MRP Performance Enhancement in 2023.2

@timshuwy Sorry, but we are the exception. Our MRP time actually went up, but I’ll explain.

I tested it on Thursday night (control) and Friday night (with new setting). We don’t do production on Fridays, so this was a good apples-to-apples test since nothing really changed from one day to the next.

  • Thursday night (control): 4:09 (4 hours and 9 minutes)
  • Friday night (new setting): 4:29

Why? Our “level 0-0” jobs took longer to create, for some likely unrelated reason.

  • Thursday night: Averaging 2:48 (2 minutes and 48 seconds) per truck job
  • Friday night: Averaging 2:59 per truck job
  • They both processed the same 171 jobs each

But the plus side is what you predicted, that the lower-level demand (level 1-0 to level 8-0) is processed faster.

  • Thursday night: 37 minutes for 6,479 parts
  • Friday night: 29 minutes for 3,379 parts

So, as promised, it does reduce the number of parts to process. The catch here is that it is a drop in the bucket (specifically an 8-minute drop) in the grand scheme of our 4-hour MRP run.

And this is how it will always be for us. MRP time is all tied up in the BOM-copy of level-zero parts.

I have spreadsheets if you’d like. I made an Excel macro to data-mine the MRP text logs.

2 Likes