Simbe Robotics announced this week that more than 3,000 of its autonomous units now sit under contract. The figure marks what the company calls the largest committed fleet of shelf-intelligence technology in retail. A decade after it first rolled out Tally, the pioneering inventory robot, Simbe has moved far beyond single-purpose scanning. It now sells a platform that fuses mobile robots, fixed sensors, computer vision, RFID and edge AI into a continuous stream of store-level truth.
The milestone arrives at a moment when retailers face persistent gaps between digital plans and physical reality. Shelves go empty. Prices appear wrong. Promotions fail to display. Those failures cost billions. Simbe’s data suggests its systems have scanned 18 billion price tags, caught 132 million errors, detected 607 million out-of-stocks and logged 5 million hours of autonomous operation. Such numbers don’t arrive from pilots. They come from years of grinding deployment across real stores.
Brad Bogolea co-founded Simbe and still serves as CEO. “Retail is one of the clearest proving grounds for physical AI at commercial scale, and this is just the beginning,” he said in the PR Newswire announcement. His tone mixes pride with understatement. The company operates with more than 75 retail banners across nearly a dozen countries. Grocery chains. Wholesale clubs. Farm supply outlets. Mass merchants. The list spans continents.
Early customers believed when few others did. Schnuck Markets deployed Tally 2.0 in 2017 and went chain-wide in 2021. BJ’s Wholesale Club signed on in 2022 as the first club operator. HomeBase USA expanded from pilot to every location in under three months after seeing a 92% drop in pricing errors and 58% fewer controllable out-of-stocks. B&R Stores brought the robots to Nebraska stores this summer. Kroger has begun rolling them out in Ohio and Washington state markets. The pattern repeats. Pilot success leads to broader commitment.
Store managers like what they see. More than 90% report that working alongside Tally improves their jobs. The robot handles repetitive aisle scans. People focus on exceptions and customer service. Shoppers seem unbothered too. Early surveys found 99% as likely or more likely to return after encountering the machines.
Safety mattered. Simbe became the first company to earn UL 3300 certification for a robot designed to work in public spaces. The standard covers service robots operating among people. Tally passed more than 40 tests. It detects humans, stops appropriately, navigates dynamic aisles. That certification removes a major barrier for risk-averse chains. Grocery Dive reported on the development earlier this year.
Yet robots alone no longer tell the full story. Simbe now pushes fixed sensors called Tally Spot. These units watch specific high-priority zones more frequently than a roaming robot can. Think produce. Meat cases. Alcohol. Cold medicine. Checkout lanes. Bottoms of shopping carts. Areas with fast turnover or high shrink. CEO Bogolea told Grocery Dive the sensors let retailers trigger immediate restocking when levels hit certain thresholds.
The shift reflects broader ambition. Simbe no longer positions itself as a robot vendor. It sells physical intelligence. Data from shelves feeds applications, workflows, AI agents and connected devices. That data becomes foundational for everything from merchandising decisions to supply chain adjustments to online fulfillment. In-store execution directly affects digital promises. Get the physical side wrong and the rest collapses.
Investors have backed the vision. Simbe raised $50 million in a Series C last year led by Growth Equity at Goldman Sachs Alternatives. Total funding now exceeds $100 million. The company doubled its team in 2024. It continues to hire as demand grows. Aidan Madigan-Curtis, partner at Eclipse, one of Simbe’s backers, captured the difficulty. “What works once in a lab doesn’t often work in the real world, and it’s brutally hard to build robotics that operate reliably across thousands of stores for the world’s largest retailers.”
Scale brings its own data advantage. Every new deployment adds shelf-level patterns across thousands of SKUs, dozens of formats and varied geographies. That flywheel compounds. Simbe claims its fleet-wide insights now support 10,000 brands. Manufacturers gain visibility into how their products actually appear on shelves. Placement. Pricing compliance. Promotion execution. Real numbers replace assumptions.
Competitors exist. Brain Corp has deployed tens of thousands of robots, many focused on floor cleaning. Other firms test inventory systems. Yet few match Simbe’s decade of continuous public operation or its certified safety record. Most physical AI efforts remain in limited pilots. Simbe has moved past that stage.
Recent coverage shows the momentum. Unite.AI detailed the fleet totals and history of version upgrades from Tally 1.0 through the just-announced Tally 4.0 with longer runtime and sharper vision. The AI Insider highlighted the platform’s expansion into fixed sensing and enterprise applications. Retailers such as SpartanNash, Wakefern and Giant Eagle appear repeatedly in deployment stories.
Challenges remain. Labor unions watch closely as Kroger and others expand. Some workers worry about job impacts. Retailers counter that the robots free associates for higher-value tasks. Pricing errors and out-of-stocks directly hurt sales and customer trust. Fixing them at scale matters more than ever in a competitive environment where online and physical channels blur.
Simbe plans further expansion. More automation in store operations. Deeper integration with enterprise systems. Continued growth of the fixed-sensor network. The 3,000-unit mark serves as validation, not endpoint. A decade of real-world data has produced something rare in robotics: proven reliability at enterprise scale.
Retail has always been a physical business. Digital tools overlaid charts and forecasts on top of that reality. Now a fleet of robots and sensors delivers the missing ground truth. Decisions once made on hope or weekly audits can rest on continuous, machine-readable observation. The stores of the near future won’t run without people. But they may run far better with this new layer of physical intelligence watching every aisle.