Control Limits for Wildly Out of Control Processes

M

MrJiveMonkey

#1
I am an intern and have been tasked with developing statistical process control charts for a number of processes at my place of employ. Unfortunately, I will not have taken a class on statistical process control until this fall; furthermore none of the industrial engineers here know how to make a control chart. Therefore I have turned to searching the internet, I've found answers to the majority of my problems, but I'm having issues establishing useful upper and lower control limits, as the data is wildly inconsistent.

I have been instructed to use a metric called CPI, the ratio of hours budgeted to a work order/actual hours taken for said work order, as the y value on the control chart with work order # as the x axis. Logically, one can determine the center line here to be 1 (if budget = actual); however I have been having problems with the upper and lower control limits.

As I said, the data is very inconsistent, and there is a tremendous amount of variability in the process at this time. So when I used a moving range to determine the control limits I got control limits somewhere in the area of LCL= 0.32 and UCL = 1.68, which are not viewed as acceptable CPI values by management, nor should they be.

With all that said, how do you determine a useful LCL and UCL when the data is highly variable?
 
Elsmar Forum Sponsor

Jim Wynne

Staff member
Admin
#2
I am an intern and have been tasked with developing statistical process control charts for a number of processes at my place of employ. Unfortunately, I will not have taken a class on statistical process control until this fall; furthermore none of the industrial engineers here know how to make a control chart. Therefore I have turned to searching the internet, I've found answers to the majority of my problems, but I'm having issues establishing useful upper and lower control limits, as the data is wildly inconsistent.

I have been instructed to use a metric called CPI, the ratio of hours budgeted to a work order/actual hours taken for said work order, as the y value on the control chart with work order # as the x axis. Logically, one can determine the center line here to be 1 (if budget = actual); however I have been having problems with the upper and lower control limits.

As I said, the data is very inconsistent, and there is a tremendous amount of variability in the process at this time. So when I used a moving range to determine the control limits I got control limits somewhere in the area of LCL= 0.32 and UCL = 1.68, which are not viewed as acceptable CPI values by management, nor should they be.

With all that said, how do you determine a useful LCL and UCL when the data is highly variable?
Welcome to the Cove. :D

In general, there are two types of limits: control and specification. The former type is calculated from the data; the latter type comes from customer requirements.

This is not going to be easy for you if you've had no training, but if you post your data (and what information you want to derive from it) we can provide better help.
 
M

Markaich

#3
When you do your course, you will not doubt (and hopefully) be told that there is little point in attempting SPC until the process is under statistical control.

SPC is only useful as a monitor of a process that is in control, and things like UCL, LCL and BarX are all data driven, from that stable process.

If I were in your position, the first thing I would try to establish is why the data appears to be out of control. There may, for instance, be time/cost booking errors when people book to work orders.

If the booking process is correct, you then need to try and establish what the causes of variation are and eliminate those that are not natural process variations.

Having gotten the process under control, you would need to let it run for a reasonable, representative time in order to collect the data that will be used to calculate the various SPC data.

Hope this helps a little
M
 

bobdoering

Stop X-bar/R Madness!!
Trusted Information Resource
#4
The other problem is you are putting the cart before the horse...and there is no horse. I recommend looking at the correct steps for implementing SPC seen here. You need to do some preliminary work to see if you are even plotting meaningful data. What are the sources of variation? What is noise (variation you can not control), what can be controlled and made constant, and what can be adjusted in the process - yes, even a transaction process. If you have nothing to adjust or special causes to remove or address, what is going to be the point of doing SPC? Also, this particualre variable measure may be so multimodal with individual contributing variables to be of no value except "report card charting." You will need to find sub-variables to actually control.

Plugging and chugging numbers into equations rarely provides more than additional confusion.

Remember, control limits are there to provide evidence of unusual conditions (special causes). The data you presented to the calculations are presumably "usual" - so even though the control limits are ugly, they are likely telling you the truth. The calculations are not designed to please management, they are designed to the expected variation based on the data you supplied.
 
M

MrJiveMonkey

#5
Thanks for the advice. It all basically confirms what I thought, that we need to stabilize our processes first and foremost. Though I suppose SPC isn't what is needed here at this time.

Based upon the replies I have heard I would like to suggest the following: I go ahead and create charts which depict how the data falls in comparison the idealized range. We proceed to perform studies upon the processes that are seen to be unstable to determine why they are so variable and attempt to reduce the variability in said processes, starting with the worst. These charts would be used more as a means of seeing what processes are most out of control and what order to do studies, rather than SPC charts.

Does that sound like a good place to start?
 

