Higher Nationals

Assignment Brief
Pearson BTEC Higher National Certificate in Manufacturing Engineering for
England: 610/1229/4
Pearson BTEC Higher National Certificate in Electrical and Electronic
Engineering for England: 610/1220/8
Pearson BTEC Higher National Certificate in Mechanical Engineering for
England: 610/1231/2

Student Name/ID Number:

Unit Number and Title: Unit 4017: Quality and Process Improvement (Y/651/0734)

Academic Year: 2024/2025

Unit Assessor: Paul Parsons

Examine the applications of statistical process control when applied in

Assignment Title: an
industrial environment to improve efficiency

Issue Date: 03/01/2025

Formative Feedback: 05/02/2025

Submission Date: 19/02/2025

Internal Verifier Name: Lois Bamidele

Date:

Submission Format:

Electronic submission in word processed format. Excel spreadsheet showing calculations and pictoral
representations of examples cited where applicable.

Unit Learning Outcomes:

On successful completion of this unit a learner will:

1. Examine the applications of statistical process control when applied in an industrial environment to
improve efficiency

Assignment Brief and Guidance:

Prepared by: Higher Education Qualifications (HEQ) Approved by: HEQ, QDAM (HE)
Version 1.0 – November 2017
DCL 1 – Public (Unclassified)
1
Higher Nationals
As an engineer in an increasingly competitive world, the need to be able to apply engineering knowledge and
engineering methods to improve the efficiency of a workplace is of utmost importance. Without efficiency
improvements a company's costs rise and it cannot compete. One of the most important of these tools is
statistical process control (SPC), a tool which allows us to transform qualitative information into quantitative
parameters.

You are employed as a production engineer in an expanding manufacturing company (“Big Marker”) and have
been tasked with looking into SPC and how it may be used to improve the efficiency of the company. A
number of studies have been carried out by an external contractor and some information has been made
available to the company which you are to examine and present to your manager in a way such that he can
understand how it relates to moving the company forward. You are to make suggestions about how to
implement SPC going forward.

Part 1

A brand new part (cartridge holder B – part no “CHB012”) and process is to be introduced into a factory as
part of an ongoing growth in “Big Marker” pen’s product line. The process of design has been completed
and approved by the successful conclusion of pre-production batches of Big Marker pens built with the
CHB012 part. Using the new part will reduce the cost of the final product by 10% and it is important that the
new production of this part (which is automated) goes smoothly.

One of a number of identical new automatic machines which the company are planning to purchase to make
the part has been trialled in production as part of its commissioning by the external contractor. The details of
the commissioning sample include information on the length of the part which is considered the critical
dimension.

Measured Length of CHB012 During Commissioning Trial (units:mm)

137.4 137.1 136.8 137.2 136.9 137.3 137.1 136.9 137.1 136.9
137.2 137.2 137.3 137.1 136.7 137.3 137.4 136.7 136.9 137.1
137.2 137.2 137.1 136.8 137 136.9 137.1 137.2 137.3 137
136.9 137.1 136.8 137 137.2 136.8 137.4 137 137.1 136.8
137 137 137.4 137.1 137.3 136.7 137 137 137.3 137

The external contractor has commented that this is an “acceptable” result for that machine which is
intending to be bought.

You are to review the report which you have received from the external contractor and to comment on what
that means in regard to the process which the machine is due to carry out. The review is to relate to the
techniques used to support quality control. You are to examine the requirements of the machine in terms of
tolerances and how you may conclude that the machine is indeed “acceptable” to your company. You are to
show clearly how your decisions have been made outlining any limitations to your conclusions or other
recommendations you may have relating your decisions to the effectiveness of the application of SPC in a
production/industrial environment.

Prepared by: Higher Education Qualifications (HEQ) Approved by: HEQ, QDAM (HE)
Version 1.0 – November 2017
DCL 1 – Public (Unclassified)
2
Higher Nationals
Part 2

Presuming that the conclusions you draw will prompt the company to purchase the machine, you are also to
make suggestions about the potential production process in which the machine will be used and how SPC
and related tools can be facilitated to control quality in a manufacturing process giving justified
recommendations of the structure of such tools and how they may be created and used. You are to make it
clear what the benefits of implementing such tools may bring to the company in terms of efficiency.

Prepared by: Higher Education Qualifications (HEQ) Approved by: HEQ, QDAM (HE)
Version 1.0 – November 2017
DCL 1 – Public (Unclassified)
3
Higher Nationals
Table of Contents
Assignment Brief ............................................................................................................................................ 1
Pearson BTEC Higher National Certificate in Manufacturing Engineering for England: 610/1229/4 ............. 1
Pearson BTEC Higher National Certificate in Electrical and Electronic Engineering for England:
610/1220/8 ..................................................................................................................................................... 1
Pearson BTEC Higher National Certificate in Mechanical Engineering for England: 610/1231/2 .............. 1
Part 1 .................................................................................................................................................................. 5
Quality Control .............................................................................................................................................. 5
Statistical Process Control (SPC)................................................................................................................... 6
Cause Variables ............................................................................................................................................. 7
Common Cause Variables.......................................................................................................................... 7
Assignable Cause Variables ....................................................................................................................... 7
Normal Distribution ....................................................................................................................................... 8
Skewed Distribution................................................................................................................................... 9
Histogram ..................................................................................................................................................... 10
Sample Size.................................................................................................................................................. 11
Cp vs Cpk..................................................................................................................................................... 12
Cp ............................................................................................................................................................. 12
Cpk ........................................................................................................................................................... 12
The 8 Wastes ............................................................................................................................................ 14
Cartridge Holder B (CHB012) ..................................................................................................................... 15
Part 2 ................................................................................................................................................................ 19
Cause and Effect Diagram (Ishikawa) ......................................................................................................... 19
Control Chart ............................................................................................................................................... 21
Scenario 1................................................................................................................................................. 21
Scenario 2................................................................................................................................................. 23
Scenario 3................................................................................................................................................. 24
Pareto Chart ................................................................................................................................................. 26
Scatter Diagram ........................................................................................................................................... 27
Bibliography .................................................................................................................................................... 28

Prepared by: Higher Education Qualifications (HEQ) Approved by: HEQ, QDAM (HE)
Version 1.0 – November 2017
DCL 1 – Public (Unclassified)
4
Higher Nationals
Part 1
Quality Control

Figure 1: History of Quality Control

Quality Control is a vital part of many industries as it ensures that products and services are meeting specific
standards and expectations. It refers to the systematic process of monitoring data, analysing it, evaluating it
and making decisions based on justified research.

