Imagine shopping in a grocery store of the future. Upon entering, scan yourself in with you palm. Moving about the store, you place your items in a smart cart that tabulates your bill, and you can leave the store without spending a single second in a check-out line.
Sound like a utopian pipe dream? To most shoppers, it still does, even though Amazon, Walmart and other retail giants have begun making big investments in the technology needed to power such smart stores. What will it take to make them a reality?
Retailers need smarter AI, and the artificial intelligence technology required to tackle these challenges must be more accessible to retailers large and small. Otherwise, cashierless stores could never truly go mainstream.
Here’s where smart retail stands in 2021 and how retailers can take smart stores to the next level.
Current Use Cases
Amazon Go currently reigns as the king of smart retail with its “just walk out” technology that relies upon in-store cameras to track shoppers’ items and bill their Amazon accounts accordingly. But with fewer than 30 active Amazon Go stores in the U.S., the experience they offer remains a rare one, and given the expensive investments required to operate such stores, scalability remains a big question.
That’s because of the hefty price tag associated. For these smart cameras to work in real time, they require tons of processing power and the algorithms must be super accurate. To save costs, enable scale, and operate efficiently, retailers are often forced to choose between performance — how fast the camera can identify merchandise — and accuracy, which is not an ideal tradeoff when most retailers already operate on very thin margins.
It’s not just smarter cameras retailers must consider here. Some retailers, like Choice Market, a Denver-based chain, enable shoppers to check in and out via a mobile app, and even allow customers to customize their shopping experience according to dietary needs or specific recipes.
[Related: Retaining the New Consumer Through the Power of Technology]
Other stores, while not fully automating the shopping experience, are using AI-based heat maps to better understand customer behavior and target promotions, product placement, and customer communications in so-called hot zones. All of these technologies require hardware to efficiently process massive amounts of data where latency is a no-go; otherwise shoppers may choose to shop elsewhere.
Considering these complexities, the financial barrier to entry has been raised. It can only be lowered by finding better ways to scale the AI algorithms responsible for these tasks so that they can run on more affordable hardware, without compromising on either performance or accuracy.
The Tech to Make It Happen
At present, many stores are relying on cloud infrastructure to process data — and while it makes scale easier to achieve, this approach can entail high cloud costs, delays in processing times as data is transmitted to the cloud, and potential privacy issues surrounding the security of data in transit.
In-store servers that process data at the edge address these concerns, but retailers themselves must purchase the necessary hardware and update the associated infrastructure to incorporate more demanding, complex deep learning models and other features.
While it’s all too easy to get bogged down in debates over whether cloud or edge processing is a superior solution, those discussions miss the point.
What retailers ought to focus on instead is leveraging their existing hardware, rather than sinking massive sums into new infrastructure when they want to innovate. Optimizing software can help save precious time and costs while delivering the same benefits retailers are chasing when they make big hardware investments. And this is where truly efficient AI algorithms could make a key difference.
Skilled data scientists will be the linchpin of this smarter approach as they bring to bear a deep understanding of the nuances involved to resolve algorithmic issues and fine-tune retailers’ AI systems.