QUESTION 1

1
1 Exploratory Analysis of the data

The given information on the problem are as follows:

1. Field failure data – Time to failure
2. Two dominant failure mechanisms – V and W
3. Censored data

Before moving to choosing the appropriate plotting methods, it is important to know the type
of data that has been provided to us. According to O'Connor et al. (2012,73), Data are of
different types, namely:
1. Complete data
2. Censored data
3. Right-Censored data (Suspended)
4. Interval data
5. Left censored.

Figure 1.1

Here the data that is given is Right-Censored(Suspended) as show in in Figure 1.1. Right
censoring is most used censoring method in the case of life data, these set of data are made of
units that did not fail.

The second step is the ranking of the data, it becomes complicated here when dealing with
the censored data.

Here, let’s use the Hazard plotting method to analyse the data, because of the following
reasons:
1. Hazard function gives insights of how the failure rate changes over time.
2. Hazard plotting can reveal patterns in the failure data.
3. Easy to compare with the bathtub curve to find at which phase the failure occurs.
4. Hazard functions provide a dynamic view of survival data.

Steps to rank the data:

1. Assuming the data fits Weibull distribution.
2. Arrange the data in ascending order according to time to failure.
3. Reverse rank the data.
4. Find the hazard value by inversing the rank number.
5. Derive cumulative hazard value.

2
Fig 1.2. Reverse order and cumulative Hazard table Fig 1.3. Cumulative Hazard-
Time to failure

Fig 1.4. Cumulative Hazard-Time to failure graph

Before analysing the failure modes and reliability of each mode, let us discuss some of the
reliability concepts.

3
2 Bathtub Curve

Figure 2.1 Bathtub Curve (O'Connor et al., 2012)

The bath-tub curve functions as a representation of the reliability

performance inherent in components. It entails a thorough examination of the
reliability performance exhibited by a substantial and homogeneous sample of
items entering service, typically commencing at time zero. When observing
these items throughout their entire lifespan without replacement, three distinct
phases become apparent.

In the initial phase, denoted as infant mortality or early failures, the population
encounters an elevated hazard function that gradually diminishes. This phase
corresponds to the burn-in or debugging period, during which less robust items
are identified and subsequently eliminated.

Following the initial phase, where weak components have been addressed and
errors rectified, the remaining population transitions into a phase marked by a
relatively constant hazard function which is equal to 1. This phase is formally
recognized as the useful life period. The third phase where components enter the
wear-out phase with hazard function increasing in nature.

4
3 Weibull Distribution
Weibull distribution is the most widely used technique in reliability engineering.
The main reason we use Weibull distribution is its flexibility in interpreting the distribution
parameters and their relation to the failure rates and the bathtub curve concept as shown in
figure 2.1.
Weibull distribution uses two parameters Beta and Eta (Mean life).
which is written as,

Figure 3.1 (O'Connor et al., 2012)

5
4 Analysing Failure mode V

Figure 4.1

Figure 4.2
Shape parameter (Beta) and scale parameter (Eta).

As we can observe from the fig 4.2 above, Beta value is less than 1, which can be compared
with the bathtub curve and said that the failure occurs in the infant mortality phase.

This could be because of the two following reasons:

1. Issue with the product design
2. Defect in manufacturing

6
5 Analysing Failure mode W.

Figure 5.1

Figure 5.2
Shape parameter (Beta) and scale parameter (Eta).

As we can observe from the fig 5.2 above, Beta value is more than 1, which can be compared
with the bathtub curve and said that the failure occurs in the wear-out phase.

This could be because the product is designed in such a way to serve the maximum possible
lifetime and then start failing in the wear out phase. We can conclude that there is no issue
with the designing and manufacturing process, as it serves the purpose. But there could be
room for improvement if the design or the process are improved.

7
6 Reliability Function R(x)

Figure 6.1

Figure 6.2

8
Survival function is concerned with the probability of the component will survive for a stated
interval of time.

(O'Connor et al., 2012)

As we can observe that the product fails from 0th hour because of the failure mechanism V.
Whereas, in failure mechanism W, it maintains a constant and starts to fall from the 150 th
hour. Which is less than the mean-life value. The probability of failing for mechanism W
happens from 150 hours to 371 hours, whereas for mechanism V it tends to fail from time 0.