There are several techniques used in Quality Control, such as:

Statistical Process Control (SPC) which involves using statistical methods to monitor and control a process,
helping to identify variations that could affect product quality. There is also Six Sigma, this is a data-driven
approach which is put in place to eliminate defects and minimise any variability in the manufacturing
processes. Total Quality Management, also known as TQM, is another key technique that focuses on the
long-term success through customer satisfaction. Failure Mode and Effects Analysis (FMEA) is another
method which focusses on identifying possible failures in the design of something, a system or a process etc.

Quality control can provide a large quantity of benefits to a company. The main advantage would be
improving the quality of a product or service, and the reliability of it, so that it will be more consistently at
this quality. This overall leads to higher quality goods and services which will have a knock-on effect for
profits as we can make more money, as well as save some from cutting away at our losses.

Prepared by: Higher Education Qualifications (HEQ) Approved by: HEQ, QDAM (HE)
Version 1.0 – November 2017
DCL 1 – Public (Unclassified)
5
Higher Nationals
Statistical Process Control (SPC)

Figure 2: Method Of SPC

Statistical Process Control has been used in the industry for a very long time. It was first developed by a
man named William Shewhart at Bell Laboratories in 1924. He developed the control chart and the concept
that a process could be in statistical control. He then went on to publish a book in 1939, titled ‘Statistical
Method from the Viewpoint of Quality Control’. By this time the SPC process had began to take over the
world and gain wide usage specifically in the military and weaponry facilities as it coincided with the times
of World War II. The desperate demand for a large quantity of weaponry and other products in this time
forced them in to looking for more efficient methods to monitor the quality of a product without
compromising the safety and SPC filled that desire. After the war, the use of SPC began to dwindle and was
then picked up by Japanese manufacturing companies until 1970 where it began to take off again and still to
this day it is a very widely used quality tool throughout several different industries.

SPC is a method used for measuring and controlling the quality by monitoring the processes within
manufacturing. Data is collected through the product, or measurements, readings etc. and then this data can
be used to evaluate and control a process. By doing this we can look at the performance a process is
operating at, and if it is not at its full potential, we can look to improve it until it is.

We use SPC for this very reason. In industries we want all our processes to be performing at their maximum
potential, with as little waste as possible. These wastes could be scrapped products, reworked products and
the time it takes to inspect, but these, wastes can all add up to a lot of financial loss for the business, as every
failed product means wasted money on the products, wages on employees, bills on the machines, lifetime on
machine and so much more. Every waste has a huge domino effect, so it is very important that we limit these
to as little as possible. As previously mentioned, SPC also reduces the need for inspection time. A lot of
companies still rely on products to be inspected for their quality after being produced to detect any issues.
By implementing SPC, we can get real-time data and make changes before we have large amounts of scrap.

To implement SPC on an already existing or new process coming in we need to first determine the main
areas where we are seeing the most waste and focus on these areas first as we will see the most benefits here.

Prepared by: Higher Education Qualifications (HEQ) Approved by: HEQ, QDAM (HE)
Version 1.0 – November 2017
DCL 1 – Public (Unclassified)
6
Higher Nationals
Cause Variables
Cause Variables are the principle that there this a cause-and-effect relationship. These need to be
investigated before and after they occur so that we can analyse them and then begin to address the issues.
Variation in a process is expected and practically unavoidable; however, by understanding control charts and
recognising different cause variations we can alter processes to minimise the chances of them occurring
again, maximising a machines potential output. These variations can be split into two different categories:

Common Cause Variables

Common cause variables are the natural variations that occur during the process. They are a part of a stable
process and usually are in abundance in comparison to the ‘assignable cause variables’ as they are often hard
to identify and remove, and if we were to try and alter and rectify a common cause, it would usually require
a big change to be made. The common causes variables are usually predictable, ongoing and relatively
consistent. An example of these could be stuff like standard wear and tear on components, computer lag
time, and measurement errors etc.

Because they are consistent and always present, it means that we can analyse them and take them into
account when building control limits for a control chart and other baselines for our statistics. We can also
identify them when we have made a control part by looking at random patterns of variation.

Assignable Cause Variables

Special cause variables are unexpected in a process and can significantly affect the process, including the
products. These variations are typically unusual, and unquantifiable as the variations have not occurred or
been observed before; and because of this, they cannot be considered for the planning within a system. They
are usually because of a very specific change that has been made during the process and often result in a big
consequence. An example of this could be a complete power outage, component malfunction or a task taking
much longer than previously expected due to an unaccounted-for crisis.

A good thing about assignable cause variables is that they are usually connected to a defect in a system that
we can address and make amendments to rectify the issue. This means we can work towards the variation
from never occurring again. We can identify these by looking on our control parts and looking for really out
of control points.

Prepared by: Higher Education Qualifications (HEQ) Approved by: HEQ, QDAM (HE)
Version 1.0 – November 2017
DCL 1 – Public (Unclassified)
7
Higher Nationals
Normal Distribution

Figure 3: Normal Distribution Example

Normal distribution is the most ‘normal’ form of distribution of random variables, also known as the
‘Gaussian Distribution’. It is a continuous probability distribution that is symmetric around the mean,
depicting that data closer to the mean is more frequent and as it stretches out further, it is less frequent. It is
represented by a normal distribution curve, which is also known as a bell-shaped curve.

