@YasminAdams0093 Barely, but yes. The notes catch nudges the clips don't show clearly. Across the candidates, most misses followed tray replacement.

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

@YasminAdams0093 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.

@EllaArcher0381 Robot cost or fixture cost? Our meeting keeps getting stuck there.

@SarahBell0676 In my own comparison, we included the cost in the complete package and identified its source as fixture work. How we categorised it didn't change the need to fund it.

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

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

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

@YasminAdams0093 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 kept the awkward samples. Normal batch variation, not specially selected demo-ruining rejects.

@SarahBell0676 Then include those alongside the easy ones. A comparison made entirely of awkward samples could misrepresent the mix just as badly as the polished demo.

@FarahAbbott0069 Empty pockets happen here too. How do we test that without improvising around a moving arm?

@SarahBell0676 Ask the person responsible for the equipment to plan that test. Prepare the tray while access is safe, and agree the expected response first

For our job, we logged assistance during changeover too. The second operator's quick result depended on constant prompts from the demonstrator, which wouldn't match our staffing.

@EllaArcher0381 Our clips crop out the person doing the nudging. Not deliberately, but the framing makes the setup look more independent than the notes support.

Wider video, job done? Or does someone still count all the help?

Keep a simple tally as well. Video helps explain an intervention; the tally makes it easier to compare runs without relying on everyone's memory

We've a way forward: a shared tray trial with placement and operator-help limits agreed first. Choosing the arm stays conditional on that comparison, rather than the existing clips.