How Artificial Intelligence Is Accelerating Innovation In Retail

A conversation with Cloudpick’s Jeff Feng

Since the widespread adoption of computers and the subsequent exponential growth of understanding around modern technology in the 1970s, artificial intelligence, or AI, has been identified as a major foundation in ushering nearly every industry into the future. Evidence of this is easy to find, from Amazon Go, which provides customers with a cashier-less shopping experience, to the retrofitting of essential infrastructure like hospitals with the Internet of Things, or IoT, sensors to ensure vital tools are in good working order. While AI in retail has long been considered only applicable for large corporations with deep pockets, companies such as Cloudpick, a startup that provides smart retail and AI solutions, are innovating to help both small and large offline retailers move into the future.

Retail IT Insights recently sat down for an interview with Cloudpick’s CEO, Jeff Feng, to discuss where AI in retail currently stands, dive into the intricacies, applications, and possibilities of this technology, as well as talk about what the retail landscape might look like after its transformation.

Where do AI and machine learning in retail stand today? How far have we progressed from earlier AI applications?

Jeff: “Right now, AI in retail has numerous applications and can be utilized to provide solutions for customer experience and service, marketing, and nearly every function within the retail space. Most of these functions are about providing insights to create different retail strategies - marketing campaigns and similar functions - that are more effective, but the potential for AI and machine learning goes much further than just insights. It’s progressed to the point now where traditional brick and mortar stores of any size can become highly digitized, and therefore cashier-less, which has massive potential for the future of the retail industry.”

How do these highly digitized spaces operate?

Jeff: “At a high level, this technology creates retail environments that are partially automated, with AI technology tracking customers' movement and product selection. This empowers customers to purchase products quickly and conveniently without being checked out by a cashier, nor by using a self-checkout station. The technology also reduces costs for the retail business, who no longer have to employ so many people, while also drastically reducing the chance and cost, of theft.”

Can a previously brick-and-mortar store be effectively converted into a smart store? What exactly does that look like?

Jeff: “Through AI, IoT, and Big Data, previously concrete stores can be converted into ‘smart stores’ via 3D reconstruction, an AI technology like that used in self-driving vehicles, but much more complex because LiDAR sensors - light detection and radar sensors - aren’t used.

Through 3D reconstruction, the physical store, and the people within it are converted into a digital version where everything, including shelves, the products on those shelves, staff members, and customers become a unique part of the digital landscape. This enables the AI to observe which customer is interacting with a certain product, or whether they favor a particular product category, providing specific insights on customer behavior that can be used to personalize their experience, marketing, and the way the business communicates with them. Of course, this digitization also means that customers don’t have to go through the typical checkout and cashier experience, because the AI can track and bill them for the items they have selected without any interaction.”

What catalysts do you think could accelerate the transition to these types of smart stores?

Jeff: “I believe we are currently experiencing a major catalyst which has the potential to accelerate the adoption of this technology: the COVID-19 pandemic. Thanks to social distancing mandates and forced closures of non-essential businesses, many retailers of all sizes are looking for ways to future-proof their businesses against similar types of global events. Smart stores can do this by not only eliminating interpersonal contact between cashiers and customers at PoS stations, but also by tracking the number of customers within a retail space, their vicinity to one another, and other actions within the store which can potentially alert an attendant to high-traffic areas that require additional sanitization.

Another catalyst could potentially be the cost of human labor. It’s growing day-by-day and fewer people want to occupy these jobs as evidenced by recent news about labor shortages. Technology like ours reduces the need for so many employees to keep a retail space functional, which is a major draw for any business.”

How do you see AI and machine learning technologies progressing in retail? Will it expand to create truly, 100 percent automated stores?

Jeff: “I think the next step of this progression won’t be so drastic as creating an entirely automated store, although that could happen one day. Instead, the ultimate next step in the progression of this technology will be around scaling, to encompass stores of any size. If you think about retail currently, we have supermarkets, minimarts, convenience stores, micro markets, and more. This technology, as it stands, is more suited to smaller stores, so I think the next step is creating the same type of technology for different scenarios. In fact, our team has been looking at possible innovations within the supermarket sector, and we’re currently working toward these solutions - which I’m hoping we’ll be able to release relatively soon.”

Could you give us a quick rundown on Cloudpick’s history and journey into retail? How is Cloudpick helping accelerate the transition to these smart stores?

Jeff: “In 2017, our founders - we all went to school together and have similar backgrounds in technology, retail, AI, and machine learning - noticed that many eCommerce companies were facing challenges with growth in the online space and were looking to add offline spaces to diversify their offerings. We took this idea and ran with it, developing and later launching our first wearable Cloudpick device in November 2017 to help empower and simplify offline retail. From there, we partnered with a range of different clients and organizations, including Intel, which helped us expand to many countries and regions while further developing our technology. Now, I’m sure we have become the leading company in this specific space.

As for how we’re accelerating the transition; Not only are we providing the technology, but we also partner with C-level retail executives in a range of markets to ensure their transition in becoming a smart store is smooth and considered, helping them to launch through training to shift their processes so they can effectively operate these new types of stores.”