You are reading the documentation for an older Pickit release (3.2). Documentation for the latest release (3.3) can be found here.

Teach box model

Teach with box model is the recommended engine for applications with randomly oriented box-shaped objects of known dimensions. Typical applications include bin picking of billets and pipes with square cross-sections, to name a few.


  • For box depalletization, and picking boxes from semi-structured patterns, the recommended engine is DeepAL.

  • For bin picking rectangular objects of varying dimensions, the recommended engine is Flex.

To teach a box model all you have to do is to specify the box width, height and length. Pickit will then generate a box model.

../../../_images/box_model.png ../../../_images/square_billet.png

By using the box model, you let Pickit optimize the detection and user experience for this specific shape. Pickit will automatically create eight pick points: Four on the long faces (below left) and four on the long edges of the box (below right).