9
 QUESTION 2

10
1 Introduction

Though coming up with a whole new design and technology for a fuel pump has its
own merits, the biggest challenge is the unavailability of any previous data. A good QRM
plan and strategy should help overcome this challenge.

To generate a structured QRM Plan and to ensure the product’s availability to the customers,
the Product Life Cycle concept can be used as shown in Figure 1.1. In this way, each phase
will need different sets of QRM tools and techniques. Let’s delve into each phase of the cycle
and discuss the appropriate tools and techniques that can be used.

Figure 1.1
Source:(Reliability Section 14, WMGAnon)

11
2 Design & Development

2.1 The Kano Model

Since the company has decided to go with a whole new design for the fuel pump all
together, it is sensible to listen to customers’ requirements and incorporate possible
requirements while the new product is in design phase.

Though the contribution of the pumps is not direct, but it plays a very crucial role in the
system.

The Kano Model as shown in Figure 2, developed by Professor Noriaki Kano (1984),
for understanding customer preferences and prioritising features based on their impact on
satisfaction.

Figure 2

For example, let’s assume and prioritise the customer requirements as follows.
1. High Fuel Efficiency
2. Good Performance
3. Cost effective for the automotive manufacturers.

Generally, A working fuel pump, is a Basic Need (Must-be Quality), where their absence
leads to dissatisfaction, but their presence doesn't necessarily increase satisfaction. But by
considering the requirements when designing the new product, the company can create a

12
better one. This will lead the customers to shift to the One-Dimensional curve, where the
direct or indirect contribution of the product is directly proportional to the customer
satisfaction.

2.2 Six Sigma

Designing and manufacturing a product from the ground up offers the company a distinct
competitive advantage and greater control over expenses, materials, and processes. While
there are alternative quality methodologies, Tennant (2017) suggests that Six Sigma stands
out by not merely being a methodology but also providing a well-defined vision and
objective. TQM, on the contrary, though plays a significant role in the Six Sigma approach
but lacks a clear vision and plan, causing participants to progress slowly and somewhat
uncertainly. Additionally, TQM lacks a robust metric for process monitoring. The goal of Six
Sigma extends beyond just waste and resource reduction; it also aims to simultaneously
enhance customer satisfaction.

Six Sigma has two main methodologies, DMAIC method for improving existing process and
Design for Six Sigma (DFSS) for designing a product and process so that defects and errors
never arise in the first place.

According to Cudney & Furterer (2012), DFSS is defined as, a methodology that can be used
to systematically design new products, services, or processes.

Therefore, incorporating DFSS (DMADV) will be the most appropriate tool for the new fuel
pump that is being newly designed.

Let’s discuss the Design of a Walker—A Design for Six Sigma Case Study and look how they
have used the Kano analysis and Quality function deployment to design a new Walker.

13
During the Defining phase, which comes before designing, the organisation decided to know
what the customer wants.
As a first step towards this, they prepared a questionnaire, to which the
customers replied through online forms and feedback. The second step is listing technical
descriptors, characteristics that will affect more than one of the customer requirements, in
development or production. Then compare the two (customer requirements to technical
descriptors) and determine relationships. Finally, Developing the positive and negative
interrelated attributes and identify “trade-offs.”

Customer needs are then analysed using QFD to identify functional requirements and
translate them into technical requirements to develop a design. The subsequent phase
involved transforming the features into measurable metrics aimed at determining the optimal
specifications to fulfil customer needs. To address design requirements, the team employed
the Kano analysis method.

For the fuel pump case, it is recommended to use QFD and KANO analysis to translate
customer requirements int technical requirements.

14
2.3 Design for Reliability (DfR)

Figure 3
Cost of design change

According to O'Connor et al. (2012), decisions during the designing process

has a strong influence on the reliability of the product. Defects in design affects all items
produced and later becomes expensive to correct as the development proceeds. As clearly
shown in the figure 3.
The expenses associated with errors and modifications in design significantly rise throughout
the progression of the product development cycle.

