Web & Social Media Analytics Previous Year Question Paper.pdf
What and how about quality function deployment
1. international journal of
production
economics
ELSEVIER Int. J. Production Economics 46 47 (19961 575 585
What and how about quality function deployment (QFD)
C.P.M. Govers*
Eindhoven Universi O, o/" Technology,, Graduate School of Industrial Engineering and Management Science, Frits Philips blstitute./or Quali(v
Management, P.O. Box 513, 5600 MB Eindhoven, The ,Netherlands"
Abstract
QFD is more a process than just a tool for product as well as production process development based on the concept of
Company Wide Quality Control. Essential characteristics are: customer orientation, team approach and a way of
concisely structuring communications and linking together information. The methodology is described to discuss
experiences and some implementation problems.
Ahhough first used by the Japanese, experiences from "Western" companies support the results of better products and
production planning. Key factor for success is the Cross Functional Management approach.
Kevwords: Quality function deployment methodology: Practice in the Netherlands; Implementation aspects: Dutch
quality award
1. Introduction philosophy and concepts that are the roots of this
method. Approaches to quality based on the con-
Quality function deployment (QFD) is a cus- cept of Total Quality Control (TQC) as introduced
tomer-oriented approach to product innovation. It by Feigenbaum [1] is fundamentally different from
guides product managers and design teams the Japanese T Q C concept. In this vision T Q C is
through the conceptualization, creation and realiz- " C o m p a n y Wide Quality Control". It is more
ation process of new products. Q F D supports de- comprehensive and characterized by deploying cus-
sign teams to develop products on a structured way tomer desires horizontally and vertically through-
that relates market demand via engineering speci- out the organization (Japan Industrial Standard
fications to parts specifications and to production Z8101 - 1981).
process variables and thus to production opera- The origins of the Japanese C W Q C are the same
tions planning. concepts of statistical quality control (SQC) and
To discuss possible improvements of develop- T Q C as they were brought over from the U.S. after
ment processes by Q F D we need to understand the World War II but they are deployed as means of
securing quality accountability at each level in the
organization [2] and combined with market ori-
entation. The "voice of the customer" drives all
* Tel.: I 4 31/40) 472285/47 21 70. Fax: ( + 31/40) 45 1275. activities.
0925-5273/96/$15.00 Copyright ,.~'~1996 Elsevier Science B.V. All rights reserved
SSDI 0925-5273(95)001 13-1
2. 576 C.P.M. Govers/lnt. J. Production Economics 46-47 (1996) 575-585
Often in many "Western" companies the execu- tion" of CWQC [3]. It means that although quality
tive's or engineer's voice dominates because there is professionals are important participants to facilit-
still a strong influence of the ideas of scientific ate the QFD process, the marketing, development
management. The separation of development and and manufacturing professionals play an even more
preparation from implementation and doing, as vital role.
advocated by Taylor, brought us to organizations An organized QFD approach follows all the
subdivided into more or less isolated functional rules for project management, which means project
departments, staffed with specialists for quality, definition, team selection and is not restricted to
cost and delivery (QCD-aspects), paying attention a single action within just one department. Teams
to output characteristics with separated perfor- should be cross-functional, expertise oriented and
mance objectives. consisting of six to eight members of comparable
Because of that orientation the TQC concept is peer levels.
too often exclusively directed to the quality of QFD is a process that can help companies to
product and service in a proper balance with costs make the key trade-offs between what the customer
and usually identified with manufacturing and as- wants and what the company can afford to build. In
sembly activities. US and European companies put essence, QFD encompasses same activities that
a greater emphasis on problem solving and effici- people did before but it replaces erratic, intuitive
ency improvement during the implementation and decision making processes with a structured meth-
production stage. In the CWQC approach more odology that establishes all relevant information
effort is put into designing quality at the develop- and experiences that are available throughout the
ment stage and the QCD aspects are managed by organization. As such, QFD lays a basis for organ-
interrelation. Roughly, the differences concerning izational learning.
