Control Phase of Lean Six Sigma Black Belt Tutorial

6.2 Section VI Control

Hello and welcome to section six of Six Sigma Black Belt training program offered by Simplilearn. In this section, we will deal with the control phase. Being the last phase of the Six Sigma DMAIC project, the control phase, in real life, is considered to be the easiest provided the previous phases have been completed successfully. Before we can move ahead, please keep your control phase tools ready, which has been provided to you as ‘control phase toolkit’. In the next slide, we will discuss the agenda of this module.

6.3 Agenda

In this section, we will start with pre-control considerations. Then, we will see how to use variable and attribute control charts. Next, we will deal with measurement system analysis, control plan, and project closure. Finally, we will be introduced to total productive maintenance. In the next slide we will look into pre-control considerations.

6.4 SectionVI Lesson 1 Pre Control Considerations

In this lesson we will look into some considerations before the Control phase. We will look into the agenda of pre-control considerations in the next slide.

6.5 Agenda

In this lesson we will understand the pre-control considerations followed by assessing the results of process improvement. We will end this lesson by doing a rational subgrouping.

6.6 Pre Control Considerations

Some of the pre-control considerations are as follows: A Six Sigma DMAIC project should move into the control stage once the improvements have been validated in the improve phase. If all improvements have failed, the Black Belt can decide to move the project as a reengineering or a DFSS project. The process must exhibit a state of statistical control by the Improve phase. Statistical validations of solutions must result in an increase in the process capability or the performance of any other business measures. All possible risks are understood, and if any potential risks are found that could harm the process, preventive measures should be in place. Post the toll-gate meeting, once the Black Belt and Champion agree to the improvement measures, the project can move into the Control stage.

6.7 Assessing the Results of Process Improvement

In this slide, we will find out how to assess the results of process improvements. It is very important to understand and differentiate changes that are observed due to improvement compared to the inherent variation in the system. These inherent variations could have been caused due to the common cause or special cause in the process. The results of process improvements are assessed as follows: To begin with, first compare the improvement measure data with the baseline data and see if there is any noticeable change. After that, we can use appropriate formulas to compare the ‘before and after’ performance of the process. Once we have validated these, we can move on to compare moving average, median, spread, defect percentage, efficiency, waste, etc. This will give adequate signals around the change to validate and confirm that a positive change has indeed happened, and is noticeable. After all these, we need to continue to monitor the data for considerable number of days to ensure the improvements are sustainable over a period of time.

6.8 Rational Subgrouping

In this slide, we will understand rational subgrouping technique. Rational subgrouping is the name given to a technique in which data are organized into subgroups for process control charts. While doing rational subgrouping, it is important to make sure that the subgroups are chosen so that within a subgroup, variation is minimized. Selection of rational subgroups should be consistent with the structure of the data from the process. Also, the selection of rational subgroups should allow for quick identification of potential corrective actions once an out-of-control condition is identified.

6.9 Summary

In this lesson we have learned the pre-control considerations, and understanding how to assess the results of process improvement. Lastly, we discussed how to do rational subgrouping.

6.10 Section VI Lesson 2 Variables and Attributes Control Charts

Let us now look into the main lesson in the control phase. This lesson is on the variables and attributes control charts. As mentioned before, if all the previous phases have been successfully completed, the control phase would deal with charting the process and coming up with a control plan. In the next slide, we will cover the agenda of this lesson.

6.11 Agenda

The agenda for this section includes the concepts of variables control charts, variables control charts, EWMA charts, Cusum charts, and attribute control charts. In the next slide, we will discuss the concepts of variables control charts.

6.12 Concepts of Variables Control Charts

Drawing charts is not a tough task with the software packages like QI Macros SIGMA XL, but before doing so one should be clear on some fundamentals. They are as follows: Variables control charts should be used with the input and output variable data, and both the variables must exhibit a state of statistical control. Xbar – R, also known as Mean – Range chart, should be plotted if the decided subgroup sample size is between two and nine. Xbar – S, also known as Mean – Sigma chart, should be plotted if the decided subgroup sample size is greater than ten. I – MR chart should be plotted, if the decided subgroup sample size is one. To plot IMR chart, data must follow a normal distribution. This is an important consideration because a lot of people draw I – MR charts for non-normal data, and end up realizing that the chart is not correct. This happens because they decided to do an I – MR Chart for the non-normal data. Xbar – R and Xbar – S charts are not dependent on the normal data and control limits. Basis of control charts are the rational subgroups. Rational subgroups mean subgroup of items that are produced under the same conditions. Often the chart may also fail if the samples are not rational subgroups. In the next slide, we will continue this discussion on the concepts of variables controls charts.

