No GUESSwork with this FashionAI store
Who says computers don’t have fashion sense? Online retail giant Alibaba recently tied up with leading fashion label GUESS to create a pop-up fashion store that relied on AI to serve customers. From clothes coordination to keeping tabs on the whereabouts of products for customers, FashionAI gave a sneak peek into the future of fashion and OMO (Online Merges Offline).
From walk-in to login
Between July 5th to 7th, Alibaba unveiled their experimental FashionAI concept store within the campus of Hong Kong Polytechnic University. It was a collaboration with American fashion brand GUESS. The two companies have been working together for the development and practical usage of AI in the fashion field.
FashionAI is effectively an AI system that takes a personalization approach to suggest fashion items to people based on their individual preferences. At the core of the system is the buying history and data of all customers who’ve used Alibaba-related services before, plus data owned by e-commerce site Taobao of which contains over 500,000 variations of fashion coordination.
Customers ‘login’ to the store by using the Taobao app on their smartphones or through facial-recognition machines at the entrance.
The store interior looked like a regular GUESS boutique but what different were the screens and various sensor devices for gathering data. RFID tags (Radio Frequency Identifier tags) were attached to all the clothes for the system to identify them, so when customers held the clothes in front of the smart mirror devices they could be recognized and the system would instantly display other items that it thought would make a good combination.
And just like shopping online, customers could pick up items at the swipe of a finger. Once inside the fitting room, you could use the installed digital interface to select the clothes you’re interested in and a staff member would promptly bring you the items one by one.
Right on track
The key bit of technology that made the FashionAI store work was the RFID tags. These were embedded with LoRa technology — a type of LPWA (Low Power, Wide Area) platform — and they made up the data transceiver system of the store. Thanks to this system, the act of selecting products became more efficient and recommendations could be more personalized. However, in truth, this system was not just for the benefit of the customers. It recorded data about what products a customer has taken off the shelf or had “shown interest” in. As such, it could even note what purchases were first considered but eventually abandoned — which in itself is important data. Basically, the system made possible the gathering of consumer behavior data on a level of detail not possible in normal “analog” stores.
Tencent and others follow close behind
Alibaba’s currently laying the groundwork to make the expansion of real FashionAI stores a reality. And rival companies are hot on their heels.
E-commerce conglomerate JD.com recently partnered with Chinese fashion brand “Southward Red Beans Finery Limited” to open an experience store in China on October 1st that makes use of their own original AI technology.
Then there is Tencent, who already from February 2017 has been using AI to research fashion trends of the younger generation born on or after 1995. Their system analyzes the 100 billion or so photos that have been uploaded by users to Tencent’s social media platforms and tries to figure out what types of clothes each user prefers and in what color. This will no doubt become useful information for future applications.
It looks like realizing a fully AI-equipped apparel store is just one step away. Such a store will no doubt require the integration of a number of different technologies. It would be able to scan not just a person’s face but other physical features as well, plus give much more detailed suggestions on how to wear particular items, as well as judge and give out much more nuanced recommendations based on a person’s preferences. Add to this the seamless merging of the online component to fully realize OMO, and we’re getting closer to the smart fashion store of the near future.
In China, there’s the rivalry between Tencent and Alibaba, of which the latter has already actualized OMO with their Hema supermarkets. On the other side of the globe, it was announced that Amazon, who’s currently developing their Amazon Go automated stores, apparently applied for a smart mirror patent at the beginning of 2018. How they plan to use the patent is still top-secret, but for the fashion and beauty industries, of where smart mirrors are already being employed, it will be worth keeping an eye on any developments. Shopping, both online and off, is about to get even more exciting.
Translation: Ching Li Tor
Original text (Japanese): Team Roboteer
