California store intelligence firm Simbe has announced Tally 4.0, the latest version of its shelf-scanning robot and the flagship data capture method of its store intelligence platform.

Built on a decade of partnership with retailers and technology organisations, Tally 4.0 introduces advancements in runtime, vision, sensing and edge AI, turning every shelf into a real-time source of truth.

With up to 12 hours of runtime, high-resolution and specialty cameras, expanded 3D and 360˚ coverage, and the full-stack Nvidia AI infrastructure platform, Tally 4.0 captures more of the store, more often, and delivers insights into what’s in stock, how it’s priced, and where it’s placed faster than before. It retains the same shopper-friendly form factor.

“Tally 4.0 represents what ten years of collaboration with the world’s best retailers makes possible,” said Jeff Gee, Simbe chief design officer. “While the robot is faster, sharper and more capable, its design has stood the test of time. Tally 4.0 stays true to the principle that has guided us since day one: technology should serve people.”

The retail operating model has changed. AI and automation are now essential to how the physical store operates, and Simbe has pioneered this shift for a decade. From Tally 1.0’s debut as the world’s first autonomous shelf-scanning robot, to the additions of RFID and Tally Spot (www.simberobotics.com/store-intelligence/tally-spot), through to Tally 4.0, Simbe has evolved alongside its clients and the ever-changing retail environment. Today, Simbe offers an AI-driven, physically multimodal platform capable of operating in any environment from regional grocer to big box store to national hardware chain.

With Tally 4.0, Simbe delivers the foundational data layer for the physical store, connecting shelf conditions to the decisions that shape initial use cases including on-shelf availability, price and promotion accuracy, and item location precision, while elevating store staff and shopper experiences with more mature applications for planogram compliance, forecasting, replenishment, omnichannel fulfilment and merchandising.

“The future of retail depends on closing the gap between digital decision-making and physical execution,” said Brad Bogolea, CEO of Simbe (www.simberobotics.com). “With Tally 4 we’re delivering the next foundation of shelf-level data infrastructure that connects the two, giving retailers a trusted source of ground truth to power AI-driven operations at enterprise scale.”

Tally 4.0 (www.simberobotics.com/store-intelligence/tally4) will be available to Simbe customers starting mid-2026. It has a more advanced optical and sensing system. Enhancements include:

• Extended runtime: Up to 12 hours of operation with shorter, faster charging cycles supporting full-day and overnight coverage.
• Sharper vision: The high-resolution imaging system delivers greater clarity on small labels, recessed SKUs and complex fixtures, improving computer vision accuracy and downstream data extraction.
• Expanded coverage: Building on coverage of hard-to-scan areas such as top stock, upper steel, coolers, freezers and hooks, Tally 4.0 adds support for fixtures such as bunkers to capture more of the store in a single pass.
• Upgraded edge computing: Nvidia Cuda, TensorRT and Doca Argus accelerate onboard processing, reducing latency and time to insight while supporting real-time autonomy alongside depth cameras powered by RealSense.
• Improved 360˚ capture: Dual fisheye cameras enable instant panoramic views, creating denser store context and digital twin-like walk-throughs any time of day or night.

“Running physical AI at the edge is critical to making robots and humans work better together in retail environments,” said Azita Martin, vice president at Nvidia (www.nvidia.com). “Simbe’s Tally 4.0 robots, supported by Nvidia’s full-stack AI infrastructure platform, demonstrate the power of real-time AI, enabling retailers to turn shelf data into immediate, high-impact decisions at the store level, as well as massive operational decisions at enterprise scale.”