Back to overview

Perspectives: AI is revolutionising the fashion business

Designed by Big Data

Artificial intelligence and its uses along the textile supply chain: from intelligent algorithms, smart fabrics and chatbots to database-based designs. Despite the fact that artificial intelligence (AI) is still in its infancy, the technology is spurring a plethora of radical innovations and reinventing the fashion business.

November 2018

From designing clothing to producing, delivering, marketing and wearing it – long-established processes are being fundamentally challenged and redefined by innovative technologies. This profound transformation is fuelled by ‘artificial intelligence’ (AI). In the form of high-performance algorithms, AI uses the vast oceans of data available to us in the age of digitalisation and big data and is already offering people and fashion a huge spectrum of new uses. “We are only just beginning to get an idea of what artificial intelligence can do for the possibilities of automation. (…) In ten years’ time the algorithm will know better than our customers themselves – and they will trust it,” said  Mark Langer, CEO of Hugo Boss, at his presentation for the Club Hamburger Wirtschaftsjournalisten, the club of Hamburg’s business journalists.

AI – the concept
Artificial intelligence is a field of computer science tasked with making artificial computer programmes “intelligent”. There are three stages of artificial intelligence in total: an AI-based programme in the first stage obeys concrete specifications and implements them; AI in its second stage can gradually learn new things; while the third stage is capable of independently appropriating contexts, analysing, arguing, making concrete predictions and giving recommendations for action. In particular, the latest generation of so-called “deep learning” algorithms is able to recognise trends at an early stage – a skill that will have a decisive impact on the future of the fashion industry. So what are the specific innovations and uses of artificial intelligence along the textile supply chain?

Designed by big data
The “AI clothing designer” by Amazon is currently one of the most impressive best-practice examples of the countless other AI-based programmes by the US giant. The intelligent tool analyses global image databases to detect microtrends early on and determine the most popular outfits and trends. Based on this, new designs for fashions and the corresponding textiles are generated. The garments and collections can then be produced on demand, i.e. flexibly and upon request. In no time at all, the programme can take orders for the production of various garments (from dresses to suits) and, where required, bundle and process individual orders.
Another AI-supported programme opening up whole new opportunities in trend research and design for fashion and textile companies is offered by IBM and goes by the tongue-in-cheek name of “Watson”. Similar to the AI clothing designer by Amazon, Watson can scan global image databases and predict what the central colours, silhouettes and styles will be in the coming season. To do this, it analyses data and comments from multiple sources such as social media channels. Fashion designer Jason Grech used Watson to analyse 500,000 images from Instagram and fashion archives. Based on the strongest design trends he developed a collection of couture dresses.

Wearable emotions
New York fashion label Marchesa has also been experimenting with the AI programme Watson. On the red carpet of the 2016 Met Gala, model Karolina Kurkova wore Marchesa’s “Cognitive Dress powered by Watson” – in keeping with the gala’s theme “Manus x Machina: Fashion in an Age of Technology”. The interactive evening dress reacted to individual Tweets posted by Marchesa fans. Watson assigned certain emotions and sentiments such as “joy”, “curiosity” or “patience” to every Tweet, for which the suitable colour tones were pre-defined by Marchesa. The LEDs integrated into the dress changed their colour depending on the emotion of every single Tweet. The intelligent dress therefore became a projection surface for the words and moods of Marchesa’s loyal fans.

Sourcing 4.0
As far as fabric research is concerned, artificial intelligence is opening up whole new opportunities for the fashion business and optimising traditional workflows and processes. American IT company Centric Software, for example, has developed an AI-based product lifecycle management tool that speeds up long-established processes and brings products onto the market faster. The programme is based on a visual search engine that helps product developers and designers with the search and sourcing of suitable fabrics. Using the AI tool, interesting textiles can be scanned, synchronised with the company’s own material archive in a matter of seconds and suitable suppliers offering a corresponding design can be located. Swiss lingerie brand Triumph, for example, used the intelligent AI tool to create a digital library of thousands of lace fabric samples to speed up future design processes.

