Process Control Chart using minitab

So, we have gathered 100 raw sample data from experiment and got this beautiful

From x-bar chart, we see that there are one point that is out of our upper control limit, which I assume its acceptable in real world manufacturing, and have no points outside of our control limit on R chart which is good. This indicates that our process was in control. However!!!!, Cp and Cpk is poor!!! meaning that our variation within and between individual samples were huge. Our process, although it was stable, is not capable of making products within its specs. If this was real manufacturing company situation, we should have went out of business due to poor reputation from this...... What we did next was trying to improve these numbers by come up with new strategy. Okay.... we decided to implement some magical method to improve our manufacturing process. lets see what it gives us after doing the same thing

 okay........... X bar chart looks better.... our Cp increased little bit... out Cpk decreased........... this is because our mean has shifted due to this change in process. Anyways... not too much improvement... we just wasted money and time.... So! now, we tried DOE, design of experiment, to see what is actually causing this variation. We had 4 factors in our DOE. Operator, hand, position, and cup level.

hmmmm............ Looks like that there is significant drop in process mean when cup level is different with operator 2. I mean overall, operator 2 have relatively larger variation than operator 1. From this data, I would say that most critical factor that influences our variation is cup level. So lets go fix cup level to optimal level!!!!!