The key features of a Normal Distribution Graph would be the symmetry aspect and the mean, median and
mode being all equal, located at the centre of the distribution. The curve, as previously mentioned, is bell-
shaped meaning the frequencies taper off equally in both directions and cluster in the centre. The spread of
the distribution, hence the shape of the curve (narrow/thick) is determined by the standard deviation. Around
68% of the data falls within 1 standard deviation of the mean, 95% within two standard deviations and
99.7% within three standard deviations.

The normal distribution has a wide range of applications that vary across all different fields from physical
sciences to social sciences. Its main goal is to transform a large set of data that is otherwise meaningless and
hard to analyse into a model that we can easily analyse and make justified decisions based on the new
representation of the data.

The data can help researchers to make predictions as there are many natural phenomena that are
accommodated for in the pattern of the normal distribution such as measurement errors (industry related) or
rainfall and other random characteristics. Additionally, the normal distribution curve can also guarantee a
level reliability and validity of your findings, because, if a graph has a sufficient and appropriate sample of
data behind it, the curve will maintain its general structure; however, if it didn’t, we could find that the curve
is distorted and losing its structure, meaning that the data may not be reliable.

Prepared by: Higher Education Qualifications (HEQ) Approved by: HEQ, QDAM (HE)
Version 1.0 – November 2017
DCL 1 – Public (Unclassified)
8
Higher Nationals
Skewed Distribution

A skewed distribution is when one tail on a distribution curve happens to be longer than the other, unlike the
standard, symmetrical normal distribution. It occurs when the data is not distributed equally on both sides of
the distribution’s peak. This does not just mean that the sample is unreliable as they have logical reasons for
occurring, such as natural limits and causes that skew the curve away from the boundary.

Figure 5: Right Skewed (Positively Skewed) Figure 4: Left Skewed (Negatively Skewed)

A Right Skewed distribution occurs when the long tail is on the right side of the curve. This happens when
the data shows that the probabilities taper off more slowly for the higher values. Because of this, you will
find extreme values far from the peak on the high end more frequently than the low. This the more common
form of skewed distribution that you will find. You can find these in distributions based on income as it
cannot be less that zero, but there are some extremely high values; and time to failure, as it cannot be less
than zero but there is technically no upper limit. In essence, these are more common as a lot of distributions
cannot be below zero, but they can have very high values.

A Left Skewed distribution occurs when the long tail is on the left side of the curve. This happens when the
data shows that the probabilities taper off more slowly for the lower values. As a result, you will find
extreme values far from the peak on the low side more frequently than the high side. These are slightly less
common but can be found with distributions that do have a maximum limit. Such as, test scores as they
cannot be above 100%.

Prepared by: Higher Education Qualifications (HEQ) Approved by: HEQ, QDAM (HE)
Version 1.0 – November 2017
DCL 1 – Public (Unclassified)
9
Higher Nationals
Histogram

Figure 6: Normal Distribution in Histogram Form

A normal distribution typically has the pattern of a bell-shaped curve, and we can use histograms to also
show a similar pattern and structure to this. We can do this by using the sample that has been collected,
plotting the data into a table and then figuring out the frequency of each data point. The more frequent data
points will be the taller bars on the histogram. As a result of this, it is almost identical to the normal
distribution curve but rather blockier.

Because of this we can also use a histogram to prove whether a sample size is representative. This is
because, we can make one and analyse the structure of it to. By analysing the structure of the distribution,
we can see where our data lies. If the data is forming something like the familiar bell-shaped curve, then it is
likely that the sample is representative of the population

Figure 7: Right Skewed Histogram Distribution Figure 8: Left Skewed Histogram Distribution

These are skewed histogram distributions. These do not typically mean that the sample is not representative,
as there are naturally occurring limits or lack there of that can result in these types of distributions, but it can
be a reason to have a look at the data again before continuing further on in the quality control process, as it
could possibly indicate that the sample size is not representative of the population.

Prepared by: Higher Education Qualifications (HEQ) Approved by: HEQ, QDAM (HE)
Version 1.0 – November 2017
DCL 1 – Public (Unclassified)
10
Higher Nationals
Sample Size

Figure 9: Sample Size Example

A sample size is the number of observations, samples or individuals included in a study or an experiment.
So, it is the number of data points that are collected from a larger population to represent it statistically, in a
quantifiable and analysable format. Typically, the larger the size of the sample, the more likely it is to be
accurate, as we have a wider range of data, allowing for a narrower margin of error. In essence, the results
from a larger sample, are more likely to be true to the whole population.

There is no ‘proper’ way to determine a sample size, and it is up to who is carrying out the study/experiment
to decide whether the sample size is ‘representative’ of the whole population; however, there are things we
can take into consideration to help up decide what a representative sample is.

First, what the study is being based on can play a big role in this decision-making process. For example, a
case aiming to study on a disease or other serious health concerns would need a much larger population size
to reduce the chance of inaccuracies as this is very ‘precious data’; on the other hand, there are some studies
that are not as serious or precious so they may not require such precision. The population size as a whole can
also determine how big of a sample size is required, generally, if you’re population size is small, you will
need to take a relatively large sample ration than in a larger population. Finally, there is the resources
available to you, you must consider the resources you have access to and find the way to get the most out of
these, while not working outside these resource limits (time, money etc.) all while obtaining the appropriate
sample size at the time.

Prepared by: Higher Education Qualifications (HEQ) Approved by: HEQ, QDAM (HE)
Version 1.0 – November 2017
DCL 1 – Public (Unclassified)
11
Higher Nationals
Cp vs Cpk

Figure 10: Cp vs Cpk on Normal Distribution Curve

Cp and Cpk are both capability indices that are used to indicate how sufficient a process is in making the
desired requirement. They both compare the spread of samples and then condense them to make it a
specification spread to measure the capabilities of the process.

They can be used when designing new processes and improving already existing processes. By comparing
what the yield of process should be to what the process is yielding can show us how much improvement is
needed on the process to get the value of the actual yield closer to the desired yield. By using Cp and Cpk
appropriately we can expect more product, lower costs and less inspection times.

Cp

The Cp index is the simpler version of the two indicators of process capability, also known as the ‘process
potential index’. This index will indicate whether the process has the potential to produce products that meet
the specifications and tolerances that are desired.