product development can be illustrated as depicted In general, the product development process
in Fig. 1. from customer-needs to manufacturing process op-
The CWQC philosophy is characterized by cus- erations, can be outlined by a step by step approach
tomer orientation, cross functional management marking the points at which the requirements for
and process rather than product orientation. It intermediate results are established and go-or-no-
refers to quality of management and the quality of go decisions can be made. Usually we can discern
work being done. Within that concept QFD pro- four phases:
vides a means of translating customer requirements
for each stage of product development and produc- Strategy
tion (i.e. marketing strategies, design, planning, and concept ~ Product Process Manufacturing
process development, production control). It is Definition design ~design ~ operations
a mechanism that serves as an "operational defini-
In the strategy phase, product policy and deter-
mination of the customer will be established and
the customer needs are translated into a product
Europe Japan
concept. The design requirements (WHAT's) serve
market research market research as input to establish the component characteristics
product definition productdefinition
establishmentof (HOW's) of the product design which on their turn
design/cOStengmeermg,
. engineering1 QCD-targets serve to define the process plans and next the
I
design engineering purchasing manufacturing process operations. Because of the
purchasenegotiation policy
cost evaMation i 1
complex relationships between the inputs and out-
I
production
I
cost reduction production puts, these relationships are mapped into matrices.
The basic structure is depicted in the relationship
~ efficiency)
-~orrective actions "7
improvements ~ ontinuous improvements[ matrix of Fig. 2.
The flexibility of the method of approach allows
Fig. 1. for adding any other information which may be
3. C.P.M. Govers/lnt. J, Production Economics 46 47 (1996) 575- 585 577
when the "voice of the customer" gets to be de-
ployed to the most detailed level of manufacturing
operations. This means deploying all phases al-
though it is possible to achieve substantial benefits
by implementing Q F D only in the first phase.
There are several useful extensions to the basic
QFD-charts which greatly assist in the trade-off
relationshipmatrix
procedure to establish the values of the How-
Much's. Decisions will be based on all the informa-
tion normally available: business and engineering
judgement as well as various analysis techniques.
Fig. 2. Basic structure QFD. Once the first chart is completed, the downstream
stages will be determined more and more by
specific technological characteristics of a particular
PHASE I organization. So the most elaborated importance-
rating systems and assessment tools arc tailored to
the first stage.
Here, in c o m m o n with other generalized intro-
ductions to Q F D , we will discuss some of the most
important and well known expedients by building
a ~'House of Quality" as it was named by Hauser
and Clausing [4].
2. Q F D methodology: "The House of Quality"
Fig. 3. Cascade of QFD charts.
Starting a QFD-project, team members should
reach agreement on issues as:
useful to the decision making. When viewing which product or product characteristic are we
a Q F D - c h a r t the first time, look for the What's going to focus on
How's relationships. Each H o w will be appraised - who do we consider as our customers
to set target goals or values, the How-Much's we - which competing products will be used as a refer-
want to achieve. These How-Much's should be ence for product evaluation
measurable as much as possible. how does the Q F D approach fit into product
Measurable items provide more opportunity for and process planning.
analysis and optimization. Using Q F D charts the In the initial phase the scope of the project has to
outlined development process can be depicted in be established and should be communicated to and
four charts (Fig. 3) although in actual use as many agreed upon by management. Management sup-
levels of charts as necessary may be used. port is always very important because all available
Of all the steps in the total production develop- expertise as well as market information will be
ment process, none deserves more and receives less required. In order to turn a pilot project into a suc-
attention than the definition of the right product cess it is critical to select an appropriate product to
for the right customer. be employed (step 1: see Fig. 4). Try to find a pro-
This first step is the most critical part of the ject with broad appeal that may pique interest from
process and it usually is the most difficult because it several areas of the company. A first project should
requires obtaining and expressing what the cus- be simple, but not trivial, and present a real oppor-
tomer truly wants and not what we think he or she tunity for improvement. Do not try to tackle your
expects. The greatest gains of Q F D will be realized toughest problem in your first QFD-effort.
4. 578 C.P.M. Govers/Int. ,Z Production Economics 46-47 (1996) 575-585
~ Whichproduct Correlatlonmatrlx
~ ...................
,(~ customer
Wh[ch customer //~0~'~"-'"'~
Design Parameters
Competetlve sat isfaction
(~) CustomerRequ|rm/ HO'~s {~ i Benchmarklna EXCITEMENT
"k
I@' MANCE
i-21,1 ' @ L SCORES
i.......