Therefore, applying reliability testing tools during the designing phase will help to cut down
costs by enabling them to make quick design improvements if so needed.

Design for Six sigma and Design for Reliability, both share similarities and some
methodologies. Meanwhile, the distinct difference between the two is, DFSS work on the
design to achieve optimised manufacturing and to reduce cost and scrap.
Whereas, DfR works on the design aiming to make the overall product maximum reliable to
the customers.

15
The common tools and techniques to ensure product reliability are listed below:
1. FTA
2. FMEA
3. RBD

Also, there are various reliability testing tools that can be used during the designing and
development phase of the fuel pump, namely:
1. Reliability enhancement testing or HALT
2. Accelerated Life Testing
3. Step-Stress Test
4. Reliability Growth/Enhancement h Planning
5. Failure-Free Test or demonstration test (Post design)

Let’s discuss about each tool in detail.

Fault tree analysis

Using fault tree analysis enables the organisation to systematically evaluate potential
failures and its modes and causes within the system. It is also a time-consuming task, where
people from various departments have a coordinate and collaborate to get accurate outputs.
FTA falls under top-down method, where different flow charts must be constructed for
different top events, in some cases, software is recommended to create an effective output.

Failure mode and effects analysis

Sounds like FTA but has a different approach and a common goal in terms of fault
analysis. Unlike FTA, FMEA is a bottom-up method. Where, each level will lead to the next
high level of failure mode, resulting in filtering and identifying the causes effectively.
But the disadvantage of this method is that the output is very large even for a simple system.
Also requires expert knowledge to predict every aspect of causes and failure modes to use it
to get desired output when needed.

16
Reliability Block Diagram
RBD can become very difficult to construct for a complex system, but in this case of a
fuel pump, it can be used to identity the cause and failures in more visually appealing way.
Like FMEA, RBD is also a bottom-up method.

Accelerated Life Testing

Accelerated Life Testing, as the name suggests, after predicting the potential
causes and failures and failure modes using appropriate tool, the prototype or the first
working model should be put into a time accelerated testing system. This will cut down the
testing time, by giving stress on the potential failure modes. This method is handy when it
comes to time bound projects, where testing should not take more time as well as cost money.
This method can be used only when there are non-linearities in the system.
Step-Stress testing is an alternative for the ALT, with even more accurate data, by connecting
the components to a step by step increasing stress levels. This will allow us to get more
accurate data in every level of stress, thus giving accurate robust of the design and reliability.
HALT
This method is commonly using the step-stressing methodology, with temperature and
vibrations. HALT is not focused on real world simulations, but it just stimulates predicted
failure modes at faster pace. Reliability measures cannot be obtained during the test.
Additionally, it cannot be used to find how the design can be more robust and can be only
used if the failure mode is well-known beforehand. Also, HALT had to be conducted in a very
special environment, which can be cost and time consuming.

Demonstration test
These tests are often done in a specially created environment, which saves times, but
at the same time, may not replicate the exact real-world conditions. These tests are usually
done before the final design enters the production phase. These are just an acceptance test,
tells us very little about the design itself. It is extremely costly, by which it is advisable to
concentrate on a well-managed reliability growth programme to avoid demonstration testing.

17
3 Manufacture

Any business’s goal is to make profits and to keep the investors happy. Profit is
influenced by various factors, but the most important aspect among those are costs. A big
chuck of the business’s operating costs goes into manufacturing. By optimising
manufacturing process, costs can be brought down. Optimising costs need not necessarily
means compromising on the quality. It is possible to have high standard quality but keeping
the costs low. This creates a win-win situation for customers, business, and the investors.
Process control tools will enable the business to gain that ability.

One such approach is Six-sigma’s DMAIC, which is a process improvement methodology,

used to minimise waste and maximise efficiency.

3.1 Statistical process control

One important component of Six sigma (DMAIC) is the statistical process

control (SPC), which supplies powerful metrics, analysis tools and control mechanisms
(Tennant, 2017).