It is measured using this formula:

𝑈𝑆𝐿 − 𝐿𝑆𝐿
𝐶𝑝 =
6𝜎

USL = Upper Standard Limit

LSL = Lower Standard Limit
σ = Standard deviation of the process
Cpk

A Cpk index is, in simplified terms, an adjustment of a Cp index that accommodates for a non-centred
distribution effect. Unlike Cp, the centre of the curve is not around the ‘mean’ of the graph. Cpk can indicate
the process’s capability of the process to produce within the limits of the specifications, like Cp; however, it
Prepared by: Higher Education Qualifications (HEQ) Approved by: HEQ, QDAM (HE)
Version 1.0 – November 2017
DCL 1 – Public (Unclassified)
12
Higher Nationals
is also able to indicate the ability of the process to be able to conform to the target values. The difference
between the Cp and Cpk will indicate how far the average of the process, and the process targets are from
one another. When these two are close to each other, the closer the gap between the two on a graph will also
be closer, and vice versa.

It is measured using this formula:

𝑈𝑆𝐿 − µ µ − 𝐿𝑆𝐿
𝐶𝑝𝑘 = 𝑚𝑖𝑛 ( , )
3𝜎 3𝜎

USL = Upper Standard Limit

LSL = Lower Standard Limit
min = Minimum values between the two terms.
µ = Mean of the process
σ = Standard deviation of the process

We can also use Cpk with ‘PPM’ (Parts Per Million). PPM is a figure to represent the number of parts per a
million produced that would be classes as scrap or waste. We use Cpk with this to give us a more
representative value of the number, so it shows

Figure 11: CPK to PPM Conversion Table

With this table we can use our Cpk values to figure

out how parts per million produced in a process are
going to be ‘waste’. This gives us a good
representation of how reliable and consistent our
process is. The industry standard for a good Cpk is
1.4, which would give us a PPM of 27.

Prepared by: Higher Education Qualifications (HEQ) Approved by: HEQ, QDAM (HE)
Version 1.0 – November 2017
DCL 1 – Public (Unclassified)
13
Higher Nationals
The 8 Wastes

Figure 12: Example of how the Motion Waste

The 8 Wastes tie in nicely with the PPM as every time we are likely to be having multiple of these wastes
within the process which account for a loss in money, reducing the maximum potential for profits within a
company. We can use the 8 wastes to separate our types of waste and begin to reduce or even eliminate
wastes as much as possible; however, there will always be some amount of waste in every process.

Defects: Defects are any product that is made that is not to the desired standards. These could complete
rejects and scrap, or products that need repairing or reworking. Every time a defect occurs, we could be
contributing to a loss in profits through material usage, employee wages, machine running costs etc.

Overproduction: Overproduction is a waste where we create a product that is not needed. This would be a
waste of our employee’s time, running costs of the machine and materials. It also reduces the space that we
could have for other products to be stored in.

Waiting: This waste is simply any time that we spend waiting on a task before another task can be carried
out. These are called dependant tasks. A lot of processes work in steps that must follow each other, so it is
all about reducing the time in between these steps as much as possible.

Non-utilised Talent: This waste is classified as underutilisation of our workforce. For example, this could
be a process has gone down because it is producing scraps, and it needs working on by the engineers to get it
right. Now not only do we have the defects waste, but we also have an employee that is not being utilised
properly, waiting for the machine to be fixed.

Transportation: Transportation refers to moving items through the process, this can either be physically or
digitally. Each time an item is moved, it adds cost which could open opportunities for more defects and
other wastes.

Motion: This is the movement of our employees or machines between steps in a process. The aim of this
waste, similarly to ‘Waiting’, would be to reduce this motion as much as possible to shorten the process,
meaning more products.

Inventory: This waste is the unnecessary overload of our stock, materials etc. In a company, we should
strive to have a balance of having just the right amount of inventory so that we have enough to carry out our
tasks but also minimise the costs and risks that come with having too much inventory.

Extra-Processing: This waste is about when processes and methods are over complicated and entail too
many steps, where a simpler method would give us the same output but in a shorter amount of time and with
less of the other wastes too.
Prepared by: Higher Education Qualifications (HEQ) Approved by: HEQ, QDAM (HE)
Version 1.0 – November 2017
DCL 1 – Public (Unclassified)
14
Higher Nationals
Cartridge Holder B (CHB012)
For the cartridge holder we are given this group of data:

137.4 137.1 136.8 137.2 136.9 137.3 137.1 136.9 137.1 136.9
137.2 137.2 137.3 137.1 136.7 137.3 137.4 136.7 136.9 137.1
137.2 137.2 137.1 136.8 137 136.9 137.1 137.2 137.3 137
136.9 137.1 136.8 137 137.2 136.8 137.4 137 137.1 136.8
137 137 137.4 137.1 137.3 136.7 137 137 137.3 137

We are deciding that this is a representative sample of the process, meaning we are going to analyse it and
should get more precise and accurate results.

Using the data, here, I have created a Histogram to represent the data.

The data creates something close to a normal distribution curve, further reinforcing that the data is
representative.

Average: 137.066
Standard Deviation = 0.1935

Prepared by: Higher Education Qualifications (HEQ) Approved by: HEQ, QDAM (HE)
Version 1.0 – November 2017
DCL 1 – Public (Unclassified)
15
Higher Nationals
Scenario 1:

For this scenario, I have chosen these values as my USL and LSL:

USL – 143.7
LSL – 130.3

With these values I was able to calculate my Cp and Cpk:

Cp: 0.43
Cpk: 0.43

And with my Cpk value I was able to conclude that my PPM would be, 193,600. This would mean that for
every million parts I produce, 193,600 of them are likely to be outside of the standard limits. This would
result in a lot of scrap and wastes for the company that could minimise our profit margins.

This example is very simplistic and does not consider a lot of other variables in the process, but it still gives
the idea of how much a company could lose.