Jen Kirley

Quality and Auditing Expert
Staff member
Admin
#6
Thanks for the advice. It all basically confirms what I thought, that we need to stabilize our processes first and foremost. Though I suppose SPC isn't what is needed here at this time.

Based upon the replies I have heard I would like to suggest the following: I go ahead and create charts which depict how the data falls in comparison the idealized range. We proceed to perform studies upon the processes that are seen to be unstable to determine why they are so variable and attempt to reduce the variability in said processes, starting with the worst. These charts would be used more as a means of seeing what processes are most out of control and what order to do studies, rather than SPC charts.

Does that sound like a good place to start?
Sounds like it to me. :agree1:
 

hngrad1977

Starting to get Involved
#7
Re: Control Limits

Many years I came across some equations that enabled me to calculate control limits for both the average and range without any actual data but have somehow mislocated them. I understand that this is not the preferred method as actual data is the voice of the machine, but need to do this task. Does anyone here in the cove by chance have those equations?
 
B

Barbara B

#8
Thanks for the advice. It all basically confirms what I thought, that we need to stabilize our processes first and foremost. Though I suppose SPC isn't what is needed here at this time.

Based upon the replies I have heard I would like to suggest the following: I go ahead and create charts which depict how the data falls in comparison the idealized range. We proceed to perform studies upon the processes that are seen to be unstable to determine why they are so variable and attempt to reduce the variability in said processes, starting with the worst. These charts would be used more as a means of seeing what processes are most out of control and what order to do studies, rather than SPC charts.

Does that sound like a good place to start?
Yes, that sounds good :)
 

Steve Prevette

Deming Disciple
Staff member
Super Moderator
#9
Re: Control Limits

Many years I came across some equations that enabled me to calculate control limits for both the average and range without any actual data but have somehow mislocated them. I understand that this is not the preferred method as actual data is the voice of the machine, but need to do this task. Does anyone here in the cove by chance have those equations?
If you are doing an x-individuals and moving range chart, the Upper Control Limit is the average (baseline) plus 2.66 times the average range, the Lower control limit is the baseline minus 2.66 times the average range.

I must disagree with the discussion of you can't do SPC on out of control processes. To me, this statement is a non-sequitor. You don't know if the process is out of control until you do the SPC chart. And then you can use the SPC results to help "peal back the onion" and determine why the process is out of control.
 
M

Markaich

#10
Re: Control Limits

I must disagree with the discussion of you can't do SPC on out of control processes. To me, this statement is a non-sequitor. You don't know if the process is out of control until you do the SPC chart. And then you can use the SPC results to help "peal back the onion" and determine why the process is out of control.
I have to disagree, when you start out collecting data you do not do so on an SPC chart. Yes, you chart the data to show how the performance varies but in the intial phases of data collection, there are no UCL, LCL, BarX data shown (given that they can't be calcultated until sufficient data is collected).

SPC is more than just charting, it is the interpretation of the charts and how the data behaves relative to the various lines included on the chart. I'd suggest obtaining a copy of John Oakland's Statistical Process Control...I have the 5th edition with ISBN 0 7506 5766 9...I've found this very useful not only for the mechanics of the approach to Process Management but also the key concepts that makes SPC more than a data collectio/charting/inspection tool
 