The key advantages of SPC are,

1. SPC can translate business processes into numbers in real time.
2. Evolving and actively discussed tool among business industries in terms of quality
and total customer satisfaction.

SPC tools that can be used in manufacturing a fuel pump are listed below:
1. Cause-and-Effect Diagrams (Fishbone diagram)
2. Process Capability Analysis

Cause-and-Effect Diagrams, also known as fish bone diagram, specifically Cause

Enumeration type can be used to list down all 5 pillars of the organisation and deriving

18
causes and failures in a fish bone’s shape. This tool is simple to create and highly useful by
giving flexibility but may get complex when then complexity of the components is high.

Process capability analysis is a statistical approach to evaluate the manufacturing process and
its ability to consistently produce products within tolerances. The two commonly used indices
are Cp and Cpk.

 Cp (Process Capability Index): Cp = (USL - LSL) / (6 * σ), where USL is the upper
specification limit, LSL is the lower specification limit, and σ is the estimated process
standard deviation.
 Cpk (Process Capability Index with respect to the centre of the specification): Cpk =
min[(USL - μ) / (3 * σ), (μ - LSL) / (3 * σ)], where μ is the process mean. (O'Connor
et al., 2012)

19
4 Testing

4.1 Environmental stress screening

To avoid infant mortality problems like what happened in with Mechanism V in the
fuel pump, we can use Environmental stress screening on the manufactured products.
Theoretically, if all the manufacturing process are under control, there will be no necessity for
the stress screening or monitoring, whereas it is near to impossible to avoid variations in the
reality.
Also, there is an intense stress screening which is known as, highly accelerated stress
screening. HASS is generally performed in the development phase, so that the stresses can
highlight the root cause and therefore, will enable the engineers to work till they eliminate the
failure mode.
ESS should be carefully used, so that, the fuel pump is not stress so much, which will then
reduce its useful lifetime, which will ultimately lead to customer dissatisfaction.

5 Use
It is also advisable that the organisation, frequently collect real life data through
customers, or field test data after the products has start being used. Many manufacturers
collect this data to improve design, process and to reduce defects in their future attempts. For
example, a mobile company will collect usage stats as part of their corrective programme. For
fuel pumps, these data can be collected from the manufacturer, who collects the data from the
service chains that they have. Listening to customer feedback, getting data from vehicle
service outlets, and random sampling analysis serves best to give inputs to the future
improvements.

20
6 Maintenance

In the realm of fuel pump maintenance, two usable approaches are Reliability-
Centered Maintenance (RCM) and Total Productive Maintenance (TPM).

These methodologies play important roles in optimizing the performance and longevity of
fuel pumps.

Reliability-Centered Maintenance (RCM) is a systematic strategy that focuses on identifying

the most effective maintenance tasks to enhance reliability. In the context of fuel pumps,
RCM involves a thorough analysis of the pump's critical components and functions,
determining the appropriate maintenance actions to prevent failures, and ensuring cost-
effective operational efficiency.

Total Productive Maintenance (TPM) is a comprehensive approach that aims to maximize the
overall effectiveness of equipment and minimize downtime. For fuel pumps, TPM involves
fostering a culture of proactive maintenance, continuous improvement, and employee
involvement. It encompasses preventive maintenance, predictive maintenance, and
autonomous maintenance to ensure optimal pump performance throughout its lifecycle.

Maintenance can be viewed in other perspective as well, that is asset maintenance.

For example, the direct machinery involved in manufacturing, and the indirect assets which
support the manufacturing and the operations of the organisation.

Among three Reliability Centered Maintenance types, preventive maintenance can be

followed for the fuel pump scenario to cut down the operating costs and avoid replacing
costs.

TPM take a holistic approach to achieve operational excellence and reliability, whereas RCM
focuses on strategically optimizing maintenance tasks based on critical analysis and tools.

21
7 Asset management
When it comes to asset management, the main factors affecting the process are, funds,
acquisition, operation, and disposal. Managing these four effectively with right tools and
predictions will help to keep the costs down and enables the business to get maximum output
from it.
Let’s discuss in detail from different aspects.

