Sunteți pe pagina 1din 154

AUTOMOTIVE ENGINEERING

Mario Vianello
vianello.clm@tin.it

PRODUCT
QUALITY DESIGN

01OFHLO

These slides can not be reproduced or distributed without permission in writing from the author. 1
F O R E W O R D

 Everybody knows, today, that a product must be concei-


ved, designed and realized on the basis of the effective
customer requirements (declared or not): this is essen-
tial to obtain a high quality product.
 Nevertheless it is not so easy to rationally predict how
much a design choice will satisfy customer’s expectations:
we have to continuously refine our prediction methods.
 But it is not enough that a product has the desired featu-
res on delivery: its performances have to be maintained
(without sensible decays) during the whole useful life.
 So reliability becomes a basic requirement for the Qua-
lity of a product.

These slides can not be reproduced or distributed without permission in writing from the author. 2
F O R E W O R D

How and when the reliability of a product has to be measured ?

Intuitively one could think to do it by experimental tests as soon as possible


(e.g. on the first prototypes), but there are 3 kinds of problems:
1. Prototypes are available only at the end of the Product Development Process
so that, if corrective actions are needed, they are very expensive (or even im-
possible) because their implementation is close to commercial launch.
2. Prototypes cannot be completely similar to the products that will be commer-
cialized, at least because of corrective actions implemented during the expe-
rimental tests. How can we estimate the reliability of our final products, on the
basis of products not yet completely definitive? How can we take into account
the differences?
3. A strict reliability “statistical demonstration” may require samples excessively
larger than feasibility. So we must learn to rationally balance theoretical pre-
scriptions against company’s needs. For this, we must accept some “reaso-
nable” risks, even if they are not statistically quantifiable. Of course the know-
ledge of Statistics is necessary. The main point is not to contradict its basic
principles in our practical applications. Mathematical expressions are not a
problem because they are provided by PC.

These slides can not be reproduced or distributed without permission in writing from the author. 3
F O R E W O R D

 It is also evident the advantage to mainly focus on pro-


active activities, in order to point out criticalities before
design-phase starts or during its development. This in-
volves important economical benefits in comparison
with putting off reliability verifications until the final expe-
rimental tests. As a consequence, it is suitable to know
the main prevention methods, at least in terms of ad-
vantages, limits and application areas of each of them.

 Of course all previous activities must be managed at the


lowest cost. Though not often repeated, minimizing
cost is always one of the most important targets for all
methods mentioned in this Course.

These slides can not be reproduced or distributed without permission in writing from the author. 4
F O R E W O R D

 This Course is focused on RELIABILITY,


 and points out the 2 phases of the Product Deve-
lopment Process in which reliability is essential:
prevention and experimental verification tests.
 The slant of this Course is aimed to industrial ap-
plications,
 with special care to automotive field.
 The practices are drawn from real cases.
 A freshen up of Applied Statistics Fundamen-
tals has been also inserted.

These slides can not be reproduced or distributed without permission in writing from the author. 5
F O R E W O R D

In details:
1. The Course starts from the achievement of customer requirements
(VOC = Voice Of Customer) using Quality Function Deployment
(QFD) method;
2. then general principles of Reliability and Robust Design are in-
troduced;
3. the Applied Statistics Fundamentals are summarized;
4. the most common Reliability prediction/prevention methods
are described;
5. the main methods for Reliability experimental verifications are
examined and discussed and the best way to manage them by a
global and coherent vision is also treated;
6. the last topic is an overview of the Product Development Pro-
cess and a mention about the global approaches to Quality, the
last of whom is Six Sigma.

These slides can not be reproduced or distributed without permission in writing from the author. 6
COURSE CONTENTS

1. Design for Quality considering customer needs and pro-


duct targets
2. Freshen up of Applied Statistics Fundamentals
3. Fundamentals of Reliability and of Robust Design
4. Measure and prevention methodological instruments for
Reliability
5. Criteria and methods to plan reliability experimental verifi-
cations
6. Managerial considerations: mention about Problem solving,
Lessons learned, Experience accumulation, Technical memory,
Global approaches like Six Sigma

These slides can not be reproduced or distributed without permission in writing from the author. 7
SYMBOLS USED TO SPECIFY THE AIMS OF EACH COURSE CONTENT

A) Fundamental (basic principles)


Fundamental topics to be learned very well (concepts, application areas
and uses): most questions (of written and oral tests) are related with
them.