....~) Relationshipmatrlx
~ . : ~ Target~
Benchmarklng
----~~ / degreeof
~-~BAS~CachieVem"
Fig. 4. The "House of Quality" showing the "rooms" of the
various steps in the QFD process.
/
Step 2 deals with the kind of customer to focus
on. Especially for consumer products a clear cus-
- market research
tomer profile is needed. A good description in- "k. one dimension
cludes the end-users but could also include profiles
of persons or interest-groups who influence the Fig. 5. The Kano model.
purchase decisions, e.g., retailers, consumers' asso-
ciations or public authorities (environmental regu-
lations!). To collect information about customer
requirements (step 3) various data collection The implicit features fall into two groups: basic
methods are available. and excitement features. The basic features are
For professional products it will be rather easy to expected. These include the fundamental functions
ascertain requirements. Improving a current (con- which must be present along with safety and relia-
sumer) product, than we already know a lot about bility considerations. Even if all of the basic features
the customers (market surveys, service calls, etc.). If are implemented perfectly we would not achieve
we take a new product this will be more difficult. In real customer satisfaction - we would only elimin-
that case a clear customer profile can help to esti- ate dissatisfaction.
mate what is important, less important or not im- The top curve represents the so-called excitement
portant. Sometimes a comparison between different features. Sometimes these are seemingly minor
target groups may help. items which the customers perceive as superior
Asking consumers about their requirements, value. Focus on the excitement features as sales
a distinction should be made between expressed points can lead to a major competitive opportunity.
requirements and implicit requirements. The Kano Requirements should be expressed in common
model [2] relates customer satisfaction to the de- parlance. So in the case of consumer products it is
gree to which product features (or requirements) important to use expressions like: easy to carry,
are achieved (see Fig. 5). modern look, natural sound in stead of: xx kg; yy
The straight line represents performance features. m m or zz Watts I-5].
We will be more satisfied if the performance ex- The "voice of the customer" needs to be worked
ceeds our expectations and dissatisfied if they fall out in order to gain a collective understanding.
short. Generally, only these (mostly one-dimen- Using a function tree a rough requirement can be
sional) requirements will be expressed by the cus- detailed into two or more levels (Fig. 6) to describe
tomer when we ask for. the " W H A T ' s " for the first Q F D - chart.
5. C.P.M. Govers/lnt. J. Production Economics 46- 47 (1996) 575 -585 579
level 1 level 2 the "'HOW's'" list. The design requirements result
from the translation of customer wishes into tech-
I//{, Easyto find information nical specifications (step 6). This list must be in
balance with the available expertise and the given
Easy to operate ~,Controls have logic positions time and cost frames of the project. To depict the
strength of the relationship between the What's and
' ~ Controls are easy accessible How's, symbols and/or an importance rating can
be used for prioritizing efforts and making trade-
Fig. 6. Function tree with two levels. off-decisions. Some commonly used symbols and
weighing factors are shown in Fig. 9. The 9-3-1
weighing often achieves a good spread between
( ~ , , ~)Importance Rating important and less important items, although any
r ~ - - 4 ~ (low) 1 - 5 (high) weighing system which makes sense, may be used.
I~z*l I ........
t--~--".;L~ (for 2 different groups) Scientifically it will be always possible to improve
the list but you should ask yourself whether addi-
tional information gathering will pay off in the
project.
Easy to use 3 5
The design parameters refer to concrete observ-
" (~ Dimensions 2 5
EnergyConsump. 3 3 able characteristics and methods of measurement
Natural Sound 5 2 (see Fig. 10). From an organizational point of view
sometimes it will be helpful if we arrange the char-
acteristics under headings like: mechanical, electri-
cal, software, etc.
