ImR Chart Rules
- November 28, 2017 at 3:01 am #55882
As a newbie in 6 sigma, I’m looking for some informations. I’m interested in ImR chart and rules of correct interpretation I-chart, mR-chart and their relation. I found a few of them, but still it’s not enough. For example in the picture below I know, that I should make a new control limits (I-chart, points 46-60) and this new process is stable (mR chart, pionts 29-60).
Also mR chart shows special cause (points 28-29)
But what to do when 1 point in the I-chart (3-rd point) is out of the control limit and mR-chart is ok? Should I run on the production floor and check what happend?November 28, 2017 at 5:24 am #202014
The rules for declaring significant changes in your process for the individuals are: 1 data point outside of 3 standard deviations, 2 out of 3 outside of 2 standard deviations, 4 out of 5 outside of one standard deviation and 8 on the same side of the target.
Looking at your individuals chart it would appear that, for the time interval plotted you had no control at all – you have no distribution of data points with the target as a center point. For the first part of the plot before the large positive spike to a point above the upper 3 sigma limit you have multiple breakouts below the lower 3 sigma and you have a run of 23 points on one side of the target in addition the plot exhibits patterns that have to be multiple occurrences of a violation of the 2 out of 3 above 2 standard deviations.
The process drifted along, out of control on the low side until the big spike upset and then it proceeded to ramp up, cross the target and continue a run until it broke through the upper 3 sigma control limit at which time it settled down to a new out of control condition. All the MR chart is telling you is that the variability associated with the regions of no control is fairly constant.
In summary I would say that your chart has given you a good estimate of the ordinary variation of your process and, since there are no indicators on the graph to note where deliberate process changes were made to try to center production on the mean, it says that you have just let the process rattle around and do whatever it feels like doing.
The graphs suggest that if you can figure out what you need to do to control the location of the process mean you should have a pretty decent process both in terms of target location and ordinary process variation about that target.November 28, 2017 at 7:20 am #202015
If one’s using SPC charts “on the floor”, then the value is responding to out of control situations with known action plans. Ideally, the first time it went out of control, one would investigate why.
I would caution you to constantly be calculating new control limits until a sustained process change has occurred and you have statistical confidence the mean or s.d. has shifted.
Great use of the SPC tool so far.November 29, 2017 at 3:47 am #202018
Thank You for the answer :)
You must be logged in to reply to this topic.