The Future is at the Mercy of Technology

Artificial Intelligence (AI) and Machine Learning (ML) have become major topics of interest in the fashion industry. Companies across the world are either already using or planning to invest in AI & ML systems. Within this decade, AI experts have speculated that AI-enabled smart factories will see a marked improvement in industrial processes. Clichéd as it sounds, smart AI is set to revolutionise the global fashion industry. Agile brands must take the leap and go smart –especially when the proverbial iron is hot as it is now.

The global fashion industry contributes to a steady 2% of the global GDP. At present, the adoption of AI & ML technologies has yet to reach critical mass. Studies show that the “global AI in fashion market size” was estimated at US$ 228 million in 2019. Clearly, there is room for growth here. Notwithstanding the early adopters, many brands and manufacturers are reluctant to take the leap. The pandemic was a wake-up call for many fashion businesses that were slow to adopt AI. In my understanding, the momentum of AI adoption will further increase in the following years.

Why Invest in AI?
In 2020, it was estimated that 44% of fashion retailers suffered from bankruptcy as a consequence of not integrating AI into their business model.   The   pandemic represented a paradigm shift in how fashion brands and manufacturers operate. The old ways were no longer as applicable as they used to be.
This precipitated an increase in investment in AI- enabled systems. As of 2022, the global spending on AI is expected to reach US$7.3 billion per year. While some may continue to harbour the same sentiment about technology and fashion, we cannot deny the reality of the situation.
Investing in AI & ML is necessary to stay relevant in a highly competitive industry. The current market is driven by the behavioural trends of the customers. Many of these trends are not known or poorly understood. As such, it becomes mandatory to have means to access information with maximum accuracy. Brands reliant on real-time data, Big Data, customer personalisation, automation and other cutting-edge technological services have shown tremendous growth in the past few years. Notably, 20% of the leading global brands are generating 144% of the industry profits!

Beneficial Applications
Let’s deep dive into a few domains where fashion brands can gain an edge through AI & ML implementation.

• INFORMATION POOL
  The most common application of AI and Machine Learning in the fashion industry is the use of chatbots. Chatbots can form a valuable data pool on customer desires and intentions. This is especially applicable for e-commerce stores. While chatbots can be fully automated, it is recommended that they be supervised by a few customer service agents. Agents can guide a customer looking for a specific product, thereby enhancing the buying experience.
  AI-enhanced information technology also helps to track customer metrics such as purchase history and other online behaviours. Such metrics can be collated into an aggregated data pool. The resulting data can help in customising brand stores and collections so as to attract more customers.
• APPAREL DESIGN
  The potential of AI in designing apparel is still nascent and untapped. Scientists are still developing new models that deploy AI for creative applications. Even so, more designers are working to realise the promising benefits of AI-enhanced fashion design. Using smart data collection methods, designers are creating collections that are better attuned to customer preferences and trends.
  One AI model to watch out for when it comes to apparel design is Generative Adversarial Networks (GANs). GANs are a type of Machine Learning technology where two adversarial models are trained simultaneously. One model is the generator (i.e., “the designer”) that learns to create images that look real. And the other is a discriminator (i.e., “the design critic”) that learns to distinguish between real and fake photos. The imaginative use of technology with GANs offers a window into the future of computer-generated designs that look aesthetically pleasing.
• VIRTUAL MERCHANDISING
  A decade ago, the e-commerce business model relied upon customers looking at static pictures (or video sometimes) of an item and buying it. Within this decade, Augmented Reality (AR) and Virtual Reality (VR) systems are expected to further revolutionise the e-commerce space. Virtual experiences for consumers will become more immersive and realistic, thereby blurring the lines between physical and online stores. AI systems have revolutionised business operations. They boost efficiency and productivity across the entire operation. This applies to the production process itself as well as the supply chains beyond.
• REDUCING TEXTILE WASTE
  AI & ML based technologies can make a fashion brand smarter as well as greener. Smart factories can deploy sustainable systems by reducing textile waste. For instance, with B2B transactions, samples of garments can be digitally created and shared for consultation. This bypasses the need to create batch after batch of fabric swatches.
  Digital samples also take less time and can be iterated upon as required.
  AI enhancements can also help on the consumer end of things. For example, customers may make an erroneous purchase. This leads to a sheer waste of the product as well
  as time and energy. AR and VR systems can help give a customer a clearer picture of the product they’re buying. This in turn can reduce customer returns and thus reduce product waste.

