Process Capability: Is Your Process Up To The Task?

Posted on28. Apr, 2009 by carolesf.

Process Capability

is a measure of how well your process performs — that is, the proportion of in-specification items that is produced by the process — when it is in statistical control. 

Note that this is not a measure of your process’s actual current performance on producing items in-spec.  (That’s batch performance.)  It is possible that your process may not be in statistical control at present, and therefore the state of current production is not necessarily relevant.

So to understand the concept of Process Capability, we need to recall what “in control” means in a statistical sense.  Statistical Process Control (SPC) refers to the use of statistics to monitor the variability of a manufacturing process.  The idea is to maintain control of the process such that some large proportion of items produced by that process are within design specifications.  Note that SPC makes no reference as to whether or not the design is a good one; a poorly-designed product will perform suboptimally no matter how well the manufacturing process sticks to the design parameters.  So Statistical Process Control does not necessarily ensure customer satisfaction or product reliability. 

However, let’s assume for the moment that the design is a robust one and the in-spec product will meet the customer’s needs.  So minimizing manufacturing defects is a definite must.  We do this using the tools of Statistical Process Control.  The most basic of these is called, naturally, the Control Chart.

Control Charts use statistical tools (based on sampling of items produced) to monitor both the central tendency of the manufacturing process, and its deviations away from the center.  The idea is to flag variations in production that may lead to items being rejected at the end of the manufacturing process. 

The manufacturing process is said to be “in control” (statistically speaking) when variation among items produced stays within the control limits.  An item falling outside the control limits, or a series of increasing or decreasing data points, should be investigated to see if there is a special or common cause that is driving these variations. 

Special cause, common cause, what’s all that? 

Say you are baking several batches of cookies.  Every batch is slightly different.  Some cookies are thinner, some are thicker, some are more done, some are less done.  However, as long as they are not charred briquettes, your spouse will eat them (in-spec). 

Now the first batch is perfect golden brown.   On the second batch, your best friend calls you on the phone just before the timer goes off.  You don’t hear it.  You forget about the cookies.  Briquette city.  Out of spec.  This represents a special cause.  Your friend called; that’s not likely to happen every time you make cookies. 

So you throw that burned batch away and start over.  Your spouse likes the cookies so well that the next week you make some more.  And the week after that, and the week after that.  Your process seems in control. 

On the fourth week, however, you start to notice that the cookies are seeming less and less done when the timer goes off.  Downright raw; out of spec.  You investigate; by putting in an oven thermometer in, you discover that your oven is not heating up as well as it used to do.  This is a common cause.  Until you get your oven repaired, none of your cookies will be baked in-spec.  (Unless your spouse likes to eat raw cookie dough.)  Your process is out of control.

Now that we understand in control, let’s revisit:  Process Capability is a measure of your process’s potential to produce items in-spec, assuming that your process is in control.

Welcome back to Lean Six Sigma Source! Thanks for your continued support.

Continue Reading