Course: Total Quality Management

Session: December 2019

Question No. -1

Answer: Quality Function Deployment (QFD) is a process and set of tool which understands and
defines the customer requirements to convert them into detailed engineering specification and plan the
process to produce the product which satisfies the customer requirements. Quality function deployment is
a graphical analysis technique that portrays customer needs and expectations, how these needs and
expectations will be satisfied, and tradeoffs between conflicting needs and expectation. It emphasizes on
the needs of maintaining quality in processes as well as in products. It is the quality system which brings
the close linkage between end users with other organizational processes like marketing, designing,
quality, production, manufacturing and engineering.

The concept of QFD was first introduced by Yoji Akao, a Japanese quality planning expert in 1966. The
main aim of this concept is to work out a method which would support the manufacturers in designing
products as per the choice of customers. QFD was first used in the Mistubishi’s Kobe shipyard in 1972.
Later, QFD was adopted by different Japanese firms. The USA automotive and electronic companies
introduced this methodology in 1980s

QFD helps organizations in many ways as to design a design a product well, a design teams needs to
know what it is they are designing, and what the end-users will expect from it. The major advantages of
QFD in which it helps the organsational causes are:

1. Customer Driven Production/Customer Focused: QFD methodology focuses on the wants

and needs of the customer not on what the organization may believe the customer wants. The
voice of customer is translated into the technical design specifications. Every product
development activities are based on the customer’s requirements. During QFD process these
specifications are driven from machine level to system, sub-system and component level
requirement. Finally the design process is control throughout the process to assure the customer
requirements are met. For example companies like General Motors Corporation and Ford Motors
adopted to understand the need of their customers.

2. Competitive Advantage: In today’s tough competitive environment it is very necessary for

the organizations to stay ahead of their competitors to successfully sustain and succeed in market.
QFD methodology provides the organizations a competitive advantage over their competitors, as
the competitor’s products and services are analyzed and compared with their own product and
services. This timely analysis puts the organization ahead of their competitors. For example
Japanese companies advantaged using QFD with compare to their US business competitors in
America which leaded for American companies to use QFD in their industries.

3. Reduced Development Time: As through QFD processes we acknowledge that what is the
requirement of the customer then now it is the time to concentrate on making amendments and
changes of product in the design stage itself. So, there will be no design changes in the preceding
 phases of the product development and also the time can be saved which was otherwise to be
used in developing other un-necessary (Not as the needs of customers) systems. This leads to
reduction in development time of the product.

4. Lower Development Cost and Improved Quality: As using QFD we know and develop the
product as per the customer needs it reduces the cost required in warranty claims, support cost,
design changes cost as the alterations are made initially in the product development phase and the
needs of the customers are considered while developing the products. For example Toyota
became successful by implementing QFD in their company worldwide with respect to their
competitors.

The above are the key areas where QFD helps a cause in the organizations, apart from the above given
benefits the implementation of QFD provides organizations a improved level of quality as the quality is
measured and assured during the development phase itself. When QFD is used throughout product and
process development it gives a clear specification of the customer needs and translates these all the way
through to the necessary process and quality controls. QFD increases the communication level between
the organsational function as the needs and requirements are well stated and delivered to the respected
departments. Hence we can say the QFD is not only good for assessing the customer requirements but it
increases the quality of the functional process and product also.

Question No. - 2

Answer:

Sampling is a procedure in which a predetermined numbers of sample/products are selected from a larger
population and testing is done as per requirement to verify that whether the product conforms to
requirement or not. By sampling individual product from larger populations the characteristics of whole
lot are decided. A population can be defined as a whole that includes all items and characteristics of the
research taken into study.

Sampling from a process prospective can be defined as the sample taken from a process which determines
the variation in the product from the actual requirements/ specifications in the current process. Sampling
in process helps us to know that the process is controlled and if the sample taken from the process defers
from the requirement then we came to know the variation in the process which enables us to rectify the
process to get the required results.

Acceptance Sampling: It is a statistical measure used in quality control. It is a process of

accepting and rejecting the lot based on the sample drawn from the lot. In this process a sample is chosen
from the given lot of finished products which is then evaluated on the basis of predetermined standards
and sample is accepted and rejected based on the result of sample tested. It allows a company to select
and check the finished lot of product. As every product cannot be checked therefore a sample is taken
from the lot which is evaluated to accept or reject the lot. For example from a process prospective a
gearbox is made up of many gears on a shaft and if any of the gear assembled on shaft is not as per
required quality then the complete gear box will be rejected or become faulty therefore a gear is taken
from the lot before assembling gearbox to verify whether that gear meets the predetermine standard or
not, this will help us in saving time and cost which have been occurred if the faulty gear has been placed.
Steps involved in Acceptance Sampling are:

1. Select a sample from a lot using various methods such as random sampling or quota sampling.
2. Evaluate the sample on the predefined criteria.
3. If the sample meets the predefined criteria accept the sample else reject it.
4. Record the results of the acceptance sampling for future use.

Double Sampling: This method is used when the singe sampling fails to provide the conclusion that
whether the lot is to be rejected or accepted. For example two samples are taken from a lot and one of the
sample fails to meet the predefined criteria and another sample passes, then it would be difficult to
conclude that whether the lot should be rejected or accepted therefore double sampling method is used.
Following steps are involved in this method:-

1. Collect a random sample from a lot for inspection.

2. Inspect the sample for required characteristics.
3. If the number of defects in sample are equal to or less than set in parameter (e.g. C1) then accept
the lot.
4. If the number of defects in sample are more then set in parameter (e.g. R1) then reject the lot.
5. If the total number of defects are more than C1 but less than R1 then select another sample.
6. If the total number of defects in both the samples are less than the acceptance number C2 then
accept the lot.
7. If the total number of defects in both the samples are more than the rejection number R2 then
reject the lot.

