Locked lesson.

Upgrade

About this lesson

The process analysis of the value stream map will reveal quality issues that are occurring within process steps. These are identified by analyzing step data and process flow data.

Quick reference

Lean Quality Analysis

The process analysis of the value stream map will reveal quality issues that are occurring within the process steps. These are identified by analyzing step data and process flow data.

When to use

This analysis is normally done during the Analyze phase. By that time, the value stream map is complete including the data in the data boxes.

Instructions

The most commonly used quality analysis with a Lean value stream map is roll throughput yield. When working with process maps instead of value stream maps, I find the assessment of tangled flow to be the most effective. RE’s analysis technique can be used with either type of map.

  • Roll Throughput Yield (RTY) – This represents the probability that any given item will correctly complete each step in the process on the first pass – no rework or updates are needed. It is calculated by multiplying together the first pass yield of every step in the straight-line value stream process. That includes the steps with non-value-added time.
  • Tangled flow – The RTY cannot be easily calculated in a process map because there is too much uncertainty about which steps will be followed by any individual item. The branches and decision points lead to multiple options for flow. Even worse, there are often loops in the process – these are steps that must be repeated. Essentially, the process is flowing backward at that time, and often work must be first undone so that it can then be redone. This analysis is a visual assessment of the entire process map, looking for bottlenecks, loops, and danglers – which are steps that either have no preceding step or succeeding step. When there is no preceding step, the activity is a disrupter since there is no way of predicting when it will occur. If there is no succeeding step, the activity is waste since the step results are never used.
  • The RE’s analysis is the simplest Lean process analysis you can do, and is often very effective for identifying quality problems. Go through the process with the operators and determine where they do any activity that begins with the letters RE. (This technique may not work in other languages). A “RE” activity implies that the work must be corrected. Every place there is a “RE” is a quality opportunity. I have found that many times the “RE” is fully embedded into the standard process and people don’t even think of it as a problem until we do this analysis.

