Automated Parcel Induction
PickOne’s automated parcel induction solution provides optimized picking and placing of mixed parcels, bags, and products for high-volume ecommerce fulfillment and distribution centers. The award-winning AI vision software easily sorts and classifies new package shapes and types, boosting throughput rates while experiencing less downtime.
How it Works
When parcels arrive in a chute or on a conveyor, the PickOne Perception Kit images the items.
Using 2-D, 3-D, and AI algorithms, PickOne identifies each pickable item in the scene and assigns it an associated confidence score.
PickOne sends the robot controller an array of locations and data for each pickable item via the PickOne API. If no items in the scene have a high enough confidence level, there are two options.
The first option is for PickOne to issue a shuffle command to the robot controller to disturb the parcels and rescan.
The second option is to issue a Yonder request. Yonder sends a picture of what the system sees to a remotely connected Crew Chief who handles the exception by selecting an item in the scene. In seconds, Yonder updates PickOne, and PickOne sends the data to the robot.
In parallel, Yonder stores the Crew Chief’s response, allowing the machine-learning algorithms to make the system smarter as it works. This ensures even higher performance over time.
Upon placement, the PickOne Perception Kit images the place zone to ensure a quality event. If it detects an out-of-spec parcel or a double, it signals the robot to resolve the issue.