Smart Robots Bring Inventory Skills, New Value to Retail

Stores tap IoT technology to improve operating efficiency. Robots harness artificial intelligence and computer vision to reduce inventory errors, optimize supply chain management, and free human workers for customer care.

 

Article Key

Image credit: Simbe Robotics

Retail robots are moving from the stock room onto the shop floor, and they are taking over mundane tasks with aplomb. While robots haven’t achieved ubiquity in stores, they are becoming more popular. Advanced artificial intelligence, machine learning, and analytic capabilities are driving improvements in robot technology and delivering more useful autonomous robots to employers and customers.

Employees, once worried technology would make them obsolete, are finding AI can improve their workday and make them more efficient. According to one study, 50 percent of workers use some form of AI at work, and two-thirds of those people are happy to be working alongside a robot. The same study found that 57 percent of U.S. workers trust a robot more than their managers, and globally 82 percent of people think robots can do things better than their managers, including maintaining work schedules, problem solving, and managing a budget.

Lowe Tech

Companies might not be ready to let a robot take over employee scheduling, but they are rolling out robots to scan for inventory and to interact with customers. The interest in using more automation is another side effect of COVID-19, as retailers look for new ways to improve the shopping experience and reduce human-to-human contact. 

Several Lowe’s Home Improvement stores in California have introduced LoweBot. The autonomous retail service robot, developed in conjunction with Lowe’s Innovation Labs, is based on the NAVii, designed by Fellow AI, formerly known as Fellow Robots. LoweBot uses artificial intelligence and 3D mapping software to roll through the store aisles, checking inventory discrepancies between actual shelf images and preset ones. It also recognizes mispriced and misplaced products. LoweBot notifies employees, via mobile device, which items need to be restocked or reshelved and where they are located.

In addition to inventory tasks, LoweBot can display special offers, find products, communicate in multiple languages, and help customers navigate the store. It responds to queries—verbal or those entered on a built-in touchscreen—and can direct or lead customers to the location of the products they seek. If a product is out of stock, the robot can send a text to the customer when it’s available, hopefully reducing lost sales.

Tally, Whoa!

Sporting goods retailer Decathlon and Schnucks grocery stores have added Tally robots to their employee rosters. Created by Simbe Robotics, Tally stands slightly taller than 5 feet and sweeps across the store aisles to monitor inventory levels and product pricing. The autonomous robot first maps the store floor, then uses AI and machine learning to improve store efficiency.

More than a dozen cameras and 40 sensors are integrated in Tally, which uses computer vision and machine learning to identify the location, presentation, price, and availability of products. Simbe Robotics uses the Intel® RealSense™ camera in its Tally robot. TallyVision uses barcodes, optical character recognition technology, and proprietary algorithms to analyze shelves from multiple angles. According to Simbe, manual inventory checks typically occur once a week and have a 65 percent accuracy rate; Tally checks inventory three times a day, resulting in 97 percent accuracy.

Tally’s captured data is encrypted and sent to the cloud using the store’s WiFi. The information lands in the Simbe Intelligence software where data is analyzed, action items prioritized, and alerts sent to employees.

Image credit: Simbe Robotics

The software can tell warehouse employees what products to restock, but its value goes beyond counting and shelving. With machine learning, Simbe Intelligence can recommend business strategies to increase sales. For example, the software can suggest relocating slow-moving items to hot-spot shelves to push more product out the door. It can identify which products sell out frequently and recommend restocking quantities—or contact delivery services directly with restock orders. Stores can even use Tally to develop displays or planograms—product placement on shelves—for maximum purchase potential.

Automation has enabled traditional brick-and-mortar stores to boost their e-commerce offerings. Without AI and robotic scanning, the inventory accuracy would be insufficient to offer store-specific online sales. In addition, Forbes suggests that using robots will improve supply chain management enough to recover as much as $1 trillion in lost inventory in the U.S. alone.