Intelligent fabrics
Another central use of artificial intelligence in the fashion and textile industry is intelligent fabrics and wearables like fitness trackers and smartwatches that measure and document the wearer’s pulse, for example. Sports apparel brand Under Armour is currently offering smart sneakers with integrated sensors that collect data on running speed, distance and step count. In the field of smart clothing, electronic components have so far been incorporated into finished garments in what has been a costly and time-consuming process. This meant that most intelligent apparel was sensitive to water and therefore unsuitable for commercial serial production.
But these days, newer textile innovations are making it possible to produce e-fibres that already contain wires. Last year, Levi’s and Google revealed their smart “Levi Commuter Trucker Jacket” to the public. The intelligent denim jacket is made with a conductive thread called Jacquard by Google and can withstand up to ten washes. Incorporated into the jacket’s sleeves are sensors that react to pressure and movement and form an electronic interface to the user’s smartphone. This means that the wearer can screen phone calls, control music or get directions with the jacket.

One year prior to that, the Massachusetts Institute of Technology developed a fibre that, in addition to wires, also includes other electronic components like high-performance chips, light-sensitive sensors and LEDs. The e-fibre can conduct light like a glass fibre and illuminate the integrated LEDs. This can be used to produce a sensory fibre that offers seemingly endless possibilities of use. If a finger is placed on the finished textile, it can record vital functions such as the wearer’s pulse in real time. Integrated sensors register even the slightest differences in the incidence of light caused by pressure fluctuations during the pulsating of the blood. In addition to all that, the smart fabric can warm up, cool down and even be used as a mobile charging station for a smartphone – a material innovation that is bringing future visions within closer reach.

Artificial intelligence is also playing an increasingly important role in the sector of home textiles: one example of this is the smart pillow by Centa Star, which provides relief to people who are having difficulty sleeping. With the help of built-in sensors, the pillow measures individual sleeping phases, the user’s movements while they sleep and the noise level of their surroundings. An intelligent programme interprets the results and, via a free app, gives the user personalised tips and recommendations on how to optimise their own sleeping patterns. The “Sleep Tracking Mat” by French electronic manufacturer Withings also promises smart sleep tracking via a health app. An intelligent textile mat is placed onto the user’s mattress to monitor their sleep cycles and also tracks their heart rate and detects snoring. The AI-based product can also be linked to other electronic devices. So just by getting into bed, it automatically turns off the lights or turns on the central heating when you wake up. Both products can be experienced in person at the next edition of Heimtextil (8-11 January 2019) in Frankfurt am Main. This January, the international trade fair for home and contract textiles will be focusing on the “Future of Sleep” and presenting everything from intelligent textiles to AI-based applications.

In the last few years, artificial intelligence has been increasingly establishing itself, particularly in fashion marketing and corporate communication.

Eloquent algorithms
In terms of customer communication, AI-based programmes like personal voice assistants and chatbots are making it easier to interact. They are capable of autonomously holding a conversation and answering questions via voice or in writing. In addition to the well-known voice assistants like Alexa by Amazon or Siri by Apple, intelligent algorithms are also being used in a targeted way in the fashion industry. Both luxury British brand Burberry and Tommy Hilfiger are using chatbots to enable their customers to contact them at any time of day or night.

Disrupting retail
Tommy Hilfiger is also using other AI-based programmes: the company’s own AI tool #TOMMYNOW has enabled the US fashion brand to minimise its production cycle from eighteen to six months and also to switch to the well-documented “see now, buy now” business model. And in addition to that, the US fashion brand is actively experimenting with augmented reality (AR) technology. AR tools like smart glasses, mirrors or smartphone apps are being used to create enhanced sensory perceptions. This makes it possible to virtually superimpose and rediscover reality with images, videos or games. At the most recent fashion show by Tommy Hilfiger in Shanghai, the audience was able to virtually try on and order new pieces from the collection directly after the show using a 3D avatar and an AR mirror. The ordered products were then sent directly to the customer’s home. As well as Tommy Hilfiger, numerous other fashion brands like Zara, Burberry and GAP are also working with AR-based concepts. At the POS in particular, this technology is offering a whole host of opportunities to emotionalise the interaction with target groups in a fun, playful way and to create unique brand experiences.

In a fashion business that is constantly gathering pace, intelligent programmes and smart processes are the key to success. They enable companies to react to changes and market trends quickly and flexibly, to better understand their own target groups and to minimise production cycles. The time has come to move into new technological dimensions, to explore the wide experimental field of AI and to play an active role in shaping the fashion and textile industry of tomorrow.

Intelligent material innovations, robot-controlled processes, AR-based tools and sourcing 4.0 – at the upcoming tradeshow duo of Techtextil and Texprocess (14-17 May 2019) in Frankfurt am Main, there will be a wide spectrum of products and applications on show that are already putting AI to good use.

Photo Credits: Courtesy of the researchers


  • Perspectives
  • Innovations