6.13 Concepts of Variables Control Charts(Contd.)

Control limits are calculated so that the probability of having an average or range falling outside the control limits is rare. The reason it is rare is that if done properly, the probability of a point falling outside the control limits is zero point two seven percent, that is, twenty seven out of ten thousand points. Control limits of all the variables charts are robust to non-normality. That means, except for the I – MR charts, the control charts work even on non-normal data. Moderate departures from normality are accepted for drawing the range and sigma control charts. Now let us understand the control limits for the range charts. The Lower Control Limit (also called as LCL) is equal to D three multiplied by R bar. The Upper Control Limit (also called as UCL) is equal to D four multiplied by R bar. Sigma value is equal to R bar divided by D four, where R bar is the average of all the ranges, and D3 and D4 are chosen on the basis of the subgroup sample size. D three and D four are taken from the control charts constants table, provided as part of the toolkit. We will further discuss the concepts of variables controls charts in the next slide.

6.14 Concepts of Variables Control Charts (Contd.)

Let us discuss the control limits for average charts using R bar. LCL is X double bar minus A two multiplied by R bar. UCL is X double bar plus A two multiplied by R bar. Basically X double bar means mean of means or average of averages. Technically though, control limits are considered to be spaced at three standard deviations from the center line, considering that the probability of finding the special causes of variation is zero point two seven per cent. Knowing the control chart constant formulas may not have a lot of practical significance, as specially designed computer packages do this calculation automatically.

6.15 Concepts of Variables Control Charts (Contd.)

In this slide, let us understand the control chart formulas with the help of an example. For a given R bar of three, a subgroup sample size of five, and the ‘average of averages’ as ten, calculate the control limits for the averages and ranges chart. The solution is as follows. For range chart, LCL is D three multiplied by R bar. The value of D three for a sample size of five is zero. Thus, the LCL for the range chart is zero. UCL is D four multiplied by R bar. The value of D four for a sample size of five is two point one one four. Multiplying this with three, gives six point three four two, which is the UCL of the range chart. Thus the control limits for range chart are 0 and six point three four two.

6.16 Concepts of Variables Control Charts (Contd.)

In this slide, we will discuss the control limits for averages chart. Here, LCL is X double bar minus A two multiplied by R bar, and UCL is X double bar plus A two multiplied by R bar. According to the control charts constants table, A two is zero point five seven seven. Using the data in the formula, we can conclude that LCL is ten minus zero point five seven seven multiplied by three, which gives the value eight point two six. Thus, eight point two six is the LCL here. On the other hand, UCL is ten plus zero point five seven seven multiplied by three. This gives us the value eleven point seven three. Thus, eleven point seven three is the UCL here. Therefore, we can say that the control limits for averages chart is eight point two six and eleven point seven three.

6.17 Variables Control Charts

Let us discuss variables control charts in this slide. On this slide, we see a graph that plots average values on the control chart. Interpreting this control chart, it is found that the process is still not in statistical control. The first set of 6 data points are all in the lower section of the graph and the rest 4 are towards the end of the graph. Over a short run, the variations may be in control, but predicting long-term variations with such a process may not be easy. In the control phase and in the days to come, adequate efforts need to be taken to fix the defaulting points.

6.18 Variables Control Charts(Contd.)

On this slide, we see a graph that plots ranges on the control chart. Interpreting this control chart, we observe that the process is still not in statistical control. There are 8 values that are showing a steady one way in the increasing order. The last 2 values have a drastic difference, and the last value goes out of the control limits. Over a short run, the variations may be in control, but predicting any long-term variation with such a process may not be easy. In the control phase and in the days to come, adequate efforts need to be taken to fix the defaulting points.

6.19 Variables Control Charts(Contd.)

Let us now interpret the charts. Interpreting both the charts, the average and the range chart, it is found that the process is still not in statistical control. Over a short run, the variations may be in control, but predicting long-term variations with such a process may not be easy. In the control phase, and in the days to come, adequate efforts need to be taken to fix the defaulting points.