From leadership and organisational context

Clarify leadership roles, responsibilities, and authorities to establish a robust
governance structure for asset management.
Cultivate a proactive organizational culture that values asset management as a strategic
function, promoting accountability and continuous improvement.

Operation
Implementing real-time monitoring for the manufacturing and supporting assets to
optimize utilization and ensure alignment with production demands. These data collected can
also be used to test the reliability of the machineries, which can serve as deciding factor when
replacing the machinery in the future, whether to stick with the same technology, or to
change.
Utilize condition-based monitoring and predictive maintenance strategies to enhance
performance and reduce downtime.

Technology Integration
Implementing advanced technologies, such as asset management software and
Internet of Things (IoT) devices, to enhance real-time monitoring, data-driven decision-
making, and predictive maintenance. Which could be expensive, but after a thorough
financial break-even analysis, if it fits and profitable for the organisation, it can be done.

Planning
22
 Developing a detailed lifecycle plan for fuel pump assets, covering acquisition,
operation, maintenance, and disposal.
Allocate budgets strategically, striking a balance between capital and operational
expenditures to ensure the sustained performance of assets.

References

British Standards Institution (2014). Asset management : BS ISO 55000:2014, BS ISO

55001:2014 and BS ISO 55002:2014. London: Bsi Standards Ltd.

Anon Quality, Reliability and Maintenance (QRM). [online] WARWICK MANUFACTURING

GROUP. Available from: https://moodle.warwick.ac.uk/mod/resource/view.php?id=2023795
(Accessed 2023).

Anon ISO 55000:2014. [online] Available from:

https://img1.wsimg.com/blobby/go/b653c9ee-535c-4528-a9c5-bb00166ad0dc/
downloads/1bsmknus2_894046.pdf (Accessed 2023).

Cudney, E. A. & Furterer, S. L. (2012) Design for Six Sigma in product and service in
development: applications and case studies. Boca Raton, Fla: CRC Press.

O'Connor, P. P., Kleyner, A. & Wiley, I. (2012) Practical reliability engineering. 5th edn.
Oxford: Wiley-Blackwell.

Tennant, G. (2017) Six Sigma: SPC and TQM in manufacturing and services. New
York;Abingdon, Oxon;: Routledge.

23
 Module Reflective writing

Quality Reliability and Maintenance module had left a very big impact in me I would say.
Before the module, my knowledge about the QRM tools and techniques were near to zero,
but later realised QRM is not just about tools and techniques, but it is a perspective on how
you approach problems within the product, organisation, and operations.
The modules and materials were so comprehensive and to the point, which grabbed my
interest even more and was easily able to compare it with my previously worked organisation.
I have already started drafting the concepts like Lean Six sigma and cause effects diagram
around our organisation, which they were struggling for years to have control over.
As a traditional and medium scale organisation, we always have thought how well the
statistical methods and tools would work. This is because of the lack of knowledge I would
say. Since, we directly deal with Japanese manufacturers ,one big challenge for us has always
been reducing defects and variability in manufacturing and in the precision of the
components. Implementing six sigma, and working towards it, I strongly believe I can bring
in some positive change in the organisation in terms of manufacturing.
The activity-based asset management lecture is what I found so interesting among all
sessions. Till then, I was in a fixed mindset, where I believed that not to touch until its
broken. But never saw the other dimension of it. Practically, on hands experienced how
strategically analysing the cost associated with replacing versus maintaining it would impact
the overall business significantly.
All these assumptions have come from the environment and a fixed mindset is what I believe.

My manager always would say about being practical and street smart, and he kind of hates
people who speak so many theories and concepts. I think that’s where the organisation and its
people are falling behind. Implementing and educating the personals in the organisation about
the right tools and techniques will show visible results, both physically and on paper.
I would compare the scenario of how I approached quality before, with a person trying to
unscrew a star screw with a flat screwdriver. It still works but will consume time and energy.

24
But using right tools for right problems will not only save time and energy, but also it saves
money, which is what is businesses are operating for, the PROFIT.

25