B) Overview (helpful to remember)


General settings of problems/approaches in their industrial environment.
This category of contents includes comparison-Tables aimed to highlight
differences among alternative methods (e.g. typical application areas).
These slides would not require demanding efforts when giving over.

C) Detail/deepening (not used for tests)


Complements and explanations useful for a better understanding of Fun-
damental Topics (A). They have been included to avoid possible misun-
derstandings and doubts, but also to be used on the job in future.

These slides can not be reproduced or distributed without permission in writing from the author. 8
PRACTICES

ACKNOWLEDGMENTS

Sincere thanks to Mr. De Blasi


for the thorough review
and the significant enrichment
of these slides.

These slides can not be reproduced or distributed without permission in writing from the author. 9
PRODUCT QUALITY DESIGN

1
DESIGN FOR QUALITY
CONSIDERING CUSTOMER NEEDS
AND PRODUCT TARGETS

These slides can not be reproduced or distributed without permission in writing from the author. 10
CONTENTS OF THIS CHAPTER

 Quality concept evolution


- Quality history outlines
- Quality/Reliability cost items and hint at their trends
- “Relative” and “Absolute” Quality
 Voice Of Customer (VOC) & Quality Function Deployment (QFD)
- “House of Quality” and their “building blocks”
- Kano model
- Customers “ranking”
- “Mathematical” and “logical” approaches
- Correlation between physical measurement and customer satisfaction ex-
pressed in a SAE scale (example of take-off indicator)
- Practical impacts of House of Quality
- McCabe method
- Strong points and strategies
- Customer simulation
 Quality measurements
- “Expected Quality”, “Planned Quality”, “Attained Quality”, “Perceived Quality”
- Mathematical expressions to evaluate “Attained Quality”
- “Customer Expected Functions” measured in relative and absolute terms
- Car “classes”

These slides can not be reproduced or distributed without permission in writing from the author. 11
CONTENTS OF THIS CHAPTER

 Quality, Reliability and need of a Statistical Approach


- Quantities related to Quality
- Reliability definition
- Other important quantities related to Reliability (Maintainability, Availability
and Durability)
- Dependability definition
- Need of a statistical approach

These slides can not be reproduced or distributed without permission in writing from the author. 12
1. Design for Quality

1.1
Quality concept evolution

These slides can not be reproduced or distributed without permission in writing from the author. 13
1.1. Quality concept evolution

1.1.1
Quality history outlines

These slides can not be reproduced or distributed without permission in writing from the author. 14
1.1.1. Quality history outlines

Synoptic Table

QUALITY
YEAR WHEN / WHERE STRATEGY
seen as …

Quantity overcomes quality


TESTING
1950-1960 luxury Compliance with specifications
(Workers)

Process control
QUALITY CONTROL
1960-1970 cost by product testing results
(Experts) and defectives selection

Preventive improvements
sales QUALITY ASSURANCE
1970-1980 on the process (investments)
rising ("Integrated" company) to avoid checking the product

Company must be rearranged


profits TOTAL QUALITY MANAGEMENT

Quality first
1980-1990 to meet customer satisfaction
increase (Customers, Market, Suppliers) ( effectiveness + efficiency )

Constant control of both


1990 optimized social INTERCOMPANY QUALITY
company processes
to date impacts ("Certified" company ) and extra-company framework

These slides can not be reproduced or distributed without permission in writing from the author. 15
1.1.1. Quality history outlines

EFQM model
( EFQM = European Foundation for Quality Management )
ENABLERS RESULTS

PERFORMANCE
PROCESSES,
LEADERSHIP

& SERVICES
PRODUCTS
People People Results

RESULTS
KEY
Customer
Policy & Strategy
Results

Partnerships
Society Results
& Resources

INNOVATION AND LEARNING

The philosophy is to achieve economical results


and also to contribute to social development (ethical aim)
These slides can not be reproduced or distributed without permission in writing from the author. 16
1.1.1. Quality history outlines

“Certification” assures a quality level


consistent with the company’s targets:
which can also be not very high.

In details, certification assures that:


• the company has official targets and they are known by
all interested people;
• a set of norms aimed to their attainment is available;
• every Employee knows the standards regarding his job.

These slides can not be reproduced or distributed without permission in writing from the author. 17
1.1.1. Quality history outlines

Quality concept evolution:

from: conformity control to design specifications

to: meeting customer requirements

QUALITY = CUSTOMER SATISFACTION


That implies installing specific indicators
to acquire the level of
quality appreciated by the customers.

