Walmart Deployed Four In-Store Technologies. Here's What Worked and What Didn't.

Quick Facts

Walmart deployed autonomous floor scrubbers across more than 1,800 stores, powered by Brain Corp.’s BrainOS platform.
The retailer cut ties with shelf-scanning robot vendor Bossa Nova Robotics in 2020, concluding human workers delivered comparable results.
Walmart removed Pickup Towers from more than 1,500 stores as curbside pickup replaced the in-store kiosk model.

Walmart serves roughly 270 million customers each week across more than 10,750 stores. Keeping those stores stocked, clean, and operationally efficient at that scale requires more than headcount. Over the past several years, the company has tested and deployed four major in-store technologies, with outcomes ranging from full rollout to quiet discontinuation.

Autonomous Floor Scrubbers

The Auto-C floor scrubber, built on Brain Corp.’s BrainOS platform, was one of Walmart’s earliest bets on store automation. A human associate teaches the machine the store layout once. After that, the scrubber navigates autonomously using cameras, sensors, and LiDAR.

Walmart started with roughly 300 stores before expanding the program to 1,500 additional locations. The machines also carry a scanning tower that monitors shelves for inventory gaps and misplaced products while cleaning. Brain Corp.’s software now runs robots across Walmart, Sam’s Club, and Kroger stores nationwide.

John Crecelius, Walmart’s senior VP of central operations, said the technology shifts associates away from riding a cleaning machine for hours toward working alongside robotics on higher-value tasks.

Shelf-Scanning Robots

Walmart began testing Auto-S shelf-scanning robots in Arkansas, Pennsylvania, and California in 2017. The roughly two-foot-tall machines used cameras to scan aisles for out-of-stock items, pricing errors, and mislabeled products. Walmart claimed the robots scanned shelves three times faster than humans and with 50 percent greater accuracy.

By 2020, Walmart had planned to bring the technology to 1,000 total stores through a partnership with Bossa Nova Robotics and NCR Corporation. The expansion never fully materialized. Walmart ended its relationship with Bossa Nova Robotics later that same year after determining human workers could achieve similar results.

The reversal was a notable setback for the shelf-scanning category, signaling that automation does not always outperform human labor once total cost and operational complexity are factored in.

FAST Unloaders

The FAST Unloader addressed a backroom problem that had gone largely unsolved for 50 years. Traditionally, unloading trucks was a manual, time-consuming process. FAST automated the scanning and sorting of items coming off trucks, organizing them by department and priority so associates could move product to the sales floor faster.

The system was designed to work alongside the Auto-S shelf-scanner. At its peak expansion, Walmart planned to deploy 1,200 FAST Unloaders across its store network alongside 1,500 Auto-C robots, 300 Auto-S units, and 900 Pickup Towers.

Crecelius described the technology as freeing associates from tasks that are, in his words, repeatable, predictable, and manual.

Pickup Towers

Pickup Towers launched in nearly 200 stores in 2017 as a direct counter to Amazon’s growing e-commerce dominance. The tall, vending machine-style kiosks dispensed online orders in-store. Customers received a barcode by email, scanned it at the tower, and retrieved their item.

The towers expanded to more than 1,500 Walmart locations before the company reversed course. Walmart removed the towers as curbside pickup grew in popularity, particularly following changes in consumer behavior during the pandemic. The physical footprint and operational cost of the towers became harder to justify as customers shifted toward drive-up fulfillment models.

What the Tech Decisions Signal

Walmart’s track record with these four technologies shows a retailer willing to test at scale but also willing to cut what does not perform. Two programs, the floor scrubbers and FAST Unloaders, remained part of store operations. Two others, the shelf-scanning robots and Pickup Towers, were discontinued after real-world results did not match initial expectations.

For operators building their own automation strategies, Walmart’s experience offers a clear lesson: deployment scale does not equal success, and the willingness to walk away from sunk costs matters as much as the initial investment decision.