6.20 Variables Control Charts (Contd.)

Now, let us look at some formulas for calculating the control limits for a Sigma chart. For finding the control limits of a Sigma chart, we need to keep the following factors in mind use sample standard deviation formula to compute Sigma value of subgroups; S bar is sum of subgroup Sigmas divided by Number of subgroups; LCL is B three multiplied by s bar; and UCL is B four multiplied by s bar. Control limits for Averages chart are: X double bar is sum of subgroup averages divided by number of subgroups LCL is X double bar minus A three multiplied by R bar UCL is X double bar plus A three multiplied by R bar Let us look at variable control charts in the next slide.

6.21 Variables Control Charts (Contd.)

On this slide, we see a graph that plots averages (that is X-Bar) on the control chart. Interpreting this X-Bar control chart, we observe that the process is still not in statistical control. There are 3 values out of 10 that are outside the control limits. Over a short run, the variations may be in control, but predicting long-term variations with such a process may not be easy. In the control phase and in the days to come, adequate efforts need to be taken to fix the defaulting points.

6.22 Variables Control Charts(Contd.)

On this slide, we can see a graph that plots sigma on the control chart. Interpreting this sigma chart, we observe that the process is still not in statistical control. Over a short run, the variations may be in control, but predicting long-term variations with such a process may not be easy. In the control phase and in the days to come, adequate efforts need to be taken to fix the defaulting points.

6.23 Variables Control Charts(Contd.)

In this slide let us look into the interpretations. Interpreting both the Average and the Sigma chart it is found that the process is still not in statistical control. Over a short run, the variations may be in control, but predicting long-term variations with such a process may not be easy. In the control phase and in days to come, adequate efforts need to be taken to fix the defaulting points.

6.24 Variables Control Charts(Contd.)

Individuals Moving Range Charts are preferred over subgroup charts like Xbar – R and Xbar – S because of the following reasons. One is that the observations may at times be expensive to get. Another reason is the output of a process may be homogeneous over a short period of time, for example pH of a solution. Production rate may be slow, this becomes the third and finally, the interval between successive observations is long. In a batch-process and in a typical ‘chemicals manufacturing’ process, within group variability is too small compared to the variability between other groups. In such a scenario, using sub-group charts is not a logical option. Let us look into an individual moving range chart on a sample data in the next slide.

6.25 Variables Control Charts(Contd.)

On this slide, we can see the first chart of the I-MR charts. This chart plots individual data points on the chart. From the chart, one can make out that the individuals chart is out of control, showing evidence of possible special cause of variation. There is a specific pattern for the data and is getting repeated at a regular interval. There are too many points going out of control at a regular interval. Special causes don’t occur frequently in a pattern, these have to be investigated further.

6.26 Variables Control Charts(Contd.)

On this slide, we can see the second chart of I-MR charts. This chart plots the oving Range data points on the chart. Contrary to what we saw on the Individual chart that was out of control, this moving-range chart is on statistical control. The data points are fairly random across both the sides of the central control limit. Though it does appear that the data points are oscillating at the first glance, a close look confirms that the data distribution is fairly random.

6.27 Variables Control Charts(Contd.)

Let us now interpret the individuals moving range chart. The Moving Range Chart is in control, while the Individuals chart is out of control showing evidence of possible special causes of variation. There are too many points out of control at a regular interval. Special causes don’t occur frequently in a pattern, the data points that are out of control need to be investigated further. Basic assumption behind doing an IMR Chart is that the data should be normal. We will look into the control chart patterns in the next slide.

6.28 Variables Control Charts(Contd.)

Control charts provide an operational definition for the special causes. One should not be confused with the references to the assignable causes, as both are similar. Special causes, most of the times can be assigned to root causes. Generally, these special causes are defined as any points that cause an observation to go out of the control limits. Most of the times, one may find points going outside of the control limits. These are due to special causes. In addition to looking for points going outside the control limits, the Black Belt must investigate the drifts, freak patterns, and cycles. These patterns indicate the presence of non-special causes. Computer packages will help the Black Belt in detecting such points. In the next slide we will discuss freak patterns and drifts.