These slides can not be reproduced or distributed without permission in writing from the author. 18
1.1.1. Quality history outlines

T A R G E T S

1st target CUSTOMER SATISFACTION

 to do things right the first time


2nd target EFFICIENCY
 continuous improvement

These slides can not be reproduced or distributed without permission in writing from the author. 19
1.1.1. Quality history outlines
Program planning and defining
Do not begin to make a new product,
without a program clearly specifying:
 Trends by Competitors

idea  Analysis of customer expectations

 Information sources from the field


 check the idea
 what to do  to whom can we sell?
 how many products can we sell?
 at what price can we sell?
Business ?
 how to advertise?

T A R G E T S :
 why us?  performances
 why to do it  reliability
 why it will be profitable?  cost
 date of Comm. L.

 how to do it  choices between alternatives

These slides can not be reproduced or distributed without permission in writing from the author. 20
1.1. Quality concept evolution

1.1.2
Quality/Reliability cost items
and hint at their trends

These slides can not be reproduced or distributed without permission in writing from the author. 21
1.1.2. Quality/Reliability cost items and hint at their trends

Quality Costs - 3 CLASSICAL CATEGORIES

 Failure costs: they may be detected Strictly speaking,


costs of poor quality,
both in Production and in Customer Ser- because related to an
vice (fine tuning before delivery to custo- inadequate quality
mers, warranty repairs, etc.). level.

 Appraisal costs regard the Production


Controls, whose inadequacy may increase Strictly speaking,
the number of defectives (which have to be costs for quality,
because they are
re-worked or scrapped). investments and
 Prevention costs include all design and operating costs
process activities (investments and opera- aimed to assure a
ting costs) aimed to assure that the most sufficient
important defects will never happen. quality level.

These slides can not be reproduced or distributed without permission in writing from the author. 22
1.1.2. Quality/Reliability cost items and hint at their trends
Quality Costs – CATEGORIES AND SHAPE OF TYPICAL CURVES

FAILURE
APPRAISAL

COSTS / QUALITY

BEST TRADE OFF


PREVENTION
TOTAL
C O S T S

0 QUALITY LEVEL 100


These slides can not be reproduced or distributed without permission in writing from the author. 23
1.1.2. Quality/Reliability cost items and hint at their trends

Quality Costs - TRENDS

PREVENTION
APPRAISAL
FAILURE
C O S T S

T EI M
TIM [ yEe a r s ]
These slides can not be reproduced or distributed without permission in writing from the author. 24
1.1.2. Quality/Reliability cost items and hint at their trends

It is important to perform
balanced choices
in order to
optimize total costs.

These slides can not be reproduced or distributed without permission in writing from the author. 25
1.1.2. Quality/Reliability cost items and hint at their trends

It is also important to evidence


two other aspects:

 The costs of poor quality are generated not only by failures,


but, in general, by every activity which we have to repeat
more than one time (although this could be prevented).

 The failure costs related to the warranty repairs are only the
easiest (but partial) measure of business losses due to poor
quality of a product; greater business losses (difficult to quan-
tify) derive from lacked sales resulting from the loss of ima-
ge due to the failures.

These slides can not be reproduced or distributed without permission in writing from the author. 26
1.1.2. Quality/Reliability cost items and hint at their trends
Perseverance in the commitment to Quality
produces substantial reductions
of the "hidden costs" associated with the "hidden factory"
and depicted as the submerged part of the iceberg.

Avoidable design changes Excessive stock levels

Longer cycle times Low stock turnover

Obsolete materials
Longer logistics flows in the warehouse

More frequent carriage


Delays in supplying

The items shown in the


Figure are examples and
not all those possible!
These slides can not be reproduced or distributed without permission in writing from the author. 27
1.1.2. Quality/Reliability cost items and hint at their trends

In the following, we will not speak any more about costs.


Nevertheless we must never forget them,
because the feasibility of every activity
(in development or planned),
must be verified by them.

These slides can not be reproduced or distributed without permission in writing from the author. 28
1.1. Quality concept evolution

1.1.3
“Relative” and “Absolute”
Quality

These slides can not be reproduced or distributed without permission in writing from the author. 29
1.1.4. Relative” and “Absolute” Quality

QUALITY IS MEASURED BY THE LEVEL AT WHICH


A PRODUCT OR A SERVICE
SATISFIES CUSTOMER’S EXPECTATIONS

This definition contains the following principles :


 Quality is a measurable concept;
 Quality is always expressed by a ratio between a fulfillment and
an expectation;
 Quality measurements refer to the “observer” system and, for
that, they are “relative”;
 Equal observers with the same expectations measure the same
quality (otherwise quality could not be considered “measurable”).