Thread starter Similar threads Forum Replies Date
N Apply control limits to a non-normal distribution Statistical Analysis Tools, Techniques and SPC 13
G I-Chart and X bar Chart Control limits Gage R&R (GR&R) and MSA (Measurement Systems Analysis) 3
V Correlating Maximum 'Range' of Values for a given "RSD" (Control Limits) Statistical Analysis Tools, Techniques and SPC 3
H Control limits for varying subgroup sizes Statistical Analysis Tools, Techniques and SPC 1
T What to do if parameters are found to be Outside of Control Limits Document Control Systems, Procedures, Forms and Templates 5
K Historical Data to establish Control Limits Manufacturing and Related Processes 4
B Cleanroom Control Limits: Cumulative or individual? Statistical Analysis Tools, Techniques and SPC 11
D Calculating Lower and Upper 3 Sigma Control Limits Statistical Analysis Tools, Techniques and SPC 2
K Advice on Calculating Control Chart Control Limits Statistical Analysis Tools, Techniques and SPC 13
K Determining Control Limits from a desired Cpk Using Minitab Software 19
P Critical Limits for Air Blower as CCP (Critical Control Point) Food Safety - ISO 22000, HACCP (21 CFR 120) 6
I Shewhart Constants vs Central Limit Theorem in calculating Control Limits Statistical Analysis Tools, Techniques and SPC 18
S Is a Stable Process (within Control Limits) required for Calculating Pp, Ppk? Capability, Accuracy and Stability - Processes, Machines, etc. 6
D Are Four Sigma Control Limits based on Shewhart's work Statistical Analysis Tools, Techniques and SPC 3
K How to Calculate UCL (Upper Control Limits) and LCL (Lower Control Limits) Statistical Analysis Tools, Techniques and SPC 18
T Conducting the "are all points within the control limits" test when there is no LCL? Statistical Analysis Tools, Techniques and SPC 2
D One-Two-Three Sigma Control Limits - Calculating Control Limits for X-bar Charts Statistical Analysis Tools, Techniques and SPC 3
B Precision Machining and Compressed Control Limits Capability, Accuracy and Stability - Processes, Machines, etc. 13
S Calculation of Upper and Lower Control Limits for Control Chart? Statistical Analysis Tools, Techniques and SPC 7
C Cleanroom Class 8 (100,000) Particulate Control and Warning Limits Statistical Analysis Tools, Techniques and SPC 1
S Computing Control Limits in Minitab Statistical Analysis Tools, Techniques and SPC 6
S Setting Control Limits for Particle Counts Statistical Analysis Tools, Techniques and SPC 9
B If I know x-bar, spec limits, and Cpk, can I calculate the control limits? Capability, Accuracy and Stability - Processes, Machines, etc. 16
A Finding Control Limits for a Non-Normal Distribution Statistical Analysis Tools, Techniques and SPC 3
P Help on Treatment of Points Outside of Control Limits Statistical Analysis Tools, Techniques and SPC 4
W Estimated Sigma and Control Limits Statistical Analysis Tools, Techniques and SPC 2
N Variable Control Limits Calculation - P Chart Statistical Analysis Tools, Techniques and SPC 8
J Moving Range Upper and Lower Control Limits when alpha=0.002 Statistical Analysis Tools, Techniques and SPC 4
K Calculating Control Limits in Average and Range Charts Statistical Analysis Tools, Techniques and SPC 3
J Choice of Control Limits - Customer wants to control at 4.5 sigma Statistical Analysis Tools, Techniques and SPC 23
G Adjusted Control Limits - Usage of Control Charts Statistical Analysis Tools, Techniques and SPC 26
J Control limits for one sided control charts using WECO rules Statistical Analysis Tools, Techniques and SPC 5
V Subsets vs. Subgroups - Calculating the control limits of a process Statistical Analysis Tools, Techniques and SPC 6
S Calculating 1 sigma and 2 Sigma limits in Control Charts Statistical Analysis Tools, Techniques and SPC 1
P Help on p-chart - Computing new control limits and chart suitability Statistical Analysis Tools, Techniques and SPC 4
M Control Charts - Points TOUCHING Limits = Stable? Statistical Analysis Tools, Techniques and SPC 12
CycleMike Setting Process Control Limits - Is this process out of control? Statistical Analysis Tools, Techniques and SPC 38
J Calculating Control Limits for nonconformities - 1000 accounting transactions per day Statistical Analysis Tools, Techniques and SPC 2
B Calculation for extending control limits Statistical Analysis Tools, Techniques and SPC 7
M Interpreting my R+R chart - NDC's and Out of Control Limits Gage R&R (GR&R) and MSA (Measurement Systems Analysis) 5
W How to justify Widened Control Limits - No Assignable Cause scenario Statistical Analysis Tools, Techniques and SPC 26
A Control Chart and Limits For Large Sample Size Statistical Analysis Tools, Techniques and SPC 6
A Tight Control Limits - Tape & Reel Process, the process of packaging ICs Statistical Analysis Tools, Techniques and SPC 27
Marc When to Recalculate Control Limits? SPC Statistical Analysis Tools, Techniques and SPC 14
S To take into account the tool wear during calculating control limits Statistical Analysis Tools, Techniques and SPC 6
C SPC: How to set up control limits on SPC chart based on historical data? Statistical Analysis Tools, Techniques and SPC 7
A Formula for I Control Chart Limits - Different values for E2 constant of IMR? Statistical Analysis Tools, Techniques and SPC 5
E Formulas and Constants - Control Limits for Individual and Moving Range Control Chart Statistical Analysis Tools, Techniques and SPC 2
A Cpk = 1.33 if control limits are +/- 3 sigma - True? Statistical Analysis Tools, Techniques and SPC 17
D Why the tight control limits? Histogram of the data is slightly skewed to the left Statistical Analysis Tools, Techniques and SPC 17

Similar threads

Top Bottom