automated depalletizing in a warehouse

AccuPickCase Study

Automated Warehouse Depalletizing


This customer is the Tobacco Authority of Thailand, a state enterprise that manufactures and distributes tobacco for local consumption and export.


Picking Randomly Positioned Boxes

TAOT had a specific requirement for a solution to enable their Comau robots to pick boxes that were randomly positioned in sets of one, two, or three at a time, and transfer them onto an adjacent conveyor belt. However, managing the entire fleet of robots was challenging due to unstable lighting conditions and the irregular placement of boxes during the handling process.


Automated Depalletizing Using AccuPick

AccuPick combines AI deep learning with machine vision for automated depalletizing by picking and placing multiple boxes at a time. In the TAOT warehouse, AccuPick enabled the Comau robots to use multiple suction grippers based on the surface area detected by the camera. Even with limited space for placement, AccuPick located the boxes, removed them from the pallet, and placed them in the desired location.

For accurate detection and scanning of the boxes, AccuPick’s ‘eye-in-hand’ mode was employed, involving the attachment of a compact RealSense camera sensor to the robot arm. AccuPick established seamless communication with the Omron PLC, effectively controlling the robots and significantly reducing cycle time to just 5 seconds per box.

The integration of Solomon’s technologies enabled the robots to achieve remarkable accuracy in picking and depalletizing, regardless of challenging lighting conditions. Additionally, our advanced vision capabilities enabled the robots to adapt to changes in object positions as the boxes moved along the conveyor belt.


Reduced cycle time to 5 seconds per box
Successfully integrated a previously unsupported robot brand with a PLC for efficient control
Achieved high accuracy and efficiency in picking despite changes in box position and orientation during depalletization
More Case
  • brown cookies on white ceramic plate

    Defect Inspection in Processed Food Production

    Solomon’s visual inspection solution, SolVision, uses AI to learn various types of food product defects and recognizes them as they pass through fast-moving conveyor systems during food processing
  • assorted metal parts

    Metal Parts Bin Picking

    Solomon’s AccuPick and SolScan technology enable accurate component orientation identification, facilitating precise positioning, while AccuPick’s ROS Motion Planning ensures collision-free picking.
  • AI Inspection of Plastic Buckles

    Powered by artificial intelligence, SolVision features four individual tools designed for different inspection tasks.
  • Close-up Photography of a Power Tool

    AI Inspection of Stamped Metal Parts

    There are many types of defects that may appear differently each time on the stamped parts, in particular oil or water stains, which are not easily detected.