Locked lesson.

Upgrade

About this lesson

A key element of the control plan is the response plan. It tells the process operators what to do when the process begins to deviate from the parameters laid out in the control plan.

Exercise files

Download this lesson’s related exercise files.

Response Plan Exercise.docx
61.2 KB Response Plan Solution.docx
61.7 KB

Quick reference

Response Plan

A key element of the Control Plan is the Response Plan.  This tells the process operators what to do when the process begins to deviate from the parameters laid out in the Control Plan.

When to use

The response plan is a portion of a control plan.  When the control plan is prepared in the Control phase for release with the implementation of the new process, the response plan must be a part of it. The Response plan goes into effect when measures from the Monitoring Plan indicate a need for a response.

Instructions

The response plan is the third part of a control plan.  The first two are the process documentation and the monitoring plan that explains what is to be measured, the target and limits, the responsibility for measurement, and the process for collecting the measurement information and recording the data.  The response plan provides guidance to the operator on what to do when the data indicates that action is needed.

There are four process states and each requires the support of a control plan.

  • The ideal state is that a process is in statistical control and delivering conforming results.  The control plan should keep the process operating in this fashion.
  • The threshold state is that the process is in statistical control but the process capability is poor, so some nonconforming results are produced.  The process is always at the threshold of very poor performance. The control plan is essential to both keep the process centered and to provide guidance for how to disposition the non-conformances.
  • The poor performance state is a process that is not in statistical control and is creating non-conforming results.  When this is the case, the Lean Six Sigma project is not complete and the team needs to find solutions before considering a control plan.
  • The false sense of security state is when a process is not in statistical control, but is creating conforming results.  The process manager and operators may think that all is OK, but due to the presence of special cause variation, the process is not predictable and may suddenly start to create non-conforming results.  Keep in mind that special cause variation can create a special condition where everything looks good, but the process cannot sustain that.  A control plan is needed to recognize this situation and provide guidance to the managers and operators concerning what they should monitor and how they should respond to this.

There are five types of responses.  A typical response plan will contain a combination of these, using the appropriate response based on the severity of the characteristic and the nature of the non-conformance.  The five approaches are:

  • Increased monitoring: This approach continues the process but with additional oversight.  This is appropriate when alarm thresholds have been reached.
  • Manual override: This approach is used for automated processes. The automation is stopped and the process is run manually in order to exert a little more caution.
  • Stop, correct, and resume: This approach is used when the process failure is clearly identified and understood – such as maintenance and calibration failures. The operator follows the corrective action procedure and restarts the process.
  • Follow special instructions: When out-of-control parameters have been identified follow instructions that were put in place for that condition.
  • Stop and escalate: When out-of-control parameters have been identified and there are no special instructions, stop the process and escalate it to the experts.

