Artificial intelligence is set to transform business, entire industries and even our everyday lives. As its impact reaches the fashion business, is the industry ready to incorporate machine learning into the design process, or is it still a frontier too far?

(Photo: Pixabay)

The fashion industry is changing, and technology is said to the biggest driver of recent shifts. According to an in-depth report on the global fashion industry conducted by McKinsey, the three keywords for 2018 were: Change, Digital and Fast (source: BOF-McKinsey State of Fashion Survey). Contemporary culture’s ‘need for speed’ and evolving consumer behaviour forces the industry to adapt. Consumers expect it all: convenience, quality, newness and affordability. The need for fast fashion has gotten even bigger in the age of social media and big data, which seems to reflect evolving consumer behaviour and the strength of social media in dictating demand.

Know what your customer wants before they do

The design process is typically time-consuming and complicated. Given that the environment is constantly changing, brands need to consistently keep up with the most current trends and predict consumer preferences for the next season. For a designer it is even harder to predict consumer preferences, as they have a much wider source of inspiration compared to ten years ago. Social media and influencers, as well as social and environmental factors, play a more important role in the decision making. This increasingly diverse pool of inspiration means consumers, as opposed to designers, now more often establish trends themselves. In an era of fast-changing preference the retailer should be able to respond to shifting demand, and tailor design accordingly – an almost impossible task for the designer to do on their own.

(Photo: Pixabay)

Digital technologies have already had a positive impact on the job of designers. However, promising results from machine learning in other processes, such as the supply chain, raise interesting questions about what artificial intelligence can offer to the designer. To what extent has artificial intelligence already moved into the design process and how much could it actually contribute?

New technologies: from creating mood boards to deploying algorithms

Recently developed machine learning technologies have the ability to produce and combine data from thousands of images and videos. This data can be used to detect trends in such an objective and precise manner that would not be humanly possible for a designer. Besides trends, also more specific characteristics such as colour, shape, length and pattern can be extracted from the data. This data-output can be used as a source of inspiration, but it is possible to go one step further and even propose outfit designs.

Machine learning in the design process is different from conventional technologies, since it provides insights in trends and consumer preferences but is also able act on it. However, as much as all of this sounds very promising, it is important to understand the limitations and usefulness of such technologies, as it is still in its infancy.

(Photo: Pixabay)

New technologies developed: algorithms pushing mood boards aside

Slowly but surely, some fashion companies are realising the necessity of machine learning in the design process, either developing tools in-house or partnering with technology companies. Together with IBM and FIT (The Fashion Institute of Technology), Tommy Hilfiger launched the ‘Reimagine Retail’ project. By using an artificial intelligence tool and an archive of Tommy Hilfiger runway and product images, FIT students created multiple forward-looking designs, which included patterns, colours and styles entirely generated by AI. It illustrated how AI can identify upcoming trends quicker and improve the design process by bringing informed inspiration to the designs.

With this additional resource of inspiration, the designers created new and inspirational products that they would not have created on their own. Earlier this year IBM and FIT announced the development of an AI application which could be used within multiple companies in the fashion industry. This proved their belief that machine learning contributes substantively to the design process.

Another more accessible example of generating the perfect blend between human creativity and technology is a custom-built app called ‘Fashion Flair’. With this technology, Huawei and ANNAKIKI’s designer Anna Yang just launched their first collection co-created by man and machine. To create the ‘Fashion Flair’ collection, the app was trained by inputting iconic fashion images along with a specific set of images of previous collections from the designer. By applying a number of filters in the app, a proposal for an initial outfit was created, which was then developed into a finished garment by the designer. Once the images are processed, the app has the ability to create infinite outfit proposals, presenting designers with endless creative possibilities. All the designer needs to do is select a design and provide the finishing touch.

Fashion Flair collection presented by Huawei and Anna Yang (Photo: Huawei)

Bringing informed inspiration

As the above examples show, there are already machine learning technologies that bring informed inspiration to designs and save valuable time at the beginning of the process. This therefore allows more focus on higher-value decision making at the crucial point where the signature of the designer is the most important. We can also learn from these examples that, while it may change the nature of the work of a designer, it won’t necessarily mean that the need for designers will disappear. Instead, it can be an additional resource that helps to connect inspiration to consumers’ needs.

My personal belief is that artificial intelligence is fast becoming one of the most promising forms of digital disruption, also in the creative processes. Early adopters will see real benefits over those who rely solely on their conventional technologies. It’s clear that the future is in data and analytics, and designers who are able to act upon this can ultimately better match the market’s needs with their designs.

About the Author

Roosmarijn de Rooij is a strategy consultant in the fashion & lifestyle industry and focusses on process optimisation. She started off her career in finance and then worked in leading positions at a department store for several years before starting her own business. She is very passionate about restructuring brands and retailers to get them ready for the future. Read Roosmarijn’s posts here or connect with her on LinkedIn.