SICK – Robot-Guided Bin Picking for Automotive Parts Handling

SICK delivered a 3D vision-guided bin picking system enabling industrial robot arms to grasp randomly oriented metal stampings and castings from bulk bins without fixtures or manual presentation. The 3D camera captures a point cloud of the bin contents; the software identifies graspable instances, calculates collision-free pick poses, and commands the robot arm to pick and place parts onto a conveyor or fixture. The system handles multiple part geometries by switching between product models stored in the software library. Grasp success rates exceed 95% for most parts, with failed picks recycled automatically. The project replaced a manual unloading step that previously required one operator per bin-feeding station. Cycle time for bin feeding was reduced and labor was redeployed to higher-skill tasks.