The notes are scrappy, but they distinguish the tray adjustments that aren't obvious in the footage. They also show most missed pickups occurring after tray replacement across the candidates.

Use those notes to plan a common tray trial, including normal replacement. Agree what counts as acceptable placement and operator help, then record both for each candidate

We had a separate transfer application where the locating step was awkward, but the continuous run looked fine. A tray replacement by another operator exposed it.

@EmmaArcher0376 How did you attribute the locating work in your comparison? Our discussion repeatedly stalls over whether it belongs against the robot or the fixture.

For my setup, package cost, labelled as fixture work. Somebody still had to pay, whichever column won.

Keeping the same gripper would make the comparison fair too. That removes another variable while you sort out the tray issue

@HassanBennett0704 Same gripper, unfair adapter? That's still a skewed comparison if the mounting suits one arm better.

@HanaCarter0970 Fair correction; I overstated that. Keep the gripping requirement consistent, document adapters and tuning, and assess the actual proposed tools where a shared gripper would distort things

Does that mean a candidate can arrive with a custom fixture and still be compared? I'm trying to understand where the boundary goes.

Yes, if it's part of what you'd buy. Keep the incoming samples and required result consistent, and include that fixture's cost and changeover work

I've settled the next decision: use a common tray trial with agreed placement and intervention limits. Any choice of the arm will depend on that result; the existing videos don't settle it.

That gives the comparison a clear purpose. Keeping the purchase conditional fits what the existing videos can actually tell you