started by appealing to big players such as government agencies and large logistics providers, their goal is to start bringing smaller ones onto the platform too, including fashion companies. (Smith previously ran a business that focused on automating the apparel and textile supply chain, while Morgan started her career on trade issues at Nordstrom.)
The fashion brands harnessing the power of AI
US sport brand NIKE launched the Nike Fit app which allows customers to scan their feet with a smartphone camera for accurate sizing. The AR technology uses a 13-point measuring system to map each foot’s dimensions. The more people that use the app, the more data is collected, and the more accurate the app becomes. This will mean that customers receive the right fit for them and help lower returns, boosting the brand’s profit margins.
UK online retailer Pretty Little Thing allows shoppers to leverage both the text and visual search tools to find an item. This boosts customer experiences, as the increased convenience increases the usability of online channels.
Swedish fashion retailer H&M has been using generative AI to trial designs for its Conscious Exclusive collection to understand which existing designs are most appealing to its customers. It then uses this data via a generative adversarial network (GAN) which proposes new designs to meet customer needs and current trends.
French luxury fashion company Kering has a personal shopping assistant called Madeline, which uses ChatGPT to improve user experience on its KNXT marketplace (an e-shop using NFTs). Consumers can give a prompt to Madeline, providing an idea of what item they are looking for or a price range they are shopping in, and the chatbot will provide them with a recommendation.
UK retailer Marks & Spencer (M&S) uses large computer vision models and CAD data to automate the design and documentation process for products such as furniture, homewares, and clothes. This involves automating tasks such as writing product descriptions and editing photos of products saving time and boosting efficiencies.
UK retailer Next uses generative AI to write responses to thousands of customer queries. This saves time and labour and ensures consistent customer service.
German online retailer Zalando uses ChatGPT to create a tool to help customers find products for their specific needs. The tool will be launched in Germany, the UK and Ireland. Customers can ask questions, and it will respond with relevant suggestions based on the criteria given by the customer.
https://www.just-style.com/news/signal-leading-fashion-brands-leverage-ai-for-supply-chain-efficiencies/?cf-view
