Computer Vision in Retail
AI in the retail industry is becoming more profound as retailers turn to computer vision technology to optimize processes. Using AI in retail stores can help identify customer movements, purchasing patterns, and busy areas. Retailers can also apply computer vision to improve merchandizing, store layout, and staff positioning. Computer vision in retail can also help to improve customers’ in-store experience, which increased satisfaction and the likelihood of them returning.
Current Trends
Key trends for online and brick-and-mortar retailers currently include improving sustainability and circularity to reduce waste, developing eco-friendly practices, and minimizing environmental impact. There are also technologies in use such as augmented reality and virtual reality which transform how customers interact with products and brands, enhancing virtual and in-store experiences. And other important trends in retail include the use of artificial intelligence (AI) and automation to optimize retail operations, particularly in demand forecasting. AI is also used for, inventory management, as well as and delivering personalized customer services.
Here we explore three important computer vision retail applications.
Cashierless Checkout with Computer Vision
Checkout technology hasn’t changed much in the past two decades but today computer vision technology can be used in retail to fully automate the checkout process to and therefore improve the customer experience. A store’s camera system captures images of each selected product, and an algorithm is trained to identify the object and its price. The system also associates the checkout transaction with a specific shopper in order to charge the correct person the correct amount when they leave. The cashier-less checkout technology helps prevent theft by observing customer behavior and analyzing patterns. An alert will be triggered when human intervention is required for risk assessment.
An AI use case example in the retail industry is Grocery retailer Aldi, who claims “Smart shopping is the future,” and piloted its Shop&Go checkout-free experience in selected stores. Another example is the many ‘just-walk-out’ Amazon Go and Amazon Fresh convenience stores that have opened up over the past 7 years. And as more retailers proactively invest in technology to fight the rise in retail crime, Amazon is also selling its solution to other companies. Meanwhile Australian software startup Tiliter promises an AI-powered solution to protect retailer profits and deliver accurate product recognition with a 5x faster shopping experience.
Inventory Management with Computer Vision
On-shelf availability is critical to brick-and-mortar retailers. If shoppers can’t find what they’re looking for on display, they’ll probably select a substitute, but as many as two out of three could leave empty-handed and may buy from a competitor instead. Computer vision technology integrated with the store's inventory management system can monitor shelves to keep track of inventory levels, detect when items are running out, and issue replenishment alerts. This AI-based inventory management system is even capable of triggering automated reordering and restocking processes, facilitating the whole process for those responsible.
An AI-powered solution can automatically update brand packaging when it changes by working with synthetic visual data (information that’s artificially generated, not produced by real-world events), referencing its own training images and any new product images.
The Californian robotics startup Simbe has developed an innovative autonomous inventory robot featuring inbuilt 3D cameras, enabled by computer vision, that navigates and analyzes the retail store environment. The AI-based robot carries out its own inventory management by detecting low inventory levels and providing alerts when a shelf needs replenishing.
With its ReShelf solution, French retailer Auchan turns everyday shelf images captured by ceiling-mounted cameras into real-time, actionable insights. In-store staff are notified when a product runs low, helping to maintain stock levels. The system can also provide out-of-stock analysis of the entire product supply chain.
Category Management with Computer Vision
Retailers create in-store merchandise display schematics or planograms to achieve the required outcomes – more sales, improved category profitability, positive customer experiences, and more. Compliance with these plans means the right product with the right price tag has the right amount of space in the right location and is optimally promoted.
Computer vision in retail helps to understand shopper behavior patterns and this real-time visual data enables the management team to optimize in-store floor layouts and workflow to optimize customer traffic. This computer vision solution for retail also provides sales data for reports that strengthen the negotiating position with suppliers.
Headquartered in Switzerland, Viso offers vision-based footfall analysis, pass-by traffic, and other customer analytics to detect in-store movements and visual data patterns so retailers can clearly see which promotions engage customers and which do not.
FAQs about Computer Vision in Retail
Why are retailers adopting computer vision technology?
Retailers use computer vision to identify customer movements, purchasing patterns, busy store areas, improve store layout, staff positioning, and enhance the overall in-store experience.
How does computer vision address long lines in retail stores?
Computer vision technology is able to make cashier-less checkouts a reality by automating the checkout process, reducing wait times and preventing an estimated $19 billion loss in sales due to abandoned purchases.
How does a cashier-less checkout work?
Cameras capture images of selected products, and algorithms identify the object and its price. The system connects the transaction to a shopper, charges them the correct amount, and helps prevent theft by analyzing customer behavior.
How does computer vision assist in inventory management?
Computer vision monitors shelves to track inventory levels, detect when items are running low, issue replenishment alerts, and trigger automated reordering processes. AI-based inventory management can also update brand packaging using synthetic visual data.
How does computer vision support category management?
Computer vision analyzes shopper behavior patterns, providing real-time data to optimize in-store layouts and workflows. It also provides sales data that strengthens negotiations with suppliers.
