Created Date: November 12,2014

Created by: franji1

Do you need to utilize a percentage? Could you simply count the number of "good " parts between "bad " parts? Basically, reset your current "good " part counter every time you receive a bad part.

Then, whenever you receive a bad part, right before you were going to reset your "good " part count, see if it is less than 50 or 75 or 100 (or whatever), then you know you got two "bad " parts between 50/75/100 "good " parts, and shut down based on not quite getting up to that value.

You could then also track your best/worst "runs " of good parts as a completely separate "statistic ".

What I am describing is not a percentage. But with a percentage, do you "quantize " on every 100 parts, or do you try to do a running average or ?. What if you want it to be 1.5% or 0.75% or ? By doing the number of good parts between bad parts, you have more resolution on the feedback (e.g. you can tweak the "between " count to be 50 or 51 or 113 or 99 or ...).

Again, you can also keep separate statistics to track global "good part " count and a global "bad " part count to calculate an "overall " bad part percentage

(100 * bad) / (good+bad)

that you reset every shift/day/month/year (or have counts and percentages for each "period "!).

Created Date: November 12,2014

Created by: evilblackdog

We currently stop our production machine for every fault that occurs in order to keep them abreast of how the machine is running. I would like to try a different method where the PLC keeps a running percentage of any of the given faults for the past 100 parts and only stops the machine if any one of the faults exceeds a 1% occurrence rate. We are using 405 series plc's. What's the easiest way of going about this?