How Digitization is Changing the Game of Sustainable Fashion (Part 1)
The lines of fashion design and technology are beginning to blur, which is an enormous plus point when it comes to making fashion more eco-friendly. The advantages include saving costs and time, minimizing carbon output (for transporting physical samples), lessening the use of treatments and dyes, and reducing the risk of overproduction and deadstock, thereby less textile and water waste is produced and companies could profit from lesser quantities of products.
While the thought of having fashion and digitization converge is definitely exciting and hopeful, some may perceive it as a threat. Are garment workers and designers going to be replaced by lines of code on a computer? Will algorithms be able to replicate the human touch and creativity on fashion design and garment making?
On another note, the looming threat of a climate crisis apparently fell on deaf ears as fashion brands somehow are going about their business as usual, with the all too common excuse that consumers aren’t demanding for eco-friendly clothing. In spite of that, the fashion industry can’t afford to wait for demands for sustainability as long as they rely on a linear “take-make-dispose” business model and sell as many products as possible. Sustainability remains a niche to this day, which is unfortunate given the sorry state of the planet. Thus the integration of digital transformation into fashion becomes a necessity rather than optional.
The current production practices of fashion are slow, manual, expensive, unethical, and unsustainable. Products were previously launched in the cycle of two seasons per year, which then became four, and eventually became a monthly or even a weekly cycle (that’s 52 seasons of clothing per year!). The increase in production seasons seemed to coincide with the rise of social media, where trends appear out of nowhere just as quickly as it fades; thus manufacturers are pressured to keep up with the trends to garner profit while exploiting labourers and the planet. Besides, garment samples are being shipped back and forth from Asia to the West, generating a hefty financial cost, as well as carbon emissions. Digital transformation could possibly put a stop to all that by virtualizing and streamlining manual processes, covering the full spectrum of the fashion industry from designing, manufacturing all the way to retail. Even so the fashion industry appeared to put digital integration on the back burner due to various factors. We’ll examine the role of digitization in fashion and how it’ll assist, change, and transform fashion as we know it into a more ethical, sustainable, affordable and efficient practice, while instantly gratifying consumers’ lust for novelty, personalization, and identity.
DIGITAL FASHION DESIGN
What if fashion design becomes a digitized platform where fashion designers, enthusiasts, and just about everyone could create designs on screen, share them across social channels, and connect to digital manufacturing facilities to allow production on demand? Isn’t that a more sustainable way of apparel production where overproduction and deadstock is eliminated, and consumers’ demands are fulfilled at the same time?
To begin with, many designers still gravitate towards manual and painstaking traditional illustration techniques, as in, with pen on paper. Others would opt for graphic design softwares which are neither intuitive, nor reflects how fabric would drape on bodies. After the designing process, a technical package (informative specifications on the construction details of the garment), otherwise known as just “tech pack” in the industry, would be put together and sent to factories for pattern-making and sampling. The samples are then sent back for review by the designer, and factories will implement the change accordingly. In a lot of cases, there could be multiple revisions of designs, and multiple revised samples will be sent back and forth until the creative vision is achieved. Such a repetitive and interactive process could cost an immense amount of time (anywhere from 2-4 weeks!), money, and carbon output, especially if designers outsource their production overseas.
CLO Virtual Fashion, a fashion design software company, is about to revolutionize the fashion industry and its pipeline. CLO3D, one of the company’s softwares, empowers designers with a plethora of tools needed to design apparels, and visualise them on readily available avatars. Designers could easily translate their craftsmanship from paper to screen, creating 2D patterns which will be sewn together to form a 3D garment, and simulates them to see how they drape on bodies. Elements such as the type and physical properties (thickness, elasticity, etc) of fabric or leather, buttons, zippers, colors, graphics and so on could be modified and simulated in real-time. Designers would have the ability to unleash their creativity and design intuitively and efficiently, minus the risk of analogue design and waste of fabric when prototyping, making the process cost-saving and eco-friendly. Aside from designers, patternmakers, manufacturers, and marketers would also benefit from a digitized designing process. Considering the convenience and flexibility of 3D data, communications between teams from every part of the fashion life cycle could be streamlined, creating a more efficient pipeline, reducing lead time in the process. Brands such as Adidas, Emilio Pucci, and Theory have adapted such design softwares in their pipeline; enhancing their design and prototyping processes to cater to growing consumption and diversifying demands.
Emilio Pucci utilizing CLO 3D for design and prototyping.
AI (ARTIFICIAL INTELLIGENCE) IN FASHION DESIGN
The use of AI (Artificial Intelligence) has also begun to bleed into all aspects of the fashion world, and not just another trendy term tossed around for marketing’s sake. As trends come and go, designers and brands need to keep up with the pace, when they could’ve fully concentrated on working on creative tasks. This is where AI algorithms come into play; by scouring images and hashtags through the internet, AI could analyze, detect, and perhaps even predict new and upcoming trends with high accuracy. This solves the problems of overproduction of garments. As advanced analytics show what does and what does not sell well to a brand’s specific target audience, aside from trendy items, brands and designers could use the data as a guideline to launch new collections that will fly off the shelves, instead of piling up as deadstock.
