It is easy to think of a liquid handler as a thing that moves liquid, but on any deck bigger than a few positions something also has to move the labware: plates onto and off the working area, lids on and off, tubes to a reader and back. That motion is its own quiet discipline, and when it goes wrong it goes wrong expensively, with a plate dropped or an arm driven into a stack. The forum threads about teaching a gripper its path and adjusting a head's park position are really about the same thing: choreographing movement so nothing collides.
The other robot on the deck
Labware transport is usually handled by a gripper or plate handler: an arm that picks up a plate by its edges and sets it down somewhere else. On some platforms it is an integrated arm that shares the deck with the pipetting head, such as a swap arm, and on others it is a separate robotic arm serving several instruments. Either way it introduces a second thing moving in the same space as the pipetting channels, which is exactly where the trouble comes from. A deck that is fine for pipetting alone can be a minefield once an arm is also traversing it.
How labware gets moved
The moves themselves are a small vocabulary.
- Plate transfers: lifting a plate from one nest or hotel position and placing it in another, or onto a module like a reader, shaker, or thermal block.
- Lid handling: removing and replacing lids so a sealed plate can be worked and then closed again, which matters for evaporation and contamination.
- Regripping and reorientation: setting a plate down and picking it up differently to change its orientation or hand it off to another arm.
- Presenting to instruments: moving labware to a barcode reader, a sealer, or an off-deck device and returning it to the deck.
Geometry, parking, and collisions
The hard part is not the pickup but the path between. The arm and the pipetting head each need a clear route, and their parked positions have to stay out of each other's way. This is why a head's park position during an arm movement is worth tuning: a head left in the wrong place is an obstacle. Getting it right means an accurate deck model, taught or calibrated positions the hardware can trust, and enough clearance that a small offset does not turn into contact. Tall labware, stacked plates, and lids all raise the collision envelope, so the deck that plans for movement leaves room above the labware, not just around it.
It is part of the method, not a preliminary
Labware movement is easy to treat as setup that happens before the real work, but it is interleaved with pipetting throughout a run: aspirate, move the plate to a shaker, bring it back, dispense. That means transport steps carry the same risks as transfer steps and deserve the same dry-running. Simulating a method should exercise the arm moves as well as the transfers, because a collision found on screen costs nothing and a collision found on the deck can cost the run and the hardware.
Before a tip touches liquid, a plate has to get there safely. Give the arm and the head clear paths and honest parked positions, or the most careful liquid class never gets its chance.