The first recorded machine used in clothing production is a simple sewing needle, from 28,000 BC. Since then machines have played an increasingly large part in the production and development of clothing and the evolution from functional garments to the multi-billion dollar global trade in fashion. Now Machine Learning (ML) and Artificial Intelligence (AI) are being utilised across the fashion spectrum, from design to manufacture and from current assistive retail technology to predicting future trends.
Machine Learning is utilised by several companies to assess current trends and predict the future. Fashion Pocket is one of those. By scanning and processing upwards of 25 million images, from both social media and retailers’ websites, they are able to monitor trends and styles. This provides dual benefits. Firstly the results enable fashion companies to more closely approximate the upcoming trends. Secondly, this facilitates more accurate production of garments, leading to a reduction in waste.
This use of ML and AI is now being assessed by researchers at Cornell and Bloomsbury Publishing, to determine to what extent this is accurate and beneficial.
Machine Learning is also being increasingly employed by retailers to provide customers with added value during their online shopping. ASOS has recently rolled out its “Fit Assistant” tool. By examining customers previous buying and returns, as well as an optional set of further questions, they aim to both ensure that the customer buys the correct size garment and minimise the returns of incorrectly fitting items.
Further value to the shopping experience lies in the use of ML to provide recommendations. Most customers are familiar with this, from shopping at larger online stores such as Amazon, however smaller businesses are now realising the benefits of accurate suggestions. Thread is an online only fashion retailer which is able via using ML and AI to provide personalised recommendations to over half a million customers, with only 10 stylists employed.
Despite the huge and rapid growth in online clothing retail, is it estimated that up to 90% of clothing sales still occur offline, in physical stores. This has led to some online only companies exploring the viability of establishing an offline presence. Here, again, both ML and AI play a part; in gathering and processing the online shopping habits and trends, what is likely to work offline is more predictable and therefore more efficiently implemented.
At the cutting edge of clothing design Machine Learning is being used as a component part of the development of smart fabrics. Myant are using ML in their Skiincore fabric, a knitted material that learns your optimal body temperature; by monitoring the temperature of both the wearer and the environment. Over time it “learns” and delivers heat tailored to the individual wearer.
Machine Learning and Artificial Intelligence are also being developed with a more socially aware end use in mind. Researchers from Google and Georgia Tech have constructed technology that allows a machine to learn how to dress itself in various different garments. This is predicted to have potential benefits, especially in the field of social and elder care.