I proposed UR5e tray unloading for small plastic enclosures, but priced the arm and gripper while leaving fixture work and staff effort vague. My colleague is annoyed (fair enough). What small pilot would let us judge the whole job?
Agree the question before the trial: whether the proposed route reduces total work for an identified enclosure mix, within a stated development allowance. Record the current route over the same boundaries, including tray preparation, rejected items and changeover rather than comparing only the transfer motion.
And let your colleague help set those boundaries. They may be objecting to being assigned hidden work, not to the robot, and a beautifully measured transfer will not answer that complaint.
On a handover I helped review, the demonstration began after someone had arranged every part. The operator could repeat the motion but not reproduce the starting condition. Put preparation instructions and who can follow them into your pilot, not just a labour estimate somebody signs after watching.
Luis, I agree, though I would bound the documentation effort too. Gabriel is trying to buy evidence for a decision, not commission an entire production system before being allowed to decide whether it is worthwhile.
Yes, enough to repeat the pilot honestly. I did not mean a finished production handover. If the preparation cannot be described yet, record the assistance it actually requires.
Keep development time separate from recurring operator time, while showing both in the decision. Otherwise the first pilot can look permanently expensive because of learning, or permanently cheap because everyone quietly excludes the work that made it run.
Who observes it while you operate? I would ask your colleague to own the time record, with you checking that assistance and interruptions are noted rather than argued about afterwards.
Give that operator room to say a step is impractical. The pilot is supposed to discover work, not recruit somebody to certify that your current instructions are reasonable.
I would be interested in the preparation result when you have it. That was the cost your original estimate could not answer, even if the robot transfer itself turns out to be straightforward.