6.29 Variables Control Charts(Contd.)

Let us first discuss the freak patterns. Freak patterns happen once in a while and are generally due to special causes. Freak Patterns show up as infrequent occurrences on the control charts, i.e., points moving out of the control limits. Now let us understand what drifts are? Drifts happen when the current state of process is determined partly by its past state. For example, wear and tear of a tool. This is not something that comes instantly. When the reasons for drifts are identified, actions should be taken to eliminate them. Let us move on to understand cycles and additional problems in the next slide.

6.30 Variables Control Charts(Contd.)

What are cycles? Cycles are often identified as oscillatory patterns in a control chart. A variables chart may not have any points outside the control limits and yet, have special causes in it. The chart, for example, may be range bound within the control limits and still exhibit frequent ups and downs. These may not be random variation after all. Cycles occur due to the nature of the process. Most of the times, they occur due to the periodic changes in the process, or the inputs. For example, such changes are observed when the process changes hourly or periodically or when the inputs are changed deliberately. On identification of cycles, adjustment can be done by changing the periods, in order to smoothen out the graph. Now let us understand the additional problems. When plotting a data set, one can expect to find different values. Having repetitive values in the data set could be because of inadequate gage resolution. We will discuss the different actions on control charts in the following slide.

6.31 Variables Control Charts (Contd.)

Merely plotting a control chart is not enough. A control chart will tell if special causes of variations exist in a process or not. Let us now discuss the actions on control charts. Once control charts are plotted, differentiation should be done between caused (special cause) and un-caused variation (common cause). If special causes are found, cause and effect diagram must be correlated with the control charts to find out what caused the observation to go out of control. Remedial actions must be taken to eliminate this special cause of variation, if found undesirable. Common causes or un-caused variation should be reduced as much as possible, evidence of which can be found from the CE Diagram. The main goal here is the statistical association and not the causal correlation. After control charts, let us understand the EWMA charts in the next slide.

6.32 EWMA Charts

EWMA charts stands for Exponentially Weighted Moving Average Charts. This chart plots the moving averages of data and assigns exponentially decreasing weights. It is helpful in detecting small process shifts as opposed to the conventional variables charts that help in detecting large shifts due to special causes of variations. The control limits namely, LCL and UCL, of an EWMA chart are not mandatory. Some practitioners may decide to skip drawing the control limits on an EWMA chart. When fixed, the control limits vary exponentially in an EWMA chart. On EWMA charts, any points outside the control limit do not mean that it has a special cause of variation. In a conventional control chart that would be the typical case. These chart templates generated with QI Macros are found in Process Shifts charts in the control toolkit. Let us continue our discussion on EWMA chart in the next slide.

6.33 EWMA Charts(Contd.)

An EWMA chart has been generated using the sample data as seen on the slide. As given in the chart, the control limits are not in a straight line. They follow more of an exponential function. Additionally, the five points at a stretch, indicated specially with the help of a red arrow, indicates an unstable region or an out of control region. We will move on to Cusum charts in the following slide.

6.34 Cusum Charts

A Cusum chart has been shown on the slide. They are another type of charts that detect small process shifts. This chart is important to a Black Belt. In real life applications one need not worry about knowing the concepts or how to draw these charts.

6.35 Attribute Control Charts

After the discussion on variables control charts, let us move on to discuss attribute control charts. Attribute control charts are used with discrete data like defects and defectives. It does not deal with data that vary from time to time on a continuous scale. Let us look into the table given in the slide. To understand the proportion of defectives one must use the p chart. The process standard deviation denoted by sigma, when using a p chart, is given by the formula, square root of p bar multiplied by one minus p bar divided by n. To understand the count number of defectives the np chart must be used. The process sigma is calculated by taking a square root of the number obtained by taking a product of n and pbar (pronounced as as "P" Bar") and one minus pbar (pronounced as "P" Bar"). To understand the defects per unit, the c chart must be used. The formula to calculate process sigma is the square root of cbar (read as "C" Bar"). And, finally to understand the average defects per unit the u chart must be used. The process sigma is calculated by taking square root of the number obtained by dividing ubar (pronounced as "U" bar) divided by n. Let us look into the control limits of p and np chart in the next slide.

6.36 Attribute Control Charts(Contd.)

Let us first look into the control limits for the p Charts. Here, LCL is p bar minus three multiplied by square root of p bar multiplied by one minus p bar divided by n. UCL is p bar plus three multiplied by square root of p bar multiplied by one minus p bar divided by n. p chart is used for varying sample size and when one wishes to find out proportion defectives. Control limits for np charts are as follows. LCL is np bar minus three multiplied by square root of np bar multiplied by one minus p bar. UCL is np bar plus three multiplied by square root of np bar multiplied by one minus p bar. Where pbar is found by dividing np bar by n, where n is the number of subgroups. Next, we will discuss the control limits of u and c chart.