Scenario 1 1000 Products (1 Day) 1,000,000 Products

Marker Price (Sold at) £1,200 £1,200,000
Labour -£120 -£120,000
Running Expenses -£250 -£250,000
Materials -£220 -£220,000
Ideal Profit £610 £610,000
Scrapped Parts -£232.80 -£231,800.00
Actual Profit £377.20 £378,200.00

From here you can see that for every 1,000,000 products produced there is the potential profit of £610,000;
however, because of the machines lack of adherence to the standard limits; therefore, resulting a lot of
‘waste’ products, we are losing out on £231,800 of these potential profits just on scrap alone. This is a huge
loss for a company, especially when it can be so easily reduced with the proper business improvement
techniques in place.

This example does not account for the potential scrap items that can be reworked, and therefore, would not
be a loss of materials, but this would still impact our labour and running expenses of the machine, so there is
still a lot of loss.

Prepared by: Higher Education Qualifications (HEQ) Approved by: HEQ, QDAM (HE)
Version 1.0 – November 2017
DCL 1 – Public (Unclassified)
16
Higher Nationals
Scenario 2:

For this scenario, I have chosen these values as my USL and LSL:

USL –149.9
LSL – 124.1

With these values I was able to calculate my Cp and Cpk:

Cp: 0.83
Cpk:0.83

And with my Cpk value I was able to conclude that my PPM would be, 12,420. This would mean that for
every million products produced, 12,420 are waste.

Scenario 2 1000 Products (1 Day) 1,000,000 Products

Marker Price (Sold at) £1,200 £1,200,000
Labour -£120 -£120,000
Running Expenses -£250 -£250,000
Materials -£220 -£220,000
Ideal Profit £610 £610,000
Scrapped Parts -£14.90 -£14,904.00
Actual Profit £595.10 £595,096.00

In this scenario, as the PPM is much smaller than the first scenario, you can see already that there has been a
massive decrease in the potential profits lost from scrapped parts. Although, the loss has been reduced by a
lot, still a Cpk of 0.83 is not the industry standard and companies will thrive to aim for even higher than this.
As you can see the profit loss is small, but it still adds up to a substantial amount, and this money could be
used for other ventures in the business, rather than being wasted on the machine.

It is important to note again that these scenarios do not account for possible reworks, but they also do not
take into consideration plenty of other wastes that would be caused by a scrap part which I am sure would
raise the loss margin considerably.

Prepared by: Higher Education Qualifications (HEQ) Approved by: HEQ, QDAM (HE)
Version 1.0 – November 2017
DCL 1 – Public (Unclassified)
17
Higher Nationals
Scenario 3:

For this scenario, I have chosen these values as my USL and LSL:

USL – 159.2
LSL – 114.8

With these values I was able to calculate my Cp and Cpk:

Cp: 1.43
Cpk: 1.43

And with my Cpk value I was able to conclude that my PPM would be, 17. This is an extremely good PPM
and a very desirable figure. A Cpk of 1.43 is close to the industry standard. Of course, we would love to
strive for 2; however, this is practically impossible as there is always some form of variation in a process.

Scenario 3 1000 Products (1 Day) 1,000,000 Products

Marker Price (Sold at) £1,200 £1,200,000
Labour -£120 -£120,000
Running Expenses -£250 -£250,000
Materials -£220 -£220,000
Ideal Profit £610 £610,000
Scrapped Parts -£0.02 -£20.40
Actual Profit £609.98 £609,979.60

Here you can see that across 1 million products, we are only losing out on £20.40, excluding other potential
wastes. This is brilliant for the company, as they are really maximising their potential earnings. With a Cpk
this high, we could essentially be happy with this process and leave it as it is, or we could strive for even
higher; however, it would be hard to get higher and maybe not worth the resources spent on the research as
we could begin to look at other processes and wastes in the company.

Prepared by: Higher Education Qualifications (HEQ) Approved by: HEQ, QDAM (HE)
Version 1.0 – November 2017
DCL 1 – Public (Unclassified)
18
Higher Nationals
Part 2
Cause and Effect Diagram (Ishikawa)
Kaoru Ishikawa was a Japanese quality control expert, and he was the man who introduced the Ishikawa
Diagram in the 1960’s. It was his aim to provide an easily digestible visual representation of the potential
causes for a specific problem. This would help to make it easier for teams to identify and address issues.

A cause-and-effect diagram, also known as an Ishikawa diagram and a fishbone diagram, is a visual tool that
essentially has all the factors of ‘brainstorming’, but in a much more professional and structured way. It
maps out all the possible causes of a problem, which can help us to try and find connections between the
symptom and its root cause. By doing this we can find the root cause of a problem and then begin to work
towards, eliminating or reducing the chance of the symptom occurring again, from said cause. It is used in
Lean Six Sigma, Quality Control, and Root Cause Analysis.

The diagram is shaped like a fishbone, hence the name, with the head being the problem/symptom and then
all of the potential causes branch off on a straight line, like a spine. It is a very simple visual representation,
but very effective and industry proven; which is why it is still used today.

Figure 13: Ishikawa on Broken Motor

Above is an Ishikawa diagram based on a made-up scenario of a motor being broken on a machine, which
would have stopped production. As a result of this, we want to find out the root cause of why this motor
broke down, causing production to stop, which led to a loss of profits for the company.

To make it I first set out my problem ‘Broken Motor’ at the head of the diagram and then branched out with
different potential causes. From these potential causes, I was able to go in more detail about what
specifically about this cause may have caused the motor to break down. The idea of the diagram is to throw
any possible ideas on to the board so that we can then begin to eliminate causes until we have found the true
root cause.

Prepared by: Higher Education Qualifications (HEQ) Approved by: HEQ, QDAM (HE)
Version 1.0 – November 2017
DCL 1 – Public (Unclassified)
19
 Higher Nationals
Here is another Ishikawa Diagram I have made, with the problem being ‘Product out of Tolerance’, to make
it more relevant to previous scenarios.

Figure 14: Ishikawa Diagram on Tolerances

Now that I have made this Ishikawa Diagram, I can begin to investigate all the possible sub-causing to
determine the likelihood of that being the cause. From here we can slowly narrow it down until we come to
a conclusion, which would be our final root cause(s). When we have find out what our root causes are it is
very important that we then begin to put our resources into investigating how we can begin to reduce or
eliminate the likelihood that it occurs again.

