Simbe’s Tally robot autonomously navigates stores and scans shelves to find out-of-stock items, incorrect tags, misplaced items, and more. | Source: Simbe Robotics
Simbe Robotics Inc. today announced advancements to Simbe Vision, the computer vision technology behind its Tally robots. The San Francisco-based company said these enhancements allow its customers to move beyond identifying problems to implementing AI-driven actions that optimize store operations with precision and efficiency.
“The most successful retailers are leveraging real-time, AI-powered insights to drive execution excellence,” stated Brad Bogolea, co-founder and CEO of Simbe.
“What sets industry leaders apart is how they’re combining cutting-edge computer vision with Simbe’s Store Intelligence platform to identify the next best action for their teams,” he added. “Together, we are enabling faster, smarter decision-making exactly when it matters most, creating a competitive advantage in today’s challenging retail environment.”
Simbe Vision has been validated by leading global retailers including BJ’s Wholesale Club, Schnucks, Wakefern, CarrefourSA, and SpartanNash. The company said that with its technology, its customers have achieved industry-leading compliance, fewer execution errors, and transformative operational efficiency at scale.
In January, Simbe launched the Tally Spot fixed camera to expand the visibility of real-time inventory management. Simbe was a 2024 RBR50 Innovation Award winner for its scale out with BJ’s Wholesale Club.
Simbe Vision adds inventory capabilities
The company said Simbe Vision enhances its inventory technology with several new capabilities. Simbe developed these in partnership with its major retail clients. They include:
Low-stock, on-shelf detection: The system combines volumetric detections and advanced depth sensing to identify products running low before they become out-of-stock, enabling proactive replenishment and preventing lost sales. Simbe said its detection system provides 1.4x more early signals about shelf condition compared with traditional methods.
Comprehensive shelf intelligence: Simbe Vision detects product spreads, plugs, misplaced items, and missing price tags using advanced AI technology, which the company said can improve planogram accuracy and reduce inventory distortion.
Breakthrough inventory monitoring: The system adapts to dynamic retail environments by combining mobile and fixed sensors. Simbe said its advanced computer vision fuses data from autonomous mobile robots (AMRs) and fixed cameras, eliminating planogram maintenance and capturing real-world shelf conditions. Its claimed that its customers can reduce complexity, improve accuracy, and obtain insights that evolve with changing stores, no matter how often their layouts or merchandising strategy shifts.
Product recognition and similarity detection: Simbe Vision uses AI-powered visual search for precise product identification across shelves. The company said it can distinguish even nearly identical products (such as flavor or size variants) to ensure correct placement and prevent mis-merchandising.
Shelf-to-stock comparison: The system now cross-checks on-shelf inventory with backroom stock data to uncover hidden “plugs” that appear fully stocked at a glance, preventing phantom out-of-stocks and improving replenishment accuracy.
Automated shelf-tag verification: Simbe Vision now utilizes barcode scanning, optical character recognition (OCR), and proprietary machine learning to detect mismatches between price tags and products, reducing costly errors.
Now accepting session submissions!
Simbe prepares for rapid scale in 2025
Simbe Vision delivers continuous, full-store visibility with 98.7% SKU-level identification accuracy and over 99.3% shelf condition recall, the company said. Overall shelf condition precision exceeds 99%, so retailers can trust the data and recommendations they receive, it asserted.
Simbe claimed it has analyzed over 60 billion retail shelf images to date, providing real-time data on inventory levels, pricing, and product displays. The company expects to process twice as many images in 2025 as in 2024, helping retailers worldwide make quicker and more informed decisions.
“What sets Simbe apart is our ability to deliver full-store intelligence rather than limited audits,” said Jari Safi, vice president for AI at Simbe. “Our deep learning algorithms provide SKU-level accuracy across our database of more than 18 million SKUs—the largest of its kind in retail—distinguishing even the smallest product differences to ensure precise inventory tracking and pricing integrity.”
Unlike most fixed-camera systems that rely on rigid planogram data, Simbe’s Tally Spot uses real-time insights from Tally’s daily store traversals. This removes the need for store teams to follow rigid workflows and delivers more adaptive, accurate shelf intelligence.
In addition, Simbe’s computer vision system can scan 5,400 items per hour, providing real-time data to enable store teams to respond quickly to inventory issues. The system detects up to 10X more out-of-stock items compared to manual audits, it said.
“Simbe Vision doesn’t just surface problems—it recommends the next-best actions based on business impact, helping store teams prioritize what matters most and shift from reactive issue handling to proactive execution that improves labor efficiency and drives real results,” said the company.
GIPHY App Key not set. Please check settings