These slides can not be reproduced or distributed without permission in writing from the author. 30
1.1.4. Relative” and “Absolute” Quality

E X A M P L E
It is unquestionable that
a Rolls Royce has an absolute quality
much higher than an economy car.

 But, considering only the absolute quality, we do not take into ac-
count that the expectations of a purchaser of a Rolls Royce are su-
rely very different from those of a purchaser of an economic car.
 Moreover, we neglect to evaluate how and how much have been fulfil-
led, in each of the two cases, the characteristics suitable to satisfy the two
sets of expectations.
 The use of absolute quality in order to compare products oriented
to different mission profiles is inadequate, because (for example)
for the purchaser of an economic car, the performances of his car
can be more satisfactory than those of a Rolls Royce!
These slides can not be reproduced or distributed without permission in writing from the author. 31
1.1.4. Relative” and “Absolute” Quality

E X A M P L E

A graphic arts company produced,


with the same absolute quality (very high),
both a series of painting art books
and a women’s fashion magazine.

It is not strange that the economical results were :


• extremely flattering for the painting art books, which have a de-
manding market ready to pay for a perfect agreement between the
printed reproduction and the original;
• in the red for the women’s fashion magazine, whose purchaser
is not interest to the good agreement with an original that he does
not know and, as a consequence, he is not willing to pay a high
price for that.

These slides can not be reproduced or distributed without permission in writing from the author. 32
1.1.4. Relative” and “Absolute” Quality
E X A M P L E
A factory of dolls based its QUALITY on:
• accuracy of finishes,
• richness of dresses,
• “human” functions: dolls able to talk, walk, drink from a feeding-bottle,
wet her nappy, etc..
 After years of unopposed leadership, the market share began to drama-
tically decrease.
 They realized that new generations of little girls preferred to have with
the doll a relationship of the kind “younger sister/elder sister”, rather
than of the traditional kind “mother/daughter”: the plays of little girls are
now oriented to “when I will be a young lady”.
 As a consequence, the old doll with a baby aspect were supplanted from
a new doll with the aspect of a girl sixteen/twenty years old and the tra-
ditional accessories (clothes, prams, feeding-bottles, etc.) were replaced
by sport-suits, cars, furnished doll houses, etc..
Clearly, a showy change in the customer’s expected functions
had been happened and had not been immediately realized
by the company.
These slides can not be reproduced or distributed without permission in writing from the author. 33
1.1.4. Relative” and “Absolute” Quality

 The meaning of Quality is always related to the observer who


appreciates it and therefore it is relative to the observer’s referen-
ce system (personal technical knowledge, needs to be satisfied, sustai-
nable costs, etc.).

 Then the Quality from the customer’s point of view


must be always measured by the “subjective” evalua-
tions of the observer, i.e. in a “relative” way.

 However relative quality is not easy to quantify, specially when


a prediction is required, because the last word is that of the custo-
mer (who needs of course at least a prototype to judge). Never-
theless a prediction of customer’s satisfaction is essential in order
to optimize the characteristics of a new product, comparing them
with those planned by the best competitors.

These slides can not be reproduced or distributed without permission in writing from the author. 34
1.1.4. Relative” and “Absolute” Quality

 Instead, absolute quality could be defined (independently of


what the user perceives) as a measurement of characteristics and
performance level that a product or service offers to the user;
 Absolute quality is measured in physical units (e.g.: kW for
engine power, dm3 for trunk capacity, pixel/mm2 for image
definition and so on): thus it is easy to measure and closely
related to cost (the higher the level, the higher the cost);
 It is especially right to “objectively” define technical targets.
 However it may have a poor correlation with customer satisfaction
(see previous examples) and therefore absolute quality is not a
good indicator of the quality from the customer’s point of view.

 Consequently it is worthwhile organizing the activity in order to


estimate relative quality level using physical measurements,
i.e. absolute quality.
These slides can not be reproduced or distributed without permission in writing from the author. 35
1.1.4. Relative” and “Absolute” Quality

Every potential purchaser


of a product or a service
has some expectations on it,
that are expressed
(because they are the reason)
before the purchase.

Generally these expectations


may be expressed
by some functions
that the customers require.
These slides can not be reproduced or distributed without permission in writing from the author. 36
1.1.4. Relative” and “Absolute” Quality

 Identifying these functions,


 evaluating their relative importance from the customer’s point
of view,
 quantifying the satisfaction level attained by the product on
each expected function.
are the means by which we are able to measure the Quality.