Prepared by: Higher Education Qualifications (HEQ) Approved by: HEQ, QDAM (HE)
Version 1.0 – November 2017
DCL 1 – Public (Unclassified)
20
Higher Nationals
Control Chart
A control chart is a quality tool that we can use to track and control a process, specifically it’s performance
over a certain period. This can be used in finance, engineering, manufacturing and several other industries.
The purpose of it is to identify any deviations in the performance of the process, from these deviations we
can begin to work towards preventing them.

It works by visually displaying the process performance over a period, so when an irregularity occurs, it is
very distinct and traceable to a certain time. We can use this to further investigate what may have caused it.
It uses UCL (Upper Control Limit) and LCL (Lower Control Limit), to help determine where an irregular
point may be, as this is the acceptable range of variation.

We can calculate these using these formulas:

̅ + 𝟑𝝈
𝑼𝑪𝑳 = 𝑿 ̅ – 𝟑𝝈
𝑳𝑪𝑳 = 𝑿

I created my own control charts, relevant to the marker scenario. I used my own sample values but
maintained the original mean and standard deviation values across all three, to keep this as the standard.
This why my upper control limits and lower control limits are the same across all 3 scenarios. For these
scenarios I have kept the USL and LSL of the previous scenario. USL = 139.3 and LSL = 134.7.

Scenario 1

Firstly, I made this table with random samples. I kept a mean of 137 and a standard deviation of 0.1935. I
inputted these values into the UCL and LCL formulas to get my control limits which can be seen below.
With the data from the table I was able to create the control chart.

Shift 1 Samples Mean Std Dev UCL LCL USL LSL

06:00 137.3 137 0.1935 137.5805 136.4195 139.3 134.7
06:30 137 137 0.1935 137.5805 136.4195 139.3 134.7
07:00 137.1 137 0.1935 137.5805 136.4195 139.3 134.7
07:30 136.9 137 0.1935 137.5805 136.4195 139.3 134.7
08:00 136.9 137 0.1935 137.5805 136.4195 139.3 134.7
08:30 137.1 137 0.1935 137.5805 136.4195 139.3 134.7
09:00 137 137 0.1935 137.5805 136.4195 139.3 134.7
09:30 137 137 0.1935 137.5805 136.4195 139.3 134.7
10:00 136.7 137 0.1935 137.5805 136.4195 139.3 134.7
10:30 137.1 137 0.1935 137.5805 136.4195 139.3 134.7
11:00 137 137 0.1935 137.5805 136.4195 139.3 134.7
11:30 137 137 0.1935 137.5805 136.4195 139.3 134.7
12:00 136.8 137 0.1935 137.5805 136.4195 139.3 134.7
12:30 137 137 0.1935 137.5805 136.4195 139.3 134.7
13:00 137.1 137 0.1935 137.5805 136.4195 139.3 134.7
13:30 137 137 0.1935 137.5805 136.4195 139.3 134.7
14:00 137 137 0.1935 137.5805 136.4195 139.3 134.7
Prepared by: Higher Education Qualifications (HEQ) Approved by: HEQ, QDAM (HE)
Version 1.0 – November 2017
DCL 1 – Public (Unclassified)
21
Higher Nationals

Morning Shift (Bob)

140

139

138
Sample size (mm)

137

136

135

134
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
Time

Morning Shift Samples Mean UCL LCL USL LSL

In this scenario, it is Bob on the morning shift. These details are important as if there were any issues we
need to be able to trace it to exactly who was present at the time of the deviation.

Fortunately, in this scenario, there is not much deviation, and all the samples stayed within the control
limits, let alone the standard limits. We could potentially investigate on the circled points to further refine
the process; however, these variations could be standard due to the nature of the process and as a result, not
resolvable. Regardless, this is a pretty good control chart in terms of performance.

Prepared by: Higher Education Qualifications (HEQ) Approved by: HEQ, QDAM (HE)
Version 1.0 – November 2017
DCL 1 – Public (Unclassified)
22
Higher Nationals
Scenario 2

Once again, I have made a table with different points of data but the same control limits, analysing the
performance of another shift.

Shift 2 Samples Mean Std Dev UCL LCL USL LSL

21:30 138 137 0.1935 137.5805 136.4195 139.3 134.7
22:00 138.7 137 0.1935 137.5805 136.4195 139.3 134.7
22:30 139 137 0.1935 137.5805 136.4195 139.3 134.7
23:00 139.7 137 0.1935 137.5805 136.4195 139.3 134.7
23:30 138.5 137 0.1935 137.5805 136.4195 139.3 134.7
00:00 138.4 137 0.1935 137.5805 136.4195 139.3 134.7
00:30 138.6 137 0.1935 137.5805 136.4195 139.3 134.7
01:00 137 137 0.1935 137.5805 136.4195 139.3 134.7
01:30 137.2 137 0.1935 137.5805 136.4195 139.3 134.7
02:00 137.1 137 0.1935 137.5805 136.4195 139.3 134.7
02:30 137 137 0.1935 137.5805 136.4195 139.3 134.7
03:00 137 137 0.1935 137.5805 136.4195 139.3 134.7
03:30 137.2 137 0.1935 137.5805 136.4195 139.3 134.7
04:00 137 137 0.1935 137.5805 136.4195 139.3 134.7
04:30 136.9 137 0.1935 137.5805 136.4195 139.3 134.7
05:00 137 137 0.1935 137.5805 136.4195 139.3 134.7
05:30 136.9 137 0.1935 137.5805 136.4195 139.3 134.7

Night Shift (Alan)

140

139

138
Samples (mm)

137

136

135

134
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
Time

Night Shift Samples Mean UCL LCL USL LSL

Prepared by: Higher Education Qualifications (HEQ) Approved by: HEQ, QDAM (HE)
Version 1.0 – November 2017
DCL 1 – Public (Unclassified)
23
Higher Nationals
In this scenario we can see that in the earlier hours of the shift we have some values that go outside of our
control limits, and even one that goes out of our standard limit.