AI systems have revolutionised business operations. They boost efficiency and productivity across the entire operation. This applies to the production process itself as well as the supply chains beyond.

Are AI & ML Cost-effective for Brands?
In a few words: yes, they are. AI systems have revolutionised business operations. They boost efficiency and productivity across the entire operation. This applies to the production process itself as well as the supply chains beyond.
The benefits of integrating AI are applicable even to businesses that are cautious operators. In 2020, Gartner Inc. conducted a survey of 200 businesses that invested in AI technologies. The survey found that 66% of them did not reduce their investments in AI. In fact, 33% of businesses even reported an increase in AI & ML technology investments. One of the key factors behind these figures was reduced operational costs. AI-enabled brands were able to integrate a range of operations into a controlled unit with maximum efficiency. This helped to reduce business costs as well as increase revenue. What’s more: an AI-enabled fashion brand can expect a quick return on investing in AI systems.

For best results, we recommend focusing on Machine Learning Operations (MLOPs). The data-centric nature of MLOPs ensures a quick ROI while also realising the full potential of AI technologies. An MLOP-style AI system can either be built in- house for a brand or outsourced from a third-party vendor.
In the case of in-house development, a brand must hire data scientists and engineers. Further costs are incurred from maintenance and repair of the hardware units. Needless to say, in-house development can involve significant initial investment. In the short term, this poses a serious obstacle to AI adoption. A survey by Rackspace technology reveals as much: 26% of the companies assessed expressed concerns about high implementation costs. Other brands also stated that they found AI & ML based technologies to be more of a cost-cutting tool than cost- effective.

To sum up, building an AI system in-house may not be the right option for many fashion brands. Outsourcing the same to a third party, on the other hand, thus may be both time and cost effective.

AI and Trend Forecasting
Changing trends and demands is a core concern for fashion retailers. While demand volatility is an issue, and sometimes unmanageable, AI has a simple solution to it – Demand Forecasting. Trend identification and customer demand prediction using Machine Learning algorithms can now be achieved with an impressive degree of accuracy. These algorithms are automated and are able to recognise patterns, identify and compile large datasets, and capture signals for demand fluctuation. Fashion brands can use information from various sources (for example social media) to anticipate trends, optimise sales and predict customer demand.Changing trends and demands is a core concern for fashion retailers. While demand volatility is an issue, and sometimes unmanageable, AI has a simple solution to it – Demand Forecasting. Trend identification and customer demand prediction using Machine Learning algorithms can now be achieved with an impressive degree of accuracy. These algorithms are automated and are able to recognise patterns, identify and compile large datasets, and capture signals for demand fluctuation. Fashion brands can use information from various sources (for example social media) to anticipate trends, optimise sales and predict customer demand.
Heuritech, a France-based company, offers such trend forecasting solution that helps brands to anticipate customers’ expectations. Heuritech has created an in-house deep-learning technique that detects the ‘early signals’ and indicators like slight shifts in the activity among edgy influencers. Heuritech helps to convert real-world photographs posted on social media into relevant information. The companies can therefore anticipate demand and trends more precisely, manufacture more sustainably and gain exceptional competitive advantage, all by leveraging the powerful AI.

AI and Textile Industry
We do have groundbreaking technologies today that are reinventing the way the traditional textile industry operates. The benefits of having an automated process of inspecting and designing textiles are directly associated with the profit margin and time effectiveness of a brand. Cognex Corp. (a Boston-based software manufacturer) developed Cognex ViDi in 1981 that can automatically inspect fabric patterns. This AI-enabled fabric inspection helps in reducing pattern defects with maximum precision and minimum labour.