The classical Marketing instrument


aimed to these goals is the
QUALITY FUNCTION DEPLOYMENT.
Before continuing to speak about
Quality measurement,
we will give a short mention of it.
These slides can not be reproduced or distributed without permission in writing from the author. 37
1. Design for Quality

1.2
Voice Of Customer (V.O.C.)
&
Quality Function Deployment
(Q.F.D.)

These slides can not be reproduced or distributed without permission in writing from the author. 38
1.2. VOC & QFD

In the planning for a NEW DESIGN we need:


1. to identify, among all the characteristics perceived by the
customer, the really important ones (Pareto Principle of Vital
Few);
2. to convert them into design technical targets;
3. to ensure the matching between technical targets and cost
targets (also defined by the Market).

These slides can not be reproduced or distributed without permission in writing from the author. 39
1.2. VOC & QFD

Besides the clear identification of targets and priority areas,


it is also key:
1. to ensure a proper “conversion” of Customer needs into
design technical characteristics;
2. not to loose the references to the priority of characteristics from
the customer’s point of view, during all phases of the Product
Development Process;

3. to coherently involve all Company’s Departments, carrying


through them the Voice Of Customer (VOC).

These slides can not be reproduced or distributed without permission in writing from the author. 40
1.2. VOC & QFD

The method to get these results is the


Quality Function Deployment (Q.F.D.),
which carries the “Voice Of Customer” (VOC)
through all the product development phases
(and Company Departments).

These slides can not be reproduced or distributed without permission in writing from the author. 41
1.2. VOC & QFD

What is Quality Function Deployment (Q.F.D.) ?

It is a methodological tool aiming at…


 contributing to plan and carry out the Product Process
Development in a structured way;

 making able the “Development Team” to clearly specify the


requirements and the needs of the customers;

 carrying the Voice Of Customer along the development of


both the design and the process;

 evaluating the capabilities of both the design and the


process to satisfy the identified requirements.

These slides can not be reproduced or distributed without permission in writing from the author. 42
1.2. VOC & QFD

The Quality Function Deployment (QFD) method was introduced in


Japan in late ’60s, by Professors Yoji AKAO and Shigeru MIZUNO.

Up to then, other methods, based on statistical process control and other Quality
improvement techniques, integrated with the studies by Dr. Juran and Prof. Kaoru
Ishikawa, have expanded their application more and more, since the end of the
second World War.

All such methods, however, were mainly aimed to solve / prevent problems during
or after production.

On the other hand, according to the definition of the ASI (American Supply
Institute), the QFD represents:

“… a system for translating customer requirements into appropriate company


requirements at every stage, from research through production design and
development, to manufacture, distribution, installation and marketing, sales and
services”.

These slides can not be reproduced or distributed without permission in writing from the author. 43
1.2. VOC & QFD

Through Quality Function Deployment (QFD) we aim to increase


competitiveness and market shares.

The market share of a company may be increased by convincing


the customers to prefer its product to those of its competitors. In
order to do this, the customers must perceive our product as the
one closest to their requirements, in terms of:
 performances,
 reliability,
 price.

Of course, to achieve such a result, it is not enough to apply QFD, but all search,
prevention and experimentation instruments have to be considered and used in
an organic and systematic way.

However QFD results are the fundamental data


so that the following activities (of the Product Development Process)
are effectively addressed to the customer’s satisfaction.

These slides can not be reproduced or distributed without permission in writing from the author. 44
1.2. VOC & QFD

The determination in overcoming competitors that originates QFD


can be well illustrated by the following Japanese joke.

Two friends go on a mountain hike. They are well equipped


(pullovers, windcheaters, mountaineering boots, and so
on). During a break, they hear a bear roaring.
One of them quickly takes off his mountaineering boots and
puts on a pair of sneakers. His friend exclaims “Also using
your sneakers you won’t be fastener than the bear!” .
The blunt replay is “I don’t need to be fastener than the
bear, I just need to be fastener than you!”.

