Robotics
Robot vision: from 2D guidance to 3D bin picking.
Computer Vision that lets the robot act: cameras, software, AI and integration to cobot or industrial robot. From 2D guidance to 3D bin picking.
Solution path
Pick the guidance type for the scene
Recognition and pose are control input, not optional extras.
Progression
From fixed presentation to bin picking
Choose by presentation variation, not the newest sensor. Sedeco is the eye-and-brain layer: sensor → AI/CAD → grasp → robot → gripper → production system. Cobots and industrial robots; Fairino where that integration is validated.
| Scene | Usual stack | Note |
|---|---|---|
| Fixed nest / known pose | 2D smart or PC | Fast cells |
| Pose variation on conveyor | 2D or 3D by tolerance | Lighting critical |
| Bin picking / clutter | 3D + PC | Cycle time budgeting |
| Force + vision loop | PC + robot controller integration | Deterministic I/O |
How Sedeco closes the loop
Select cameras and optics for the task. Detect reliably under real variation. Send pose or offsets to cobot or industrial robot, with PLC handshake when the cell demands it. Precision placement may need local refinement or extra sensing.
Put guidance into motion
From cell concept to closed-loop control on the floor.
When the bin is messy
3D bin picking path
If parts arrive in bulk with occlusion and changing pose, continue to the dedicated 3D bin picking solution and decision guide — not only generic robot vision.