6.37 Attribute Control Charts(Contd.)

Control Limits of a u – Chart are as follows: LCL is u bar minus three multiplied by square root of u bar divided by n. UCL is u bar plus three multiplied by square root of u bar divided by n. Use u – chart if the sample size is varying. Control Limits of a c – Chart are as follows: LCL is c bar minus three multiplied by square root of c bar. UCL is c bar plus three multiplied by square root of c bar, where c bar is the sum of subgroup occurrences divided by the number of subgroups. Let us look into a p chart plotted using sample data in the next slide.

6.38 Attribute Control Charts (Contd.)

From the toolkit, one can use the sample data and plot a p chart. Here is an example showing how it will look like. One can observe that the proportion defective from the time period 3 has been consistently growing and moving in a fairly single direction, increasing the proportion defective from 0.1 to around 0.23. We will discuss an np chart plotted using the sample data in the following slide.

6.39 Attribute Control Charts(Contd.)

From the toolkit, one can use the sample data and plot an np chart. Here is an example. One can observe that the ‘defective peaches’ ranges from 20.2 to 29.5, and are on both the sides of central control limit. A specific observation on the X-axis (represented by date, time, or period) is that between 6 and 7 there is the biggest jump in the defective peaches from 20.2 to 29.5 defectives. Let us move on to plot the u chart with the sample data in the next slide.

6.40 Attribute Control Charts(Contd.)

From the toolkit, one can use the sample data and plot an n chart. We can see an example on the slide. One can observe that the defects ranges from 0.08 to 0.18, and a majority of them are on the top side of the central control limit with values 0.15 or higher. A specific observation is that there is a specific pattern that is repeated continuously after every 4 range points on the X-axis. We will conclude with plotting c chart with sample data in the following slide.

6.41 Attribute Control Charts(Contd.)

From the toolkit, one can use the sample data and plot a c chart for defects. Let us look into an example in this slide. One can observe that the defects ranges from 2 to 4, and almost half of them are with the value of 3 just above the central control limit of 2.92. A specific observation is that there is a specific pattern that has been repeated continuously after every 4 range points on the X-axis. Let us summarize what we have learned so far in the next slide.

6.42 Summary

In this lesson we have learned the concepts of attribute and variables charts, Xbar R, Xbar S, I – MR, EWMA, and different charts namely, p, np, u, and c chart. In addition, we have also learned how to calculate the control limits for different charts manually.

6.43 VI Lesson 3 Measurement System Analysis Control Plan and Project Closure

In this lesson we will discuss the measurement system analysis, control plan and the project closure. These are some of the important things to do in a Control phase.

6.44 Agenda

In this lesson, we are going to cover details on the Measurement Systems Analysis followed by understanding how to create and use a Control Plan. Finally, we will discuss the activities to be done in a Project Closure.

6.45 Measurement System Analysis

In this slide we will learn the importance of Measurement System Analysis (MSA). This activity is discussed in the Control phase as it is important for this phase. Measurement system analysis is done in the Measure phase to validate the measurement system, and check if the data to be collected is valid and reliable. MSA in the Measure phase is also done to re-compute the baseline data. In the control phase, measurement system analysis is done to check if the measurement system has adequate resolution to measure the data with reduced variability. MSA re-analysis is important because in the long run, the measurement system must be adequate and effective enough to measure the reduced variability data seamlessly. We will look into control plan in the next slide.

6.46 Control Plan

Control plan is the official document and also, a part of the knowledge transfer mechanism across different teams. This plan has to be updated by the Six Sigma team, checked by the Black Belt, and approved by the project Champion. Control plan will also provide details on document archiving, and should be archived along with the other documents. We will continue our discussion on control plan in the next slide.