Hints & tips

  • Create a response for every measure and out-of-control condition.  You don’t want the operators to guess what to do.  Either they will do nothing or will start to tamper with the process and possibly make things worse.
  • If using the “escalate” option, the subject matter experts need to respond quickly or the process management will likely override and ignore the control plan.
  • The process state of the original process was determined during the Measure phase. Determine the process state of the new process with early process operation. However, be aware of the danger of the Hawthorne effect. When it is new everyone is paying special attention and it has not yet become routine.
  • The Response plan actions could be a tiered response. For instance, the first tier may be a manual override and if that doesn’t work the second tier is to stop and escalate.
Login to download
  • 00:05 Hi, I'm Ray Sheen.
  • 00:06 I saved one aspect of the control plan for its own special module, and
  • 00:10 that's because it is so important to the success of the project.
  • 00:14 That's the response plan.
  • 00:16 And just to be clear, this is a subset of the overall process control plan.
  • 00:21 I will start by considering the control strategy
  • 00:24 based upon your process performance.
  • 00:27 There are four different strategies you could pursue,
  • 00:30 these strategies are based upon the typical performance of your process.
  • 00:34 It could be in statistical control, meaning there are no special cause
  • 00:38 variation, only comic cause, or it could also be out of statistical control,
  • 00:43 meaning that there is the presence of special cause variation.
  • 00:47 Second, your process may have been producing conforming parts, or
  • 00:52 it may have been creating some errors and defects.
  • 00:55 Now, it should be no surprise that our goal is to have
  • 00:58 a process that is statistical control and always producing conforming items.
  • 01:03 This is the ideal state, we should try to design a solution that gets there and
  • 01:08 then put in place a control plan to stay there.
  • 01:11 Sometimes the process is in control, but there are some non-conforming results.
  • 01:17 This is a process capability issue.
  • 01:20 The, Specification limits are tighter than what the process is able to consistently
  • 01:24 achieve.
  • 01:25 We call this threshold performance because we are on the threshold of being able to
  • 01:30 get to the ideal state.
  • 01:31 From a control strategy standpoint, we want to be maintaining control and
  • 01:36 constantly monitoring for
  • 01:37 those instances when the results are not conforming so they can be reworked.
  • 01:42 This is also a candidate for a further Lean Six Sigma project.
  • 01:46 The third strategy is for the case when the process is out of control and
  • 01:51 producing parts that are non-conforming.
  • 01:53 In this case, the process is broken and typically everyone knows that there
  • 01:58 are ongoing problems and chaos as evident at times.
  • 02:02 You should not be in this state at the end of a Lean Six Sigma project,
  • 02:06 although this is often the case at the beginning of a project.
  • 02:09 If your team is still in this condition, your project is not over yet.
  • 02:15 The fourth condition is actually the most dangerous one.
  • 02:18 The process is not in control, but it has been producing conforming results lately,
  • 02:24 that means that some special cause is overriding the normal
  • 02:27 process performance and making sure that the result is acceptable.
  • 02:32 This condition can lead to a false sense of security and also,
  • 02:36 it's difficult to create a control plan for this condition because the process
  • 02:40 does not have the inherent ability to control what's happening.
  • 02:44 It is the special cause that's overriding everything in the process, and
  • 02:49 as soon as that special cause changes, the results will likely be disastrous.
  • 02:54 So with the strategy in mind,
  • 02:56 now let's look at how the response plan fits with the control plan.
  • 02:59 For starters, the response plan lets the process operators know which inputs and
  • 03:04 outputs are critical and which are not.
  • 03:06 Based upon the type of response and nature of the response,
  • 03:10 the operators will have a sense of what is most important.
  • 03:13 When a defect is detected, and that could be an output defect or an input defect,
  • 03:18 the response plan tells the operator and the managers what to do.
  • 03:23 First, what to do with respect to notifying the customers of the process and
  • 03:28 second, what is to be done within the process, both are important.
  • 03:32 In some cases, the response plan will prescribe a response even
  • 03:37 though a defect is not occurred yet.
  • 03:39 This normally happens when a process is starting to degrade, for instance,
  • 03:44 a run-chart on an input may indicate the presence of a new special cause variation
  • 03:49 with a shift in the input values.
  • 03:51 The magnitude of the shift may not be enough to create a defective output, but
  • 03:55 the fact that a special cause variation has been detected is an indication of
  • 03:59 the process degradation.
  • 04:01 And the response plan will likely give directions to the operator as to what they
  • 04:05 should do.
  • 04:06 Sometimes the process is directed to do something that is not part of the standard
  • 04:10 process.
  • 04:11 There's a custom order or a special case that must be run.
  • 04:14 The response plan provides directions to the operator as to how they should
  • 04:19 respond to the special request on unusual situation.
  • 04:22 It isn't that it's necessarily a bad problem, it's just different
  • 04:27 from the normal process and therefore requires special attention.
  • 04:31 So let's review different types of responses.
  • 04:34 There are a number of factors that should be considered when determining what
  • 04:39 an appropriate response would be to a situation.
  • 04:42 These factors include the characteristic of the process and the process parameters,
  • 04:47 the magnitude of the out-of-control or unusual situation, and the breadth and
  • 04:52 depth of information that is needed to respond to the situation.
  • 04:56 One type of response is to increase the monitoring of the parameter in question.
  • 05:00 In this case, the process is continued but with extra oversight,
  • 05:04 there may be increased testing or increased frequency of sampling.
  • 05:08 This is often done when alarm limits are hit on a parameter,
  • 05:12 the output is not yet out of specification, but it's getting close.
  • 05:16 Another response is the manual override, this is often used with automated
  • 05:21 equipment that has identified a possible defect or error.
  • 05:25 An operator inspector or manager reviews the defect and
  • 05:28 determines if it's a valid defect or if the item can be continued in the process.
  • 05:33 I often find this approach is used when a process has to handle custom orders or
  • 05:38 custom items.
  • 05:39 There is no way for the automation to know all the possible options in a custom job,
  • 05:44 so the automation flags it as a possible error and someone must respond to it and
  • 05:48 make a determination.
  • 05:50 The third one is probably the easiest one to understand, stop, correct and resume.
  • 05:56 This is often the response used for setup errors or manual assembly errors.
  • 06:01 This is the approach used for
  • 06:02 items that would be discussed in a troubleshooting guide for the operator.
  • 06:06 When they have a specific type of problem, they should stop,
  • 06:09 take a specific corrective action and then they can resume the process.
  • 06:13 It can be a great approach but it requires a process designers to have already
  • 06:18 identified the possible defect, it doesn't work for brand new problems.
  • 06:22 The fourth response could be considered a variation on the third response, again,
  • 06:28 the operator stops work, but now they begin to follow a specific procedure.
  • 06:32 This is normally used if the problem or
  • 06:35 error is not a quick fix that can be easily corrected and the process resume.
  • 06:39 This is often used with items that are very expensive or
  • 06:42 that are a sensitive item with respect to external stakeholders.
  • 06:46 In those cases, we normally don't want to throw anything away and start over,
  • 06:51 so there are special instructions for what to do in those special situations.
  • 06:56 The fifth response approach is a fallback approach,
  • 06:59 when you've not been able to prepare one or the other four.
  • 07:03 The operators responses to stop and ask for help, bring in an expert and
  • 07:07 let them determine what is happening.
  • 07:10 I recommend this response for new process introduction on critical processes until
  • 07:14 the process can become stable.
  • 07:16 When there is a problem,
  • 07:18 I want to subject matter experts to be leading the investigation.
  • 07:22 Most response plans will have a combination of these approaches and
  • 07:25 I usually provide the response with monitoring information.
  • 07:28 So the operators don't need to be referencing multiple procedures or
  • 07:32 documents.
  • 07:33 A clear response plan will make it much easier for the process manager and
  • 07:37 the operators to deal with the inevitable problems that will come up.
  • 07:42 Don't neglect this.

Lesson notes are only available for subscribers.

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

Gift this course