Simbe Robotics Inc. today announced the launch of Tally 3.0, the latest version of its autonomous mobile robot for retail inventory. The San Francisco-based company said it has enhanced the robot’s optical system and improved its durability and maneuverability. Tally 3.0 also includes an embedded data processor.

Simbe introduced Tally in 2015. The robot is designed to improve retailer efficiency by autonomously providing store associates with timely information. Tally uses a suite of sensors to operate safely during normal store hours alongside shoppers and employees, and it does not require any infrastructure changes, said the company.

“At Simbe, our ongoing commitment to best-in-class design and unparalleled data accuracy has empowered our partners with the real-time shelf insights they need to protect their bottom line and competitively serve customers in a rapidly evolving retail landscape,” stated Brad Bogolea, Simbe co-founder and CEO.

Tally 3.0 includes new optical sensors, endurance

Simbe said it has added autofocus capabilities to Tally 3.0 with more cutting-edge sensors that improve depth perception for inventory detection. The company also made the mobile robot‘s base taller to make room for additional Intel RealSense depth and RGB cameras to see more, including top-shelf overstock items.

Tally’s new optical system can accurately capture shelf tags and read data from up to 30 in. away, enhancing the robot’s recognition accuracy to nearly 99% and without slowing its traversal speed, claimed Simbe. The robot can scan an entire store up to three times per day and autonomously return to its dock.

In addition, Simbe said it used the newest mobile technology to create a more cost-effective and resilient Tally robot with the same compact, slim form. Each Tally 3.0 is designed to traverse more than 5,000 miles in its lifetime – at lower cost to scale, said the company.

Enhanced computing with hybrid data processing

As Tally moves through store aisles and captures real-time shelf data, it sends inventory information to Simbe’s proprietary, secure cloud platform. The software uses machine learning to enable store teams to stitch together high-definition 3D images of the store shelves.

Tally 3.0 also features the NVIDIA Jetson GPU-accelerated embedded computing platform, which Simbe said enables it to be more efficient by running deep computer vision and AI algorithms on the edge.

The additional processing capability and hybrid approach improve the efficiency of Tally’s data capture, providing Simbe’s customers with shelf information more quickly, according to Simbe. The cloud/edge approach also helps minimize the costs of capturing and uploading the robot’s data to retailers’ back-end systems, it said.

Simbe claims category leadership

Simbe said that Tally’s combination of RFID scanning technology, computer vision, and AI-powered analysis enable it to provide actionable, e-commerce-level insights to physical retailers, solving for the $1.75 trillion “ghost economy” defined by out-of-stock items, inaccurate price execution, and lack of product location optimization across the industry.

“The amount of critical data and valuable insights that Tally continues to provide is immeasurable,” said Dave Steck, vice president of IT infrastructure and application development at Schnuck Markets Inc. “Our long-standing partnership with Simbe has enabled us to provide better decision making throughout our supply chain while offering store teams a powerful tool to ensure products are available for our customers.”

Simbe cited a global consulting firm that reported that Tally can do the following things:

• Detect up to 10x more out-of-stock items than manual audits conducted by store teams
• Average a 20% reduction in out-of-stock items at the store level
• Pay for itself as quickly as the first month of deployment at a store
• Yield more than 2% in annual sales lift due to better data and improved store execution

Tally has been deployed in more than a dozen of the top 250 global retailers in the Americas, Europe, the Middle East, and Asia. Through its work with retail partners including Carrefour, Decathlon Sporting Goods, Groupe Casino, Giant Eagle, and Schnuck Markets, Simbe said it is building one of the most valuable datasets in the retail ecosystem.