6.47 Control Plan(Contd.)

A process control plan assures that a well thought-out reaction plans are in place, in case an out-of-control condition occurs. It also provides a central vehicle for documentation and communication of control methods. Special attention is typically given to the potential failures with high RPNs and those characteristics that are critical to the customer. Here, RPN is the risk priority number, which is calculated as a product of risk impact, risk occurrence probability, and detectability of the risk. A control plan deals with the same information explored in an FMEA (pronounced as F-M-E-A) plus more. The major additions to the FMEA that are needed to develop a control plan are identification of the control factors, the specifications and tolerances, the measurement system, sample size, sample frequency, the control method, and the reaction plan. In the next slide, we will find out where to use a control plan.

6.48 Control Plan(Contd.)

Now that we have covered FMEA in details, let us move on to its usage. The primary question here is where to use it. It is used to ensure that the problem solutions are permanently effective. There are three questions that are to be addressed for this. The questions are as follows. What has been done to prevent the process problems? How is it known when problems occurs, and what will be done when problems in fact do occur? What are the written descriptions of the systems for controlling parts and processes? The control plan to be updated is attached as part of the toolkit.

6.49 Control Plan(Contd.)

Here is a snapshot of the control plan template. A table is shown on the slide that needs to be filled out for a complete and effective control plan. The first item in the table is Process Step. Here, list down each of the process step where the control plan is applicable. For each of the process step, mention the input and output variables, and the parameters. After this, in the column for the Process Spec, lower specification limit, upper specification limit, and the target need to be captured. Corresponding Process capability (that is C-p-k date for sample size) have to be captured in the next column. Followed by this are the Measurement system and percentage R and R for the same. After this, the information from FMEA for current control method needs to be documented, along with the information like who will do this, where, and when. The last but not least is the reaction plan, where we document the reactive steps to be taken in case something goes wrong.

6.50 Project Closure

Finally let us understand the Project Closure. A project is considered closed only after: an increase in capability and performance indices is noted; process remains in statistical control future risks and preventive solutions are identified; corresponding reduction in DPMO levels are noted; increase in bench Sigma levels is noted; RPN reduces for failure modes; project storyboard and other documentation are completed and approved; and finally, control plan is complete, checked, and approved. Let us summarize what we have learned so far in the next slide.

6.51 Summary

In this lesson we have learned how to do another round of measurement system analysis to validate and confirm whether the measurement system is still accurate. Then, we reviewed the process for Control Plan and Document Archiving. And at the end, we looked at the Project Closure and reviewed the last activities that need to be done in a Control Phase. Once these activities are done, the project is considered closed and the Six Sigma team can submit a formal project closure document to the Champion.

6.52 Section VI Lesson 4 Introduction to Total Productive Maintenance

In this lesson we will be introduced to Total Productive Maintenance also known as TPM.

6.53 Agenda

In this lesson, we will discuss total productive maintenance. We will begin with an overview of total productive maintenance (also called as TPM). Followed by that, we will know how to implement TPM in an organization. Next, we will understand the eight pillars of TPM. Then, we will look into the details of computing the key metrics of TPM, namely Overall Equipment Effectiveness (OEE) and Total Effective Equipment Performance (TEEP ). The next slide deals with total productive maintenance.

6.54 Total Productive Maintenance(TPM)

TPM is used for better utilization and maintenance of production resources that result in improved process capability. The unreliable uptime is due to machine breakdowns, which further impacts the productivity and flow. Operators and teams of a company are the best people to seek and identify the problems in the equipment, before they assume the damaging proportions. Let us look into some facts of TPM in the following slide.

6.55 Total Productive Maintenance(TPM)(Contd.)

TPM philosophy was originated in Japan in nineteen fifty one, as a part of the preventive maintenance. Toyota embraced TPM first as a part of TPS in nineteen sixty. Nippondenso, a part of Toyota, mandated the principle of autonomous maintenance due to the high degree of automation in their company. Total productive maintenance (TPM) has 3 major goals. First one is to achieve zero product defects. Second is to strive for zero equipment unplanned failures (Please note that a scheduled system downtime is not factored by TPM as detrimental). The third and most important is zero accidents. We will find out how to implement TPM and its 8 pillars in the next slide.

6.56 Total Productive Maintenance(TPM)(Contd.)

Let us now understand how to implement TPM. TPM is done as follows. Gap analysis on the historical product is done to see if they are defects or accidents or failures. Then physical investigation on the equipment is done using Genchi Genbutsu and Jijutsu. Finally, corrective and incremental actions are implemented. The 8 Pillars of TPM are as follows: Focused improvement also known as Kobetsu Kaizen; Autonomous maintenance also known as Jishu Hozen; Planned maintenance; Training and education; Early phase management; Quality maintenance also known as Hinshitsu Hozen; Office TPM; and SHE (pronounced as S-H-E). We will understand OEE and TEEP, the two critical metrics TPM focuses heavily on in the next slide.

6.57 Total Productive Maintenance(TPM)(Contd.)

The two critical metrics TPM focuses on heavily are OEE and TEEP. OEE and TEEP are the key KPIs of TPM. OEE stands for overall equipment effectiveness. TEEP stands for total effective equipment performance. Both the above metrics show how well facilities in a site are utilized. OEE is Availability multiplied by Performance multiplied by Quality. In other words, OEE is A multiplied by P multiplied by Q. Availability is the time the equipment is available out of the total time. Performance is the performance of the equipment. Quality is the defect rate out of the equipment. Next, we will discuss availability and performance calculations.

6.58 Total Productive Maintenance(TPM)(Contd.)

Let us look into availability calculations first. Available production time per day is four eighty minutes. Scheduled break time is thirty minutes. Scheduled production time is four fifty minutes. We get that by subtracting the scheduled break time from the available production time. Scheduled downtime is sixty minutes. Available production time is three ninety minutes. This value is obtained by subtracting the scheduled downtime from the scheduled production time. Availability is eighty seven per cent. To get the availability figure, divide the available production time by the scheduled production time. However, dividing scheduled production time by available production time is incorrect. Now let us see the performance calculations. Performance is indicated by products produced multiplied by ideal cycle time divided by available time. Number of parts is hundred. With the actual time taken per part, three minutes, the available time for production is three ninety minutes from the availability calculations. Thus, performance is hundred multiplied by three divided by three ninety which gives the performance as seventy seven per cent. It should be noted that the performance calculations doesn’t factor quality. Let us move on to quality calculations in the next slide.

6.59 Total Productive Maintenance(TPM)(Contd.)

Let us now do the quality calculations. Quality is given by number of good units divided by total number of units. Quality is ninety divided by hundred, assuming out of hundred parts one will get ninety good units. Using the data we calculated before, OEE is eighty seven percent multiplied by seventy seven percent multiplied by ninety percent. That gives us a total of sixty point two one percent. Industry OEE benchmark is eighty five percent which indicates that there is a possible room for improvement. We will discuss TEEP in the next slide.

6.60 Total Productive Maintenance(TPM)(Contd.)

TEEP is loading multiplied by OEE. Loading is given by scheduled time divided by calendar time. Out of a calendar time of seven days and twenty four hours, the scheduled time is five days and twenty four hours. Thus the loading percentage is seventy one point four per cent. Multiply loading with OEE which gives us TEEP. From the calculations, TEEP is seventy one point four per cent multiplied by sixty point two one per cent giving us a TEEP of forty two point nine eight per cent. This tells us that only forty three per cent of the total effectiveness of the equipment is visible, which means that there is room for improvement.

6.61 Summary

Here is the summary of what we have learned in this lesson on Total Productive Maintenance (abbreviated as TPM). We started with understanding the overview of Total Productive Maintenance. After that we went into details on how to implement TPM in an organization. Followed by that, we understood the eight pillars of TPM. Then, we went in details of computing the key metrics of TPM, and understood how to calculate Overall Equipment Effectiveness (abbreviated as OEE) and Total Effective Equipment Performance (abbreviated as TEEP).

6.62 Tools to Refer

This slide contains a list of tools and their file formats to refer, which is provided in the toolkit.

6.64 Thank You

With this we have come to the end of the course. Thank you and happy learning.

  • Disclaimer
  • PMP, PMI, PMBOK, CAPM, PgMP, PfMP, ACP, PBA, RMP, SP, and OPM3 are registered marks of the Project Management Institute, Inc.

We use cookies on this site for functional and analytical purposes. By using the site, you agree to be cookied and to our Terms of Use. Find out more

Request more information

For individuals
For business
Name*
Email*
Phone Number*
Your Message (Optional)

By proceeding, you agree to our Terms of Use and Privacy Policy

We are looking into your query.
Our consultants will get in touch with you soon.

A Simplilearn representative will get back to you in one business day.

First Name*
Last Name*
Email*
Phone Number*
Company*
Job Title*

By proceeding, you agree to our Terms of Use and Privacy Policy