Our UR5e transfers the pouch reliably in the supervised example. I have since made guides for two more shapes, added sorting and found myself smoothing the pouches before loading. The manual packer used to pick them up as they came. I am starting to suspect I have automated the easy movement and adopted everything else as my new occupation.
Compare the complete current task with the complete proposed task, using representative orders and the same start and finish. Keep stopping as a legitimate result; the time already spent should not force you to keep designing around every new pouch.
I have manual timings from packed pouches arriving to filled trays leaving. The robot trial starts after sorting and smoothing. So the comparison I showed was incomplete.
What was the original aim: less time, less repetitive handling, or removing an awkward reach? A faster-looking transfer can fail one aim while helping another.
And what happens to the pouches that don't sit nicely in the guide? I'd follow a real mixed pile, not just the ones that made it into the demonstration tray.
On a handling job I helped review, the machine removed one repeated reach but the feeder gained repeated wrist work preparing each part. We had to look at where the effort moved, not only how many robot cycles succeeded.
You can observe that before making another guide. Have the packer walk through the whole proposed task with the current guides, including the set-aside pile and the change between shapes.
Nora's aim question matters to the stop decision. Record the packer's experience of the repeated movements alongside time, otherwise we could call a less comfortable method a success because its stopwatch result is slightly better.
Exactly. Rebecca, does smoothing the pouches bother the packer physically, or is it mainly an extra operation they dislike doing? Let them describe it rather than deciding for them.
She says the repetitive smoothing is more annoying on her hands than the original placement. That is her experience, not a measurement I've made. She is willing to compare the whole order with me.
Good that you asked her. Keep the awkward ones in the order count too; our quality hold pile once made a process look quicker simply by hiding its unfinished work.
I wouldn't assume another guide is the next step from that answer. You may need a different presentation process, or no automation here. Let the comparison decide whether further design has a useful question to answer.
We've agreed to observe an ordinary mixed order, with sorting, set-asides, guide changes and the packer's comments included. No new guide work while we do that.
Omar, that would help, and I'd keep the written observations beside any totals; two methods can take similar time while asking very different things of the person feeding them.
Our supervisor will record while the packer does the work. I am finding the pause harder than expected, but at least I am not spending another evening proving I can make a fourth guide.
The pause is part of making the decision, not a commitment to abandon it. Keep the alternatives open until the whole-order comparison is in, including the option to stop without immediately inventing a narrower replacement trial.