The design parameters must reflect a valid
Fig. 7. What's list and importance rating.
measurement of customer requirements. As already
said, there u'ill be no one-to-one relationship and
The list of "WHAT's" should be sufficiently de- the interactions vary in intensity. The weighing
tailed to make judgements about the importance of and completion of the relation matrix (step 7) trans-
each item to the customers on whom we focus. late the project objectives into a technical priority-
Depending on the target group the relative import- list.
ance of the various requirements can be rated. This This also permits us to cross-check our thinking.
can be done using a scale from 1 (not very impor- Blank rows or blank columns indicate places where
tant) to 5 (very important) (step 4 see Fig. 7). In the our translation of What's into How's has been
case of obvious rating differences between the tar- inadequate!
get groups, it is necessary to consider to develop The operationalisations of How's are the HOW
tailored product types. (in Fig. 7, for example, a de- MUCH's. The How Much's should be measurable
vice for "sound freaks" or for "entertainment"l. as much as possible. If How Much's are not
Customers choose between products of different measurable or nondescriptive, then we have not
brands. Therefore, it is of strategic importance to been detailed enough in our definition of the How's
know how the products of our most important (another cross check of our thinking! 1.
competitors match up to the customer require- We want HOW MUCH's for the following
ments compared with our own product. Competi- reasons:
tive benchmarking (step 5) answers the question To determine priorities and directions for im-
"WHY" we should focus on which requirements provements of the How's (Sometimes an extra
and will allow a plan to be derived for improve- row to indicate this is added to the How-Much
ment. This comparison is shown in Fig. 8. mappingt.
The heart of the Q F D methodology in the first To provide an objective means of assuring that
phase is the generation of the design parameters: requirements have been met.
6. 580 C.P.M. Govers/Int. J. Production Economics 46-47 (1996) 575-585
Competitive Benchmarking
SCORES
poor very good
1 2 3 4 5
Easy to use 3 5
Dimensions 2 '5
EnergyConsump. 3 3
Natural Sound 5 2 .... Ul~.........:O i own product
O:''~lll ,& competitor A
i1~-~ ............
I.~ • competitor B
Fig. 8. Competetive benchmarking.
- To provide targets for further detailed develop- competitors. This kind of benchmarking provides
ment (step 8). a check for consistency of the relation matrix (step
The targets are enumerated in the bottom part of 7) and the competitive benchmarking data (step 5).
the house. For instance, a high score for customer require-
To set the targets, it is quite common to perform ment X should be reflected in high scores for design
a competitor's analysis on technical data. The parameters which are strongly related with that
benchmarking on technical performance (step 9) requirement.
reveals our technical position with respect to our Mostly you will find interdependency between
the design parameters. In the attic of the house
supporting and conflicting design parameters are
identified by a correlation matrix (step 10). Differ-
relationship symbol weighingfactor ent degrees of interaction will again be represented
by symbols. (see Fig. 11).
WEAK Triangle 1 The assignment of positive or negative correla-
tions are based on the influence of How's on achiev-
MEDIUM Circle 3 ing other How's regardless of the direction in which
the How Much values move. Positive correlations
STRONG Dot 9 are those in which one How supports another. The
other way, negative correlations are those in which
Fig. 9. one How adversely affects the achievement of an-
9. C.P.M. Govers/lnt. J. Production Economics 46 47 (1996) 575-585 583
specification setting. This was especially useful for
the experts. Each of whom who master a part of the
technologies could not formerly make use of the
sO@ "hidden" knowledge of his colleagues.
3 o A A =1 The use of Q F D revealed that:
s- Ao A - the knowledge about customer requirements was
1 OA @ O =3 insufficient
s 0 0 @ =9 data on competitors was handled incorrectly
-
2 @ O@ interdependences between technologies were
2@ A @ only partly known.
The outcomes of the approach (first phase) are
summarized/translated into one or more
"scenario's" to support management decisions
concerning strategies. The information is widely
Fig. 12. Importance rating. spread among specialists of the department so
everybody can contribute to the process by his
observations. It is difficult to express benefits in
exact figures with respect to diminished lead times
control points clarified (reduced development or costs. However it has become clear that top
time; better planning) management get more insight in expected conse-
effective communication between divisions (de- quences (e.g., the need for adaptations or develop-
partments) ments of production technologies) or not yet solved
design intent is carried through to manufacturing problems. Compared to the past they have become
(Quality is built in "upstream"). the "real" decisionmakers. Before sometimes solu-
However it is difficult to obtain specific case tions had to be selected on ad-hoc basis by lower
material to witness the improvements. Generally management levels with restricted insight into pos-
companies are very reluctant to broadcast their sible consequences of their decisions.
