Skip to main content

​
Project background

A chemical producer in Anhui previously relied on manual workers to depalletize and move soft packages. The packages deform easily and use different packaging materials, creating efficiency, dropped-package, and package-breakage risks.

​
Solution

The project adopted a 3D vision soft-package depalletizing solution. An RGB-D camera captures package texture and 3D point-cloud data. The algorithm identifies package deformation and stack position, then provides the result to a robot arm for picking and transport to the unloading port.
Chemical soft-package depalletizing system on site

​
Implementation workflow

1

Recognize the package

The camera captures texture and point-cloud data and identifies the position and shape of the target package.
2

Plan the grasp

The system generates the robot-arm grasp sequence and provides target positions to the control system.
3

Move the package

The robot arm moves the package to the unloading port and hands it to the next process.
The depalletizing algorithm runs on the camera, reducing external compute requirements. Deployment still requires validation against ambient light, package condition, and robot cycle time.

​
Results

The project upgraded soft-package depalletizing from a manual operation to an automated workflow, reducing manual participation and improving process consistency. The on-site demonstration is shown below.