Measure the last parts in the bin as well as the first
A full bin may offer many easy grasps near the top while the last few components sit against walls, dividers or corners. A successful top-layer demonstration therefore does not establish the useful emptying rate. Include the lower and near-empty states of the actual container in a 3D bin-picking trial.
Record the remaining parts and their positions when no usable pick is found. Distinguish missing depth information, an unrecognised part, an inaccessible grasp and a robot path that cannot reach the proposed pose. Those are different problems and may call for different changes to camera viewpoint, tooling, presentation or recovery. Zivid's technical guidance identifies bin walls and dividers as causes of occlusion in stationary camera arrangements.
Agree the production objective: complete emptying, an allowable residual quantity or a controlled handover for the final difficult parts. Include any operator intervention when calculating useful throughput across a full bin. This makes the performance statement relevant to replenishment and staffing, not only the fastest individual pick. Send the actual bin dimensions, dividers and photographs of parts at the bottom when requesting a feasibility review.
Why test a nearly empty bin separately?
Walls, corners and dividers can hide or obstruct the last parts even when the top layer is easy to pick.
What should be recorded when a bin-picking cell reports no pick?
Record whether the cause is missing image data, recognition, grasp access or robot reach, plus the remaining part positions and required intervention.
Review 3D camera selection and viewpoint · Compare fixed and robot-mounted vision