 Trend identification and customer demand prediction using Machine Learning algorithms can now be achieved with an impressive degree of accuracy.

Another example of AI-aided textile designing is Datacolor, founded in Lucerne, Switzerland. It is a colour tolerancing system where the intelligence ensures that the original design colours match the colours in a finished textile product.
AI can be used to measure and eliminate wrinkles and defects in a fabric. AATCC (American Association of Textile Chemists and Colourists) methods are commonly used in measuring fabric wrinkle performance. However, the process requires mechanical reformation to cut costs and time.
Trend identification and customer demand prediction using Machine Learning algorithms can now be achieved with an impressive degree of accuracy.

The Present Scenario
“Companies must use Big Data analytics to redefine the traditional core elements of retailing – Consumers, Merchandise and Stores – and the relationships amongst those elements to upgrade current formats and create new retail occasions,” was a remark made by Daniel Zhang, CEO of Alibaba. The company is already ahead in making big strides in AI-aided fashion endeavours and experiences. It has invested US$15 billion in its R&D labs as a push to dominate as an AI leader. Alibaba has already launched AI-powered smart stores with intelligent mirrors, garment tags, and Bluetooth chips embedded within every product. There are other brands like H&M, Tommy Hilfiger and Amazon that are adopting AI technologies to refine their service and personalisation.
Tommy Hilfiger has founded a project called ‘Reimagine Retail’ that teaches the fashion designers to use AI for designing. Macy’s came up with AI-powered shopping assistant with the objective of improving a customer’s in-store shopping experience. They use NLP (Natural Language Processing) for their ‘On-Call’ tool that is able to answer the customer’s queries and help them navigate through the store.
Amazon, by far, has bagged most of the credits for revolutionising e-commerce. By using AI-enabled fashion designer algorithm, they can design apparel by copying the design styles of the trending clothes. Amazon is also using Echo Look, an AI tool that can analyse a customer’s style and make fashion recommendations.

AI and Fashion in India

The Indian e-commerce market has expanded by leaps and bounds over the past few years, and was estimated to have reached US$84 billion in the last year (as per the RAI report).
The domestic fashion sector is facing some revolutionising changes, mostly at the hands of young entrepreneurs who have integrated AI with their fashion business or start-ups.
Former graphic designer Meghna Saraogi launched StyleDotMe (seed funded by Indian Angel Network) in 2016. The App offers instant fashion advice after collecting data through instant votes and polling options. As a next step, the Delhi-based start-up launched MirrAr at the Bridal Asia exhibition in August 2018. This platform enables customers to virtually try on accessories to feel the item and make the right purchasing decision. Later, they partnered with Tanishq. Now they have PC Jewellers, Farah Khan Jewellery, Amrapali and several small retail jewellers across India as their clients.
BigThinx, founded by Shivangi Desai and Chadralika Hazarika in Bengaluru, is another such game changer. Their flagship product, Lyflike App can create 3D avatars that look exactly the same as the user. One can recreate the self using different outfits from a virtual closet or even design outfits by picking from the libraries of different garments, cuts, patterns, graphics and fabrics. Once the user avatar is all dressed up, the App can direct to the online stores that may have similar outfits.
Tryb4uBUY from PurpleApple Infosystems focuses on AR and VR aspects of shopping. It benefits the offline shopping experiences where customers can try out the outfits and accessories on big screens before making the purchase.

In Conclusion
AI & ML are game-changing technologies that can disrupt and reinvent the operations of fashion retail. The biggest players in the industry concur that the use of AI can facilitate sustainable business practices by eliminating wastes and trial & error costs. We can claim that we are at a stage where the fashion industry has already realised that the future of retail is at the mercy of Artificial Intelligence. There is no turning back!