01 · The challenge
The client’s suction-only gripper handled boxes well but failed on polybags, soft goods and blister packs: about a fifth of their customers’ catalog.
02 · What we did
- 01
Built a test cell with 300 representative SKUs and a scripted pick protocol so every concept was scored on the same data.
- 02
Explored five gripper architectures in eight weeks, converging on a compliant-finger design that cradles items while suction seals.
- 03
Moved to machined aluminum and SLS PA11 with a quick-change tool plate.
- 04
Cycle-tested three units to 1.2 million picks on our automation rig.
03 · The outcome
The new end-effector raised pick success on the hardest SKU set by 18 percentage points and shipped on the client’s next-generation cell.
“They built a test cell before they built a gripper. That is why it worked.”