These slides can not be reproduced or distributed without permission in writing from the author. 45
The Q.F.D. basic logic: the House of Quality
( ** ) (*)
Commonly used values : Commonly used values:
strong relationship =9 synergy: strong = +9
moderate relationship =3 (positive) weak = +3
HOWs vs. HOWs
weak relationship =1 contrast: strong = -9
relationships(*) (negative) weak = -3
WHATs vs. HOWs
matrix of impacts HOWs
(product's Technical Characteristics)
WHYs

relationships (**)

in comparison to the best competitors


x x x

Customer PERCEPTIONS
x

of each customer need


Relative IMPORTANCE
(customer needs)

relationships(**)
WHATs

x
x

x
x x

Importance
HOW MUTCHes

HOW MUTCHes
of each Technical Characteristic
importance)
to assess

HOWs vs.
(criteria

matrix
Technical measurements
compared to the best competitors

Targets
of each Technical Characteristic

These slides can not be reproduced or distributed without permission in writing from the author. 46
1.2. VOC & QFD

The starting point is to identify Customer Needs.


Such preliminary analysis is typically done through the application
of one or more of the following tools:
• Brainstorming,
• Cause and Effect Diagram,
• Kano Model
• Interviews, Focus Group, etc.

CHARACTERISTICS PERCEIVED QUALITY

PERFORMANCES AESTHETICS

CUSTOMER’S
SATISFACTION

RELIABILITY SERVICEABILITY

DURABILITY CONFORMITY

These slides can not be reproduced or distributed without permission in writing from the author. 47
1.2. VOC & QFD
Different categories of customer needs
CUSTOMER NEEDS (WHATs): KANO diagram

excitement
DELIGHTER QUALITY
(not expressed needs = WOW factors )
CUSTOMER SATISFACTION

needs not fulfilled


needs well fulfilled
MORE IS BETTER (one-dimensional) QUALITY
(usually expressed needs, whose satisfactions are
MUST-BE QUALITY
proportional to the performance )
claims (implicit needs usually not expressed )

OFFERED QUALITY
These slides can not be reproduced or distributed without permission in writing from the author. 49
1.2. VOC & QFD
Example of Customer Needs

and, of course, RELIABILITY

These slides can not be reproduced or distributed without permission in writing from the author. 50
1.2. VOC & QFD

CUSTOMER'S REQUIREMENTS (WHATs)

BASIC " Products with characteristics useful to fulfill


required functions"
Negative Quality

NEEDS

NEEDS TAKEN Needs required but non expressed :


FOR GRANTED unwanted side effects, safety aspects not conside-
(" must be ") red by law, etc..

LAWS AND Compliance with laws, regulations and other con-


REGULATIONS straints of each Market.
Positive
Quality

"HIDDEN" Features not demanded by the customer, but such


CUSTOMER'S that, once discovered, are for him source of remar-
EXPECTATIONS kable satisfaction (WOW factors ).

These slides can not be reproduced or distributed without permission in writing from the author. 51
1.2. VOC & QFD

MAIN STEPS TO ORGANIZE CUSTOMERS INTERVIEWS


AIMED TO QFD

 Brainstorming (to provide a written questionnaire or


even as an alternative to interviewing customers);
 Interviewing the customers in order to bring out the
actual reasons which made them decide to purchase;
 Grouping of similar sentences (e.g. by Tree Diagram
or CEDAC;
 Definition of main topics;
 Integration of similar items within each topic and
elimination of duplications (if any);
 Elimination of little importance items.
These slides can not be reproduced or distributed without permission in writing from the author. 52
1.2. VOC & QFD
TRANSLATING CUSTOMER WORDS INTO MEASURABLE
CHARACTERISTICS
TECHNICAL CHARACTERISTICS

Car door

armrest grip to
characteristics

characteristics

close the door


Position of
of opening

of opening
Dynamic
Customer's requirements Customer's requirements

the door

position

position
Static
close
Load
(expressed by its own words) converted in

to
generally not measurable measurable characteristics

(fictitious Load to close the door


Easy to close

but realistic outside

Easy to close
Dynamic characteristics of opening position
Static characteristics of opening position
Load to close the door
Dynamic characteristics of opening position

example) inside Static characteristics of opening position


Position of armrest grip to close the door
Load to activate internal door handle
Load to activate external door handle
Easy to open
(inside) Dynamic characteristics of opening position
EASY Shape of internal door handle
TO Position of internal door handle
OPEN / CLOSE It does not Static characteristics of opening position
rebound Dynamic characteristics of opening position
It must stay ope-ned at a Static characteristics of opening position
wanted position
Dynamic characteristics of opening position
Winder easy
Winder position
to get
Winder position
Winder easy
to grasp Knob diameter
Knob length
Load to activate window regulator
F E A T U R E S

WINDOW REGULATOR Winder position


EASY Easy