Locked lesson.

Upgrade

About this lesson

In this lesson the strategy for how to use Control Charts is discussed. In addition to monitoring the process, the charts can be used to set performance baselines, validate the impact of improvement, and identify sources of variation.

Quick reference

Statistical Control Stategies

SPC Control Charts can be used for more than just maintaining control of an existing process.  The charts can be used to set performance baselines, validate the impact of improvement, and identify sources of variation.

When to use

Statistical control strategies, especially those involving control charts, can be used throughout the improvement process in addition to maintain control during operations.  In particular, on a Lean Six Sigma project they can be used during the Measure phase to establish a performance baseline, during the Analyze phase to identify special causes and common causes for variation, during the Improve phase to validate the impact of a performance improvement, and during the Control phase to ensure the change is fully implemented and becomes the new standard process.

Instructions

Statistical process control can be used both during normal operations and during the improvement process.  The strategies in each case changes slightly.  During operation the strategy is to reach an ideal state of process performance.  During problem solving the strategy is to improve performance and regain stable operations.

Normal Process Operations

During normal process operations a process can be in one of four states depending upon whether or not it is in statistical control and whether or not it is consistently providing results that conform with the process specifications or requirements.  This is shown in the diagram below.

When the process is in statistical control and producing conforming results, an “Ideal” state exists.  The process will consistently deliver conforming results. 

When a process is in statistical control but not always delivering conforming results, it is in a “Threshold” state.  Because of the statistical control it is predictable, but that predictability includes a cost of quality for those instances when the results are not conforming.  The process manager is on a threshold.  He or she must decide whether the cost of improvement to reduce common cause variation is greater than the cost of non-conformance.

When the process is not in statistical control and not producing conforming results, we say it is in a state of “Chaos.”  Fortunately, this is normally very obvious and therefore it is relatively easy to initiate an improvement project.   The SPC charts provide insight into how bad things are.

The final quadrant is the most dangerous.  In this case the process is not in statistical control, but luckily it has not yet created any non-conforming results.  We call this state the “False Sense of Security.”  Since it is not in statistical control, it is not predictable.  The result tomorrow could be non-conforming. However, unlike the Chaos and Threshold states, we don’t expect non-conformances and often are not checking for them.  That is why an SPC chart in this case is a crucial element of process management.  It will alert you to problems.

Process Improvement Projects

SPC charts and statistical control can also be used very effectively during an improvement project.  I will illustrate this by discussing how to use SPC control charts in the phases of a Lean Six Sigma project.

Define – Control charts may occasionally be used in this phase to help justify the need for a project.  However, this is not common.

Measure – Control charts are used in this phase to measure and record process performance.  During this phase, the process is measured and mapped.  Control charts are an excellent tool for use in this phase.

Analyze – Control charts are used in this phase to assist in the identification of special cause variation and an assessment of the level of common cause variation.   Control charts do this very well since they were designed to identify special cause variation.

Improve – Control charts are often used in this phase to demonstrate the impact of a corrective or preventive action.  A “before” chart is created that illustrates the process performance including mean and control limits before the change is made.  Then after the change is made, an “after” chart is created with the new mean and control limits.  Generally, we seek to center process mean and narrow the band between the control limits.  These charts show whether that improvement has been successful.

Control – Control charts are a primary tool for this phase.  Once the improvement is in place, control charts will notify the operator and process manager of problems maintaining the improved performance.

