Choose 2D robot guidance
Flat conveyor or tray, top-down view, known part height, minimal overlap. Vision finds position and rotation in the plane.
Knowledge · Robotics
When is 3D bin picking needed?
Not every robot vision task needs full 3D bin picking. Use this frame to decide between planar 2D guidance, structured 3D guidance, and picking from messy bins with occlusion.
Decision frame
Messy bins and occlusion vs planar pick
3D bin picking applies when parts arrive in bulk, overlap, and change pose cycle to cycle. If parts are singulated on a flat surface with stable height and orientation, 2D or structured 3D robot guidance is often enough.
Choose 3D robot guidance
Structured feed with height variation or tilt, but parts remain separated. You need depth and pose, not full bin search.
Choose 3D bin picking
Random pile in a bin or tote, mutual occlusion, unknown top part, and collision-aware grasp planning per cycle.
Signals
Questions that push you toward bin picking
- 01
Can a single 2D view see the grasp surface?
If another part hides the pick face, 2D fails regardless of detector quality.
- 02
Is part height or tilt unpredictable?
Depth and pose estimation become mandatory, not optional.
- 03
Must the robot avoid bin walls and neighbors?
Collision checking and approach vectors are part of the vision stack.
- 04
Is singulation upstream too costly?
If mechanical separation is rejected, bin picking is the Computer Vision path.
Unsure which path fits?
Describe your feed, bin geometry, and cycle target. We map it to 2D guidance, 3D guidance, or full 3D bin picking.