Artificial intelligence has the potential to revolutionize a host of processes and operations, and the prospect has some manufacturers feeling delighted, and others, dubious.
Experts speaking at the Sewn Products Equipment and Suppliers of the Americas (SPESA) Advancements in Manufacturing Technologies Conference in Downtown, Los Angeles on Thursday aimed to quell fears and deliver a dose of reality to the discourse surrounding AI, speaking to both its growing capabilities and its limitations.
Related Stories
Kyle Stephens, a veteran of Wrangler and Lee parent Kontoor Brands and CEO and co-founder of KolAi Denim, said “it can be hard to wrap your head around what it all means” when it comes to advanced technology like AI.
“But every single one of these AI systems takes a prompt of digital information from you and translates it into some other kind of electronic information. That’s it.”
Stephens illustrated a quintessential use case for AI in the design of denim. Desirable design elements like whiskering, chevrons and other finishing details commonly found on a pair of jeans are most often hand-drawn in photoshop by a designer—and it’s a time-consuming process. The breezy, lived-in look that so many shoppers desire from their denim is meticulously crafted behind the scenes. “It usually takes maybe eight hours or a day to get it looking good,” he said.
Throughout his past experience with larger brands, product development teams would have to travel overseas to work with producers, spending hours if not days waiting for the laser files to be developed from a “target garment,” or reference piece. “That just is not an efficient way to work,” Stephens explained.
“If you have a lot of images of target garments and a lot of images of laser files, you can train an AI” to develop future designs. “That’s basically what we did,” he said. “You photograph your garment, you illustrate it to make a file the laser machine can understand and then you take the machine and you apply it.”
According to Stephens, “The AI isn’t always good,” but the designer’s job “has now gotten a lot easier, and it takes a lot less time and requires a lot less skill.” A process that once took a full work day can now be accomplished in 10 seconds, and the designer can simply make adjustments where needed.
“AI is not is not new, it’s been here for a long time,” Ketty Pillet, Lectra’s vice president of marketing for the Americas, added. Tools like Apple’s Siri, for example, have been a part of consumers’ lives for some time. But generative AI tools like ChatGPT, which are capable of creating content using prompts from users, are a newer form of technology that the public is still working to understand and harness.
Pillet believes that any anxieties about AI replacing manufacturing jobs are overblown. The technology is only as good as the information and instruction it’s fed, she said. “You can have the most advanced AI but if you have people who do not have the data or expertise to drive it, it’s not going to work,” she added.
AI technologies can be useful throughout the supply chain from a product’s conception through production. It can also help businesses with inventory planning, using historical data and other market insights to determine what types of products demonstrate the greatest capacity for sell through. Lectra’s expertise and software solutions encompass everything from on-demand manufacturing to product design and development and traceability.
Alvanon, a retail technology platform that creates digital mannequins based on the human anatomy for garment production and patternmaking, is another AI tool developed to help brands optimize product development and drive sales. The group’s Americas executive director, Eric Lee, said such design tools are rapidly advancing and proving their worth to retail.
“With more AI adoption and machine learning, more recently, I’d say in the last two to three years, there’s been a dramatic shift in the ability for the software system to simulate things realistically,” he said. “In the past it could take the software days to stitch together a pattern and simulate the fabric and adjust everything so that it looks possibly realistic.” That’s a lot of time to spend fiddling with software when a brand could just create a physical sample, he added.
But in 2024, these tools are saving real time and money. Issues with digital simulation, like the drape of different fabrics, have been largely worked out. Designers can import a pattern file and quickly see their creations rendered on the screen. “What that means is there’s going to be a pickup in the adoption of these 3D softwares” for design purposes, he said.
Alvanon is also working on creating “digital twin bodies” for web shoppers, with the goal of dressing them in 3D garments. “The holy grail for e-comm is virtual try-on,” he said. While technology providers have been working on the concept for some time, he said progress in the space is accelerating, and successful application could have manifold effects, from driving sales conversion to providing useful data for inventory planning.
Accuracy and speed are of the utmost importance for fashion brands and retailers, according to Tirsa Parrish, co-founder and managing partner of Fashion Index. “Time has always been up against us in this industry, because we have to develop things so far in advance,” she said. “What the customer actually wants a few months later is often very different” from what was initially created, she added.
The best way to combat consumer fickleness is to reduce the amount of time it takes to get a product into their hands, Parrish said. Using digital design tools based in AI and machine learning can speed things up greatly, while also helping ensure that products fit and perform better, “getting a lot more accuracy and helping with specific functions.”
AI can also streamline processes throughout the supply chain, saving brands money in the process, she added. Streamlining “each one of the steps, be it at the front office or the manufacturing floor, integrating all the way to retail and the customer experience, is a tremendous win with cost savings,” Parrish said.