results because the results of a Q F D process are Nowadays it is common policy to incorporate
highly confidential and of strategic value for the Q F D into all development processes of this par-
company. ticular department. Most projects require product
One of the first applications of Q F D in The expertise from many different sections of the de-
Netherlands was within the Philips Corporation. partment. Therefore teams are cross functional
Philips concentrated attention on Q F D since 1986. groups of individuals representing appropriate dis-
The first successful application within the frame- ciplines like product planning, marketing, engineer-
work of the Quality Improvement Program was ing and manufacturing. Sometimes, in the case of
achieved in the Chungli monitors factory in Taiwan less complicated projects, the facilitator himself fills
[5, 7]. Afterward (early 1989) it was introduced at in the QFD-charts by consulting the specialists.
the Eindhoven research and development depart- The most important contribution of the facilitator
ment of high-end TV-tubes. This product is an is concentrating on the preparation of scenario's to
extremely complex mixture of many interdepen- communicate the Q F D outcomes to people who
dent technologies which resulted in a present are less familiar with the methodology.
House of Quality of approximately 150 times 120 An approach within ~Van Doorne's Trans-
positions. Filling in all the technologies (How's) missie" (VDT) was quite out of the ordinary. Here
and technical know-how in the chart provided the driving forces to apply Q F D by some interested
a growing insight into the interactions between the groups were very complex problems of process con-
weighed What's on one hand and the How's on the trol. Because the product (a metal push belt fl)r the
other. This knowledge facilitated decision-making automatic transmission} already existed they
and decreased dependence on "good feeling" for started to analyse the importance of and relations
10. 584 C.P.M. Govers/lnt. J. Production Economics 46-47 (1996) 575-585
between process requirements and design variables 4. Implementation
(phase three). The results of this project persuaded
quite a number of designers into the application of Q F D is not a panacea for solving design prob-
Q F D in other areas of the development process lems nor for developing "perfect" products. The
and they will start to discuss the wants of the aim is improving the planning and control of the
automotive industries with respect to the develop- development process. This implies that the other
ment of new generations of continuous variable processes of production are more or less under
transmissions. control. A company that still struggles with the
The concepts of Q F D are not restricted to the quality performance at the expected and the speci-
development of products or services. Sullivan [8-] fied level, has to stress basic quality techniques first
describes an approach for Policy Management in and to change the culture towards more Total
conceptual terms but he does not offer practical Quality Management. With respect to this the
reference. Also in that field clear communication European Quality Award assessment model can be
and customer orientation are fundamental for suc- seen as a reference to derive criteria for production
cess. In an MSc-graduation project his ideas were and development functions but also for the mana-
worked out for the development and management gerial functions. The Dutch Quality Award is
of an annual policy declaration for a production copied from the European Award but for the inter-
plant of an international IC-manufacturer [9]. pretation of the established quality level, the path
As far as I had contacts with Dutch Q F D facili- towards T Q M is split up into five phases. Each level
tators most of them reported technological im- roughly corresponds with a score for the award:
provements. Initial projects of Q F D do not yield all Phase I Activity orientation: Problems are sol-
expected benefits in one go. Early applications re- ved, but the process is not receiving
quire more time and additional effort but results as attention. (0-200 pts)
knowledge transfer, better products and better Phase II Process orientation: Based upon pro-
understanding of the customer expectations are cess control, problems are solved in
directly exploitable. Once a team has gained experi- a systematic way. (200-400 pts)
ence lead time and cost reduction as well as further Phase III System orientation: All functions of
reduction of product and production problems will the organisation are controlled.
be accumulated by fostering a better understanding Customer orientation is achieved in-
of customers' requirements and what is needed to cluding internal customers. Attention is
meet these requirements. paid to prevention of problems.