Hints & tips

  • Control charts are easy to create and use, so use them.  They are a big help in both managing a process and improving it.
  • It is often hard to get approval to improve a process that is the “False Sense of Security” quadrant.  Since it is not currently a problem an improvement is deferred; the business wants to focus on current problems.  The use of a control chart will at least notify the process operators and process management when things suddenly get worse so they can at that time take action.
Login to download
  • 00:04 Hi, I'm Ray Sheen.
  • 00:06 Statistical process control can be used both for managing existing processes but
  • 00:10 also during the improvement process to aid the improvement team.
  • 00:15 Let me start with a discussion of process control during normal operation.
  • 00:19 I'll focus on the state of your current process performance.
  • 00:23 Processes can be divided into four categories of process, performance and
  • 00:26 control.
  • 00:27 Processes can either be in statistical control or out of statistical control.
  • 00:32 And processes can either be providing results that are conforming
  • 00:35 to the customer's expectations or
  • 00:37 results that are not conforming to the customer requirements.
  • 00:40 At least some of the results are not conforming.
  • 00:43 And so let's look at each of these four conditions.
  • 00:46 When a process is under control and providing results that are conforming,
  • 00:49 it is in the ideal state.
  • 00:51 The process performance is good and is likely to remain good.
  • 00:54 That is the place where we want all processes to end up.
  • 00:58 On the opposite extreme, when a process is not under statistical control and
  • 01:02 is not providing results that are conforming, we have a chaotic state.
  • 01:06 There is a high cost and excessive delays that are caused by constantly reworking
  • 01:11 results or re-running the process hoping to get an occasional good result.
  • 01:16 This process needs immediate improvement in the SPC
  • 01:19 control charts could be used as evidence of the need.
  • 01:22 Then there's the state of a process that is under statistical control but
  • 01:26 not producing results that are consistency conforming to the customer requirements.
  • 01:31 This can happen when the customer requirements and
  • 01:33 the tolerances are tighter than the normal common cause variation in the process.
  • 01:37 The process is under control,
  • 01:39 it's just not capable of always providing what the customer wants.
  • 01:43 We refer to this as threshold performance.
  • 01:46 The process is on the threshold of becoming capable.
  • 01:49 The business needs to decide whether the cost of quality for
  • 01:51 fixing the problem is greater or less than the cost of an improvement project.
  • 01:56 SPC control chart will assist process managers to determine the cost of quality.
  • 02:01 The final stage is the most dangerous.
  • 02:03 In this case the process is not under statistical control but so
  • 02:07 far you've been lucky in providing results that are conforming to your customer's
  • 02:11 expectations.
  • 02:13 This is a false sense of security, since it is not under control,
  • 02:16 the results are unpredictable.
  • 02:18 It may be good today, but who knows what the results will be tomorrow?
  • 02:22 SPC control charts are vital to this process.
  • 02:25 They will provide the alert to process managers when something needs to be done.
  • 02:31 So let's look at the business and
  • 02:32 process strategies that are supported with control charts.
  • 02:36 SPC control charts can be used throughout the improvement of a process.
  • 02:40 That means that we can use them in several phases of a Lean Six Sigma project.
  • 02:45 A control chart can be used to establish baseline process performance during
  • 02:48 the early data collection phase of an improvement project.
  • 02:52 In a Lean Six Sigma project, that means the measure phase.
  • 02:55 It can also be used to assist in the analysis of data to recognize special
  • 02:59 cause variation and
  • 03:00 common cause variation which leads to understanding of the root causes.
  • 03:04 This application of a control chart will be used during the analysis phase of
  • 03:08 a Lean Six Sigma project.
  • 03:10 Then during the improve phase of the project,
  • 03:13 SPC control chart could be used to illustrate the impact of a process change.
  • 03:18 This is often done by comparing the before and after process performance.
  • 03:22 And of course, control charts are used in the control phase
  • 03:25 as part of ongoing operations which we discussed in the previous slide.
  • 03:30 Control charts are not needed for every situation.
  • 03:33 During the improvement process, if there's a clear and
  • 03:35 obvious special cause, a control chart is not needed.
  • 03:39 Likewise, a proliferation of control charts for processed parameters
  • 03:43 that are not clinical can make the important ones less effective.
  • 03:47 Let's look at some of these in a little more detail.
  • 03:50 First, the measurement baseline.
  • 03:52 In this case the control charts are used to establish the current performance
  • 03:55 baseline that could be used to establish business expectations for a new product or
  • 04:00 process and clarify current performances before starting an improvement effort.
  • 04:05 The baseline demonstrates that the problem is stable or unstable.
  • 04:09 When it is stable there is virtually no special cause variation, and
  • 04:13 what there was has already been addressed.
  • 04:15 So now there's only common cause variation.
  • 04:18 And that level predictability can be established.
  • 04:21 What we're trying to do when using control charts in this manner
  • 04:24 is to ensure everyone understands the current state.
  • 04:27 This helps to manage expectations for process performance and sets the stage for
  • 04:31 possible process improvement.
  • 04:33 There's a realistic and reasonable understanding of the process mean and
  • 04:37 the common cause variation that we see illustrated in the control limits.
  • 04:41 Of course we can use the control limits during analysis to understand the sources
  • 04:45 of variation.
  • 04:47 Control charts are designed to reveal a special cause variation.
  • 04:51 The control chart shows exactly when an extreme point occurred.
  • 04:54 Or it can show when a change in the average level has occurred, or
  • 04:58 the beginning of a trend has occurred.
  • 05:00 But even beyond the special cause variation,
  • 05:02 control chart control limits, show a reasonable expectation for
  • 05:06 the process mean and the zone of common cause variation.
  • 05:09 If that variation is too large, or the mean is too high or
  • 05:13 too low, the process managers and designers know
  • 05:16 that a process design change is needed to impact the common cause variation.
  • 05:21 The control charts can also help an improvement team
  • 05:23 to not be tempted into tampering with the process.
  • 05:27 Once they see the normal variation, then if that variation exceeds the customer
  • 05:31 requirement, they know the fundamental process design change is needed.
  • 05:35 Not just a little tweak at the controls.
  • 05:37 Finally, we can demonstrate the impact of improvement activities
  • 05:41 through the use of control charts.
  • 05:43 A process may be under control, but the magnitude of the random common cause
  • 05:47 variation is still too large to meet the customer requirements.
  • 05:50 That process is in threshold performance and
  • 05:52 there is a desire to get to ideal performance.
  • 05:55 The process change is made and
  • 05:57 the impact of that change can be shown on the control chart.
  • 06:00 Often, there's a shift in the mean, or a change in the control limits.
  • 06:04 Using the control chart,
  • 06:05 in this way illustrates the business impact of the design change.
  • 06:09 One caution, at the time of that change the control chart will likely show
  • 06:14 an out of control condition.
  • 06:15 Well, this is expected.
  • 06:17 A special cause has occurred,
  • 06:19 you just change the process to implement an improvement.
  • 06:23 Control charts are not just for production operations,
  • 06:26 they can be used in many stages of the process life style.
  • 06:30 And many of the phases of an improvement project.

Lesson notes are only available for subscribers.

PMI, PMP, CAPM and PMBOK are registered marks of the Project Management Institute, Inc.

Gift this course