A great case study would be Amazon’s AI fashion designer, where the AI algorithms take the role of a designer in place of a human. Considering social media platforms are pretty much goldmines of fashion trends and inspiration, Amazon’s AI combs through those images to put a finger on what’s becoming on-trend, what’s more likely to sell, and what’s probably ending up on sale bins. By analyzing the batch of images and labels and spotting what’s trendy, the algorithms can then craft similar-looking clothes quickly. The designing process can be hooked with Amazon’s patented on-demand clothing manufacturing warehouse, so as soon as a customer’s order is placed, manufacturing and shipping would take place accordingly, rather than mass-manufacturing apparels that are mostly disposed of.
8 by YOOX, an AI-powered, human-designed collection.
YOOX, an Italian fashion e-commerce platform has also embraced AI with their latest endeavor, a private-labeled project known as “8 by YOOX”. In this case however, Yoox’s AI tools will work alongside human designers instead of replacing them entirely. By harvesting fashion content across social media in key markets, while concentrating on content from fashion influencers. This insight is combined with predictive indicators of emerging lifestyle and fashion trends, as well as data analysis from products sold on its site, customer feedback, purchasing trends of the industry, top trending searches and image recognition. The algorithm then produces a dynamic moodboard that reflects ongoing and emerging trends, shifting industrial trends, and the behaviour and preferences of their customers. This moodboard allows Yoox’s human designers to draw their inspiration from it, and to better understand their consumers, before proceeding to conceive collections that customers prefer to buy.
The role of AI designers seems to gradually overlap with those of human designers. The first question that comes to mind is, will consumers go for AI-made or human-made apparel?
A human-made garment encompasses his or her experience, craftsmanship, heritage, memories, history, choices, preferences, and even emotions, all of which an omnipotent AI lacks. On the other hand, an AI-made garment draws inspiration from not just a human being, but the entire humanity’s behaviour, minus the human touch.
“I find it hard to imagine that AI algorithms can produce anything avant-garde or creative on purpose,” says Natalie Nudell, a fashion historian who teaches at the Fashion Institute of Technology in New York.
Still, algorithms proved its worth is improving trend forecasting in the fashion industry. As technology progresses and machine learning gets more sophisticated, trend prediction will improve and it will fundamentally change how designs spread. Retailers could respond a lot faster to changing preferences thanks to AI.
Regardless, whether it is piles of code that is sketching out dresses for the next season, fashion still needs a touch of humanity, at least for now.
AI IN LOGISTICS & MANUFACTURING
Conventional apparel manufacturers, especially those concentrated in South East Asia, are notorious for exploiting and overworking garment workers. A typical labourer could work up to an astonishing 14-16 hours per day, 7 days a week. Many brands and manufacturers chose to turn a blind eye to these inhumane practices.
The introduction of AI will prove to be useful in these circumstances. However, that is not to say that labourers will be replaced, essentially, with a click of a button; but rather, their workloads would be vastly relieved. Labor-intensive tasks such as sewing, sorting, and so on could be managed by algorithms with a higher rate of accuracy at faster speed. Quality assurance processes could also be improved as AI could easily identify defects in fabric, as well as streamlining colors and textures, ensuring that they match the original design intent.
To ensure garment workers are not left behind by automation, manufacturers have the obligation to train workers to be digitally literate, to be able to operate and maintain the systems. "Nearly 30% of the workers have been displaced from the sweater sub-sector of the readymade garment sector because of automation. We should impart skills training to them so that they can make themselves competent for advanced jobs," said Bangladesh Garment Manufacturers and Exporters Association vice-president Faisal Samad. While it may seem like a rather uphill battle to upskill workers in time before machines and AI take over, that is not the case as line workers understand clothing construction. Fundamental knowledge of clothing construction will come in handy when workers are exposed to garment modeling in 3D. Therefore, manufacturers, fashion brands, technological solution providers should unite and collaborate on skills development, education, R&D, and work to establish a generation of digitally-proficient workers, enabling them to transition smoothly into the age of digitization.
When it comes to logistics and supply chain, machine learning technologies could be utilized to manage and optimize supplies. Given the volatile nature of the fashion industry, retailers may find it challenging to keep up with trends and stock up according to consumer preferences. Inventory levels and sales are typically rough estimations based on sales data from the previous year, however that aspect alone is neither reliable nor accurate. Unpredictable factors such as changing trends, economy, climate change, political and social climate and so on could influence sales. Companies and retailers could leverage machine learning algorithms to make more accurate predictions of inventory demand based on the aforementioned factors, and thus reduce forecasting error by as much as 50 percent. In addition to reducing errors in predictions, that also lessens the amount of clothing and wastage produced, and therefore less energy and resources are used as well.
AI can be applied to expedite and automate logistics for faster transit time, saving on shipping costs, manpower, and carbon footprint. Machine learning technologies could analyze and identify faster or alternate routes for delivery derailed by unexpected circumstances, such as construction or bad weather.
It is evident that technology may be the answer to many of the challenges and problems of the fashion industry. Algorithms could very well be a competitive advantage for businesses, assisting human employees to churn solutions after solutions. The coalition of machines and human beings make for a more efficient, cost-effective, eco-friendly workforce. Therefore digitization is something that the fashion industry should adopt and implement as soon as possible, especially in the light of current events.
These are very interesting yet exciting times we live in today, as there is a lot of interest and innovation taking place all over the world; where the norm could just be left behind in a matter of time, in favor of something more innovative and productive. And we can’t wait to see where digitization will go to truly drive change towards a future of sustainable fashion. It won’t be easy or straightforward, but we need it now more than ever.
Let’s continue the conversation in our next article, where we’ll touch on the roles of digitization in shopping and retail.
Subscribe to us here to stay in the know!