Most problems they had to untangle were related (400-600 pts)
to organizational circumstances like project defini- Phase IV Chain orientation with suppliers and
tion and project management as well as team selec- customers: Optimal use of knowledge
tion and team building. and capabilities for customer satisfac-
A critical factor concerning project definition tion. Cooperation is sought in order to
is the "Voice of the Customer" (see Kano model) minimise costs. (60~800 pts)
and what are the Critical Quality Characteristics Phase V Total Quality Management: Philos-
[10]. ophy and strategy are based upon a
With respect to project management and team sense of responsibility within the so-
selection, it can be mentioned that focus on the ciety. (800-1000 pts)
expertise is required but take care that the ranks of The majority of "good" companies is at the level of
the team members should be about equal in order approximately 400-500pts which corresponds
to avoid decisions being manipulated by ranking. with the obtaining of an ISO certificate. Around
Keep in mind that support from (top) management 700 pts we find serious candidates for winning the
will be needed. It would be the best to have recep- award.
tive, open-minded members on the team who are To implement Q F D successfully a company
willing to challenge established practice. should have reached roughly the upper level of
11. C.P.M. Govers/Int. J. Production Economics 46 47 (1996) 575 585 585
phase It. This also can be an explanation that at When undertaking a project it is critical to take
present mostly the bigger firms benefit by the Q F D the time to plan and organize your efforts. An ap-
method. Not only because they have the resources propriate project needs to be selected with respect
but probably more because they have already es- to its scope and objectives. Key factors for initial
tablished a system (chain) orientation. success are management support and the constitu-
When an organisation is ready for QFD, there is tion of the team. The team members need to be
the need for a good facilitator who knows the given the time to establish its rules of operation and
method very well and has also the social skills to training requirements.
build and to manage a team. Usually the first pro-
ject will be more time consuming and appeal to
open minded discussions. So look for people who References
has a positive attitude towards new approaches
[1] Feigenbaum, A.V., 1961, 1983, 1986. Total Quality Con-
(early innovators).
trol. McGraw-Hill, New York.
[2] Akao, Y., 1990. Quality Function Deployment, Integrating
Customer Requirements into Product Design. Productiv-
5. Conclusion ity Press. Cambridge, MA.
[31 Sullivan, E.P., 1986. Quality function deployment, Quality
It can be said that Q F D is a synthesis of numer- Progress. June, 39 50.
[4] Hauser, JR. and Clausing, D., 1988. The House of Quality,
ous methodologies originating from the USA (e.g., Havard Bus. Rev. 63 73.
Value Engineering and market research) but they [5] Nijdam, L.M., 1990. Quality function deployment, in:
are perfected and integrated by the Japanese. Suc- Quality Matters, a publication of the CQB of Philips
cess stories from Japan are fairly known but do not International, no. 21.
forget that they have already a longstanding experi- [61 King, R., 1989. Better Designs in Half the Time. Implemen-
ting Quality Function in America. 3rd ed. Goal QPC,
ence. Problems encountered by facilitators in the Methuen. MA.
"West" concentrate on the realization of Cross [7] Rijsingen. P.H. and de Vries, R.F., 1991. Quality ['unction
Functional Management. deployment in picture tube production, in: Quality
There is a wide variety of improvement tools that Matters, a publication of the CBQ of Philips International,
will enable companies to achieve high quality. no. 28.
[8] Sullivan. L.P.. 1988. Policy management through quality
Tools alone however cannot provide results on function deployment, Quality Progress. June, 18 20.
themselves. They must be developed to reflect the [9] Philips, M., Sander, P, and Govers, C., 1994. Policy formu-
companies' culture and management vision. To im- lation b~ use of QFD techniques: a case study. Int. J.
plement Q F D successfully a company has to be Quality Reliability Management, 11(5): 46.58.
system oriented. Q F D provides activities that bring [10] Govers, C.P.M., 1993. Quality of services "applicable to
production'?", Int. J. Prod. Econom., 30-31:385 397.
together all required disciplines to work and plan [11] Goal/QPC, 1990. Quality Function Deployment: A pro-
the development efforts in a highly disciplined, cess for Iranslating customers' needs into a better product
communicative and effective manner. Q F D as such and profit, Research Report no. 89-10-02. Goal/QPC
is not a high technology rather it is a technology Methuen, MA.
[12] Americala Supplier Institute, 1987. Quality Function
developed by users based on common sense and
Deployment, Instruction Manual ASI, Dearborn, MI.
effective information transfer. Many Japanese and
American companies have experienced Q F D to be
worth the effort.