As planned, today was dedicated to the manual programming of a robot.

In the morning hours, the Fuji instructor walked us step-by-step through the process.  In basic terms, this involves teaching the robot each position it needs to move through to pick up bags/boxes from an in-feed conveyor and load them onto a pallet in a predetermined pattern.  Sounds relatively simple at first, but it’s very much an exercise in precision, as four linear/rotational parameters need to be determined for each position.
The trainees were given their opportunity to get dirty during the afternoon.  Each of us was tasked with programming a robot to load a pallet in a sensible pattern.  The robot programming is done using a teaching pendant, aka a corded remote control.  So you can physically make the robot move in certain ways with button presses.

In my case, I was palletizing bags of kitty litter – 5 bags to a layer, 3 layers to a pallet.  My initial goal was simple, don’t crash the arm of the robot into anything.  After a few minutes of operation, I felt comfortable enough to take on the real challenge of creating the program.  I walked through the steps demonstrated for us earlier and after about 90 minutes had a program ready to test.

We broke the test up into two parts.  The first part was to run the robot through all of the motions, but without product.  The purpose here is to observe the sequence of steps and identify any fundamental motion flaws that may be obvious.  All was OK, so we then ran the test with product.  There were some product positioning tweaks that were needed.  But the software is flexible enough to allow those to be made at the control panel rather than running the robot through the steps again.  All in all, it went well – in other words, no crashes.

Tomorrow will again be focused on robot programming, but in this case utilizing the library of standard pallet patterns that Fuji has created.  The approach here is very different and should prove to be more efficient in the long run.  Until then, Siyanora!