Goods to Person: Segmented Tote Picking

[ The Problem: ] A global cosmetics brand is expanding its ASRS system to 8 goods-to-person pick stations, effectively doubling its capacity. Their goal is to robotically pick from 4 of these stations.

[ The Solution: ] The PickOne Perception System identifies the pick points of randomly oriented items in a tote to allow a robot to pick them. The System verifies placement and addresses errors by issuing robot commands to reposition errant items.

[ Human In The Loop: Yonder ] In the instance that an item cannot be picked, PickOne initiates a Yonder Request for a remote Crew Chief to intervene. In seconds, the Crew Chief informs the robot what to pick and the system continues.


Data retrieval start. 42%… 57%… 94%… Transmission successful. Streaming now.

This software module extends the power of PickOne for applications involving picking cases, cartons, trays, and bags from mixed random and homogeneous skew pallets. 

Depalletizing cycle start. Layer by layer. Picking initiated.

The PickOne software analyzes the 2D, 3D and AI data to identify each pickable item on scene and assigns each one an associated confidence level. If no items on scene has a qualified confidence level, PickOne generates a yonder request for a Crew Chief to handle the exception by selecting an item to be picked.

Item successfully picked.

Yonder stores the Crew Chief’s responses allowing the machine learning algorithms to make the system smarter as it works. This ensures higher performance over time.

Message successfully transmitted. Ending transmission now.