Hints & tips

  • When calculating RTY, always multiply the percentage of success (yield) not the percentage of failures.
  • If a loop in a process map cannot be easily eliminated, consider if there is a way to shorten the loop so that fewer process steps must be repeated. Check for the need for the loop sooner, or only go back to the minimum number of steps – not a complete do-over.
Login to download
  • 00:03 Hi, I'm Ray Sheen.
  • 00:06 Let's take another look at the Lean value Stream map,
  • 00:11 but this time I want to focus on how we analyze quality.
  • 00:15 Let's start with the most common process-level quality analysis and
  • 00:19 measurement, the role throughput yield.
  • 00:21 If you're working on a process improvement project with quality issues,
  • 00:25 you'll probably have a low roll throughput yield.
  • 00:28 This is normally the case when there is a high scrap or rework rate in the process.
  • 00:33 There are several ramifications of this problem.
  • 00:36 One is the uncertain delivery quantity and
  • 00:38 timing often leads to extended lead times to account for the delays when
  • 00:42 a unit is scrapped and another one must be completed to fulfill the customer request.
  • 00:47 And of course, there's the cost of fixing the problem.
  • 00:50 But there's also the cost of increased inspection or test to watch for
  • 00:54 the problem and increased supervision to manage the decisions that must be made to
  • 00:59 create workarounds when a critical unit is scrapped.
  • 01:02 To calculate a rolled throughput yield, we start with the step yield for
  • 01:06 each process step.
  • 01:07 This is normally one of the items in the data box and
  • 01:10 is often referred to as first pass yield.
  • 01:13 It is the frequency with which the operator or
  • 01:15 system completes the process step correctly on this first try.
  • 01:19 Each of these step level first pass yield values are multiplied
  • 01:22 together to get the roll throughput yield.
  • 01:25 The roll throughput yield then is the total process metric that represents
  • 01:29 the likelihood that when an item starts the process,
  • 01:32 it will be processed correctly, no rework or corrective actions.
  • 01:36 At each step of the process until it finishes the process,
  • 01:40 it goes all the way through error free.
  • 01:43 It's very common for this number to be well below 50%,
  • 01:47 meaning most items are reworked at least once at some point in the process.
  • 01:51 Roll throughput yield works well with a value stream map, but not process maps.
  • 01:56 With these maps, look for loops and danglers.
  • 01:59 A value stream map should always have straight line flow, so
  • 02:02 there are no loops or danglers.
  • 02:04 However, you may be working with a process map that shows branches and decisions.
  • 02:09 In this case, I recommend analyzing these maps for loops and danglers.
  • 02:13 Loops are just that, a loop in the process flow.
  • 02:17 On the map, it normally shown as backwards flow, and then repeating steps.
  • 02:22 These are quality issues requiring rework in the process.
  • 02:25 Of course, a major question with loops is, how many times will it loop?
  • 02:29 Each loop is adding cost and time to the process item, but not adding any value.
  • 02:34 An improvement goal will be to eliminate or
  • 02:37 shorten all of the loops in the process.
  • 02:40 Danglers are any process step that is not connected with at least one incoming
  • 02:45 arrow and one outgoing arrow.
  • 02:47 Danglers that do not have a preceding arrow are unpredictable disruptors in
  • 02:52 delays.
  • 02:52 The step is initiated by something totally outside of the control or
  • 02:56 cognizance of the process operators in the process control, therefore,
  • 03:01 its start cannot be planned.
  • 03:03 That means that sometimes it may occur and distract or
  • 03:06 disrupt the team, but at other times, the teams will need a result from that step or
  • 03:10 a step in the path that was started by that step.
  • 03:13 And the process must wait until the unpredictable and
  • 03:16 uncontrollable event occurs.
  • 03:18 Danglers that don't have a succeeding arrow that goes to the next step
  • 03:21 are a waste of effort.
  • 03:22 If a step results are not used by another step in the process and
  • 03:26 are not generated as a process output, then why bother with them?
  • 03:31 Often, these are legacies from an older process and
  • 03:34 can be eliminated with no problem.
  • 03:36 The final Lean quality analysis I would like to introduce is the Re's analysis.
  • 03:41 A Re is any action in a process step or
  • 03:44 in your process that begins with the letters RE.
  • 03:49 Let's look at a few examples of what I mean.
  • 03:51 Rework, recall, remeasure, reship, rewrite.
  • 03:54 Well, you see what I mean here.
  • 03:55 The fact that you must redo something means that you didn't do it correctly
  • 03:59 the first time.
  • 04:00 I'm careful now when I describe this technique.
  • 04:03 I used to say that anything in your process that started with an RE is waste
  • 04:07 and should be eliminated.
  • 04:09 Until I was in a room one time where everyone turned and
  • 04:11 looked at an individual and said, Rebecca.
  • 04:14 So now I'm careful to say that it's any action that starts with the RE
  • 04:19 letters should be eliminated, it's because Res are failure points.
  • 04:24 Re steps in your process are essentially planned failure.
  • 04:27 The previous failure rate has been so high that now we just plan on it happening, so
  • 04:32 we plan the step and to fix it.
  • 04:33 Since we never got to the underlying root cause of a problem to eliminate it,
  • 04:38 we now have to assume that everything will initially have the problem and
  • 04:42 we need to rework it.
  • 04:43 A great example of this is when a process introduces a reinspection or
  • 04:48 retest because of a former problem, and it then becomes the new standard process.
  • 04:53 You can assume that everything in a Re step is non-value added item.
  • 04:57 Yes, if there's problem, we need to fix it, but
  • 05:00 there should not have been a problem to begin with.
  • 05:03 The value-added time was the first time the activity was attempted in the process.
  • 05:07 Any repeated attempts are non-value-added.
  • 05:10 The Re step may also be fixing a problem that occurred earlier in the process.
  • 05:15 If you don't think that you are able to fix the root cause,
  • 05:19 consider whether the RE step can be moved closer to the problem point,
  • 05:23 then the amount of reactivities would be minimized.
  • 05:26 What I mean is that if we can't stop the problem from happening,
  • 05:30 we should at least fix it as quickly as we can.
  • 05:33 So there are quantitative, graphical, and even grammatical
  • 05:37 techniques that can be used to analyze your process flow for quality.

Lesson notes are only available for subscribers.

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

Gift this course