Sequential Acceptance Sampling: It is further refinement of double sampling plan. In this sampling
one or more samples are taken randomly from the lot and tested one by one individually. In this sample
size is not set in advance. After selected the samples the hypothesis test is taken that whether the taken
samples attain the predefined characteristics or not. Each time sample is selected a decision is taken to
accept the lot or reject the lot or to continue sampling. For example if in a process of testing headache
medicine, the medicine works ones but it doesn’t work on other time then the medicine get rejected as it
doesn’t get the results as A plan in which the consumer randomly selects items from the lot and inspects
them one by one.

Question No. -3

Answer 3a: The sampling is done to verify that the product conforms to the predefined requirements.
Instead of checking each product samples are taken from the whole production lot as per pre-defined
methods.

As given in the question the 10 samples are taken from a population size of 50 products each one sample
is selected from lot of 5 finished products. Here using p-chart to measure the proportion of defective items
in a sample.

The fraction defectives are calculated as shown in table below:

 Fraction Defectives (p) =
Sample
No. of Defects No. of defectives /sample
No.
size
1 2 0.04
2 3 0.06
3 1 0.02
4 2 0.04
5 3 0.06
6 4 0.08
7 2 0.04
8 1 0.02
9 3 0.06
10 2 0.04
23 0.46

Upper control limit and lower control limits can be calculated by using the following formula

UCL = 𝑃̅ + 𝑧𝜎𝑝

LCL = 𝑃̅ − 𝑧𝜎𝑝

Where,

z = Standard normal variable (set as 3)

𝑃̅ = Sample proportion defective

𝑃̅(1−𝑃̅ )
𝜎𝑃 = Sample deviation of average defective proportion = √ 𝑛

n= Sample Size

𝑇𝑜𝑡𝑎𝑙 𝑁𝑜. 𝑜𝑓 𝑑𝑒𝑓𝑒𝑐𝑡𝑖𝑣𝑒𝑠

𝑃̅ =
𝑇𝑜𝑡𝑎𝑙 𝑛𝑜. 𝑜𝑓 𝑖𝑡𝑒𝑚𝑠 𝑖𝑛𝑠𝑝𝑒𝑐𝑡𝑒𝑑

23
𝑃̅ = = 0.46
50

̅̅̅̅̅̅ ̅̅̅̅̅̅)
0.46(1−0.46
𝜎𝑃 = √ 50
= 0.0705

Now as we have the values of 𝑃̅ and 𝜎𝑃 therefore by putting a these values in the above given formulas of
UCL and LCL we will get

UCL = 0.46 + 3(0.0705) = 0.671

LCL = 0.46 – 3(0.0705) = 0.249

Answer: 3b A process capability index is an analytical measure that compares the performance of a
product with its predefined specifications. 𝐶𝑝 is the relation between the specification range and the
process width i.e. the spread of the distribution. The formula for 𝐶𝑝 is as follows:

𝑈𝑆𝐿−𝐿𝑆𝐿
𝐶𝑝 =
6𝜎

As values of USL and LSL is given in the question which are

USL LSL Recorded

Results
47 25 49
47 25 19.4

47 25 47.6

𝜎 can be calculated from the following formula

1
𝜎 = √𝑁−1 ∑𝑁
𝑖=1(𝑥𝑖 − 𝜇)
2

𝑆𝑢𝑚 𝑜𝑓 𝑟𝑒𝑐𝑜𝑟𝑑𝑒𝑑 𝑅𝑒𝑠𝑢𝑙𝑡𝑠

𝑊ℎ𝑒𝑟𝑒, 𝜇 =
𝑁𝑜.𝑜𝑓 𝑅𝑒𝑠𝑢𝑙𝑡𝑠

49+19.4+47.6
𝜇= = 38.667
3

N=3

Calculating 𝑥1 , 𝑥2 & 𝑥3 i.e.

𝑥1 = (49 − 38.667)2 = (10.333)2 = 106.771

𝑥2 = (19.4 − 38.667)2 = (−19.267)2 = 371.217

𝑥3 = (47.6 − 38.667)2 = (8.933)2 = 79.798

Now,

∑𝑁
𝑖=1 = (𝑥1 + 𝑥2 + 𝑥3 )

∑𝑁
𝑖=1 = (106.771+371.217+79.798) = 557.786
 1
𝜎 = √𝑁−1 𝑥 557.786

1
𝜎 = √3−1 𝑥 557.786

𝜎 = √278.993 = 16.700

By putting the values USL, LSL and 𝜎 we can find 𝐶𝑝 i.e.

47−25
𝐶𝑝 =
6𝑥16.700

𝐶𝑝 = 0.22

As value of 𝐶𝑝 is less than 1 which means process capability is good

𝐶𝑝𝑘 is the process capability indicator which considers the inconsistencies that are observed in 𝐶𝑝

𝑈𝑆𝐿−𝜇 𝜇−𝐿𝑆𝐿
𝐶𝑝𝑘 = min of ( ),( )
3𝜎 3𝜎

Putting the values of 𝜇 = 38.667

USL = 47, LSL = 25

& 𝜎= 16.700
47−38.667 38.667−25
𝐶𝑝𝑘 = min of ( 3𝑥 16.7 ) , ( 3𝑥 16.7 )

𝐶𝑝𝑘 = min of (0.17, 0.27)

𝐶𝑝𝑘 = 0.17

The value of Cpk factor (0.17) is lower than the Cp value (0.22). This implies that the LSL is more
difficult to achieve than the USL. Non-conformities exist at both ends.