This is Alan coming on the night shift after Bob. From the moment he gets on the machine, it is already out
of the control limits; however, it is not immediately a cause for concern, especially to stop product as these
products are within the tolerance limits for the markers.

These points would need investigating after the process has been completed though, because of its time we
could draw conclusions that it could possibly be something due to how Bob left the machine after his shift or
anything that could have altered the machine in between the shifts (maintenance) or how Alan has set up the
machine for his shift, or it could be an issue unrelated to this, regardless it will need investigation.

Next there is a point outside of the standard limits. At this point, the process would need completely shutting
down and working on in real time with engineers to resolve the issue as it is currently producing scrap
product.

After this point, there are more points above the control limits. These would need investigating too, but it
could be down to the engineers working on the machine until they get it right. This is a very bad scenario so
we can assume that something was seriously wrong and it took a few adjustments until the machine was
back on track, which it eventually does do.

Scenario 3

Once again, I have made a table with different points of data but the same control limits, analysing the
performance of another shift.

Shift 3 Samples Mean Std Dev UCL LCL USL LSL

06:00 136.9 137 0.1935 137.5805 136.4195 139.3 134.7
06:30 136.6 137 0.1935 137.5805 136.4195 139.3 134.7
07:00 136.8 137 0.1935 137.5805 136.4195 139.3 134.7
07:30 136.9 137 0.1935 137.5805 136.4195 139.3 134.7
08:00 136.9 137 0.1935 137.5805 136.4195 139.3 134.7
08:30 136.2 137 0.1935 137.5805 136.4195 139.3 134.7
09:00 136 137 0.1935 137.5805 136.4195 139.3 134.7
09:30 136.4 137 0.1935 137.5805 136.4195 139.3 134.7
10:00 136.6 137 0.1935 137.5805 136.4195 139.3 134.7
10:30 136.5 137 0.1935 137.5805 136.4195 139.3 134.7
11:00 137 137 0.1935 137.5805 136.4195 139.3 134.7
11:30 136.9 137 0.1935 137.5805 136.4195 139.3 134.7
12:00 136.6 137 0.1935 137.5805 136.4195 139.3 134.7
12:30 136.8 137 0.1935 137.5805 136.4195 139.3 134.7
13:00 137 137 0.1935 137.5805 136.4195 139.3 134.7
13:30 136.7 137 0.1935 137.5805 136.4195 139.3 134.7
14:00 136.9 137 0.1935 137.5805 136.4195 139.3 134.7

Prepared by: Higher Education Qualifications (HEQ) Approved by: HEQ, QDAM (HE)
Version 1.0 – November 2017
DCL 1 – Public (Unclassified)
24
Higher Nationals
Morning Shift 2 (Bob)
140

139

138
Sample Size(mm)

137

136

135

134
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
Time
Morning Shift 2 Samples Mean UCL LCL USL LSL

Scenario 3 is Bob on the morning shift again. There are 2 points that go outside of the control limits, these
would not stop production but thanks to the control chart, we can identify them and begin to investigate
further.

This scenario is interesting as none of the samples go outside of the standard limits, resulting in no scrap, so
you would think there is no issue; however, because of the use of the control chart we can spot a very clear
trend within the data. The sample sizes tend to stay below the mean across the whole shift. This would be
cause for serious investigations in the process, as the machine is clearly not performing to its desired values.
This could be a problem with how the machine is set up, or the capabilities of the machine with a certain
pattern etc.

Prepared by: Higher Education Qualifications (HEQ) Approved by: HEQ, QDAM (HE)
Version 1.0 – November 2017
DCL 1 – Public (Unclassified)
25
Higher Nationals
Pareto Chart
The Pareto Chart originated in 1897. It was created by an Italian economist called ‘Vilfredo Pareto’ when he
made a formula representing the uneven distribution of wealth, this would then later be known as the 80-20
rule, we know of today as the Pareto Chart. This 80-20 rule is explained as ‘80% of the problems are caused
by 20% of the causes’.

The Pareto chart is now used in all industries as a tool to rank various categories/causes of a problem by
charting them as bars in a descending order of frequency, with the largest being on the left and the smallest
being on the right. We use The Pareto Chart to help establish areas of priority in a process by dividing the
causing and seeing which are most present. By identifying the most common causes, we can begin to
investigate and work towards reducing the frequency of these causes occurring. The aim is to reduce the
frequency of the causes, as much as possible, working from the most common to least common.

Below I have made a table of random data for a made-up scenario.

Causes Frequency Cumulative Frequency Percentage Cumulative Percentage

Machine Parameters 58 58 58 58
Worn parts 23 81 23 82
Operator Error 11 92 11 92
Poor Quality Material 5 97 5 97
Belt Slipping 2 99 2 99
Other 1 100 1 100

With this data, I was able to create the Pareto Chart below.

With this Pareto chart we are easily able to identify where our main areas for concern are. We can then
begin to work towards reducing the likelihood of this cause occurring. This will help to reduce the amount
of ‘waste’ product we are getting. This has a domino effect for savings, saving money on the running
expenses, employee wages, materials, machine lifespan, tool lifespan and many more.
Prepared by: Higher Education Qualifications (HEQ) Approved by: HEQ, QDAM (HE)
Version 1.0 – November 2017
DCL 1 – Public (Unclassified)
26
Higher Nationals
Scatter Diagram
The Scatter Diagram was first introduced in 1833 by an English scientist who goes by the name of John
Frederick W. Herschel, while he was conducting his study of Orbits of Double stars and then later in 1886 it
gain a broader recognition, as a result of Francis Galton, a polymath who popularised it and introduced the
statistical concept of correlation.

A scatter diagram is a graph that pairs numerical data with a variable on each access, so that we can look for
a relationship between them. If there is a relationship between them, the points will fall along a line or curve,
and the better a correlation is the tighter these points will hug the line.

We can use a scatter diagram after we have made our initial fishbone diagram as it is a good way to analyse
the causes brought up in this quality tool and see if they truly are the root cause of the problem.

Figure 15: Scatter Diagram Example

Above is an appropriate example for a scatter diagram used in car accidents. They have set out a table with
data and are trying to see if any of the causes are in correlation with the number of accidents in a day.

There are a few examples above, the top right one, has a very clear correlation with the number of accidents
in a day as the points are tightly hugging each other and forming a curve/line. In the bottom left diagram,
there is no real correlation between the points as they are too far apart and far too random. In the bottom
right, while they are slightly more random and spaced out, there is some form of correlation between the
points as they follow a similar line; however, it would be a loose correlation.

Prepared by: Higher Education Qualifications (HEQ) Approved by: HEQ, QDAM (HE)
Version 1.0 – November 2017
DCL 1 – Public (Unclassified)
27
Higher Nationals
Bibliography
ASQ (2019a). What are histograms? Analysis & frequency distribution. [online] Asq.org. Available at:
https://asq.org/quality-resources/histogram.
ASQ (2019b). What is a Scatter Diagram? Scatter Plot Graphs | ASQ. [online] Asq.org. Available at:
https://asq.org/quality-resources/scatter-diagram.
Croft, D. (2023). Understanding Process Capability Index (Cpk) [With Calculator] - Learn Lean Sigma.
[online] www.learnleansigma.com. Available at: https://www.learnleansigma.com/control/understanding-
process-capability/.
Frisco, C.D. (2024). What Are The 8 Wastes of Lean? - Cannsult Inc | Business Consulting Services & Lean
Certifications. [online] Cannsult Inc | Business Consulting Services & Lean Certifications. Available at:
https://cannsult.org/post/what-are-the-8-wastes-of-lean/ [Accessed 19 Feb. 2025].
Frost, J. (2021). Skewed Distribution. [online] Statistics By Jim. Available at:
https://statisticsbyjim.com/basics/skewed-distribution/.
geeksforgeeks (2023). Normal Distribution| Definition, Formula, Properties, Uses. [online] GeeksforGeeks.
Available at: https://www.geeksforgeeks.org/normal-distribution/.
latestquality (2017). The difference between Cp and Cpk and when they are used. [online] Latest Quality.
Available at: https://www.latestquality.com/difference-between-cp-and-cpk/.
Learn Lean 6 Sigma. (2025). Master The Fishbone Diagram For Root Cause Analysis. [online] Available at:
https://www.learnleansigma.com/root-cause-analysis/fishbone-root-cause-analysis/.
MoreSteam. (n.d.). Pareto Chart (Manufacturing Example). [online] Available at:
https://www.moresteam.com/toolbox/pareto-chart-manufacturing.
prabh jot (2024). Why Normal Distribution Is Important In Statistics? [online] StatAnalytica. Available at:
https://statanalytica.com/blog/why-normal-distribution-is-important/ [Accessed 19 Feb. 2025].
Quality Gurus (2023). Quality Control: Importance, Benefits, Approaches and Strategies. [online] Quality
Gurus. Available at: https://www.qualitygurus.com/quality-control-101-understanding-its-importance-
benefits-approaches-and-key-strategies/.
quality-one (2015). SPC | Statistical Process Control | Quality-One. [online] Quality-one.com. Available at:
https://quality-one.com/spc/.
Ruiz, N. (2023). What Is Statistical Process Control (SPC)? [online] Sixsigma DSI. Available at:
https://sixsigmadsi.com/what-is-statistical-process-control-spc/.
Staff, C. (2024). What Is Sample Size? [online] Coursera. Available at:
https://www.coursera.org/articles/what-is-sample-size?msockid=3a79d920018167fe0b25cc2f00b06689
[Accessed 19 Feb. 2025].
Staff, iSixSigma (2022). Common Cause vs. Special Cause Variation: What’s the Difference? [online]
isixsigma.com. Available at: https://www.isixsigma.com/methodology/common-cause-vs-special-cause-
variation-whats-the-difference/.
Status.net. (2023). Ishikawa Diagram: Examples and Applications. [online] Available at:
https://status.net/articles/ishikawa-diagram-examples-applications/.
Syque.com. (2022). The Quality Toolbook: Examples of the Scatter Diagram. [online] Available at:
http://syque.com/quality_tools/toolbook/Scatter/example.htm [Accessed 19 Feb. 2025].
TimesPro. (2024). What is Quality Control. [online] Available at: https://timespro.com/blog/what-is-quality-
control-a-comprehensive-definition-and-overview [Accessed 19 Feb. 2025].
Usmani, F. (2022). Common Cause Vs Special Cause Variations |. [online] PM Study Circle. Available at:
https://pmstudycircle.com/common-cause-vs-special-cause-variations/.
Usmani, F. (2024). Scatter Diagram (Correlation Chart): A Guide with Examples | PM Study Circle.
[online] PM Study Circle. Available at: https://pmstudycircle.com/scatter-diagram/.
Wallstreetmojo.com. (2024). Control Chart. [online] Available at: https://www.wallstreetmojo.com/control-
Prepared by: Higher Education Qualifications (HEQ) Approved by: HEQ, QDAM (HE)
Version 1.0 – November 2017
DCL 1 – Public (Unclassified)
28
Higher Nationals
chart/.
Webflow.io. (2020). What Is A Sample Size? A Guide to Market Research Sample Sizes | starlight analytics
blog. [online] Available at: https://starlight-stage.webflow.io/article/what-is-a-sample-size [Accessed 19
Feb. 2025].
Whatissixsigma.net. (2019). Seven Basic Tools of Quality. [online] Available at:
https://www.whatissixsigma.net/7-qc-tools/.
www.learnleansigma.com. (2023). Understanding Process Capability Index (Cpk) [With Calculator] -
Learn Lean Sigma. [online] Available at: https://www.learnleansigma.com/control/understanding-process-
capability/.

Prepared by: Higher Education Qualifications (HEQ) Approved by: HEQ, QDAM (HE)
Version 1.0 – November 2017
DCL 1 – Public (Unclassified)
29