Tally 3.0 store robot gets camera and processor upgrade to spot out-of-stock items, improve sales

The new Tally 3.0 retail robot moves through stores to track out-of-stock inventory. Upgrades include a LiDAR solid-state camera from Intel and an Nvidia GPU for edge processing. (Simbe Robotics)


Venture-backed Simbe Robotics announced a upgraded version of its Tally retail robot with an Intel RealSense LiDAR camera for greater resolution and an embedded Nvidia Jetson GPU for edge processing.

The new Tally 3.0, announced Thursday, updates the original that debuted in 2015 and has now been deployed in more than a dozen of the top 250 global retailers including Giant Eagle and Decathlon Sporting Goods.

Free Daily Newsletter

Interesting read? Subscribe to FierceElectronics!

The electronics industry remains in flux as constant innovation fuels market trends. FierceElectronics subscribers rely on our suite of newsletters as their must-read source for the latest news, developments and predictions impacting their world. Sign up today to get electronics news and updates delivered to your inbox and read on the go.

The tall, slender robot is used for inventory management, moving up and down store aisles up to three times a day. It relies on RFID scanning tech, computer vision and AI to find out-of-stock items and offer insights on ideal product location.

Schnuck Markets, based in St. Louis, Missouri, uses Tally for better decisions about providing supplies to its stores located in five Midwestern states. “The amount of critical data and valuable insights that Tally continues to provide is immeasurable,” said Dave Steck, vice president of IT for Snuck in a statement.

Simbe claims that Tally can detect up to 10 times more out of stock items that human store teams and pays for itself in the first month of deployment.  With improved data, stores can see a 2% improvement in annual sales, the San Francisco-based company said.

Simbe offers the Tally robots as a service based on a monthly subscription fee that is calculated on the number of items in a store, a spokeswoman said.

The enhanced optics in the new version include Intel RealSense depth and RGB color cameras to view more items on the shelves, including the top shelf overstock items. The cameras have been added to a taller version of the robot. A new autofocus optical system captures shelf tags and can read data from 30 inches to bring recognition accuracy to nearly 99% and allow the robot to continue moving without slowing down.

The company also estimates the robot is five times more resilient than earlier models, allowing 5,000 miles per robots. Inventory data from store shelves is sent to a Simbe cloud platform, which is where high-def 3D images are stitched together.  The embedded computing platform in the new version of the robot uses an Nvidia Jetson GPU, allowing AI to run at the edge. Using a hybrid of edge and cloud allows stores to get shelf information more quickly and lowers the cost compared to uploading data to a retailers’ data system, Simbe said.

RELATED: Bossa Nova’s robots are trained to perform in sun glare

Suggested Articles

In June, Su stood out for heartfelt commentary on social injustice after the killing of George Floyd. She challenges industry to do more.

New chip will be incorporated in coffee mug-sized device to cost $199 next year

In a nutshell, Edge ML is a technique by which smart devices process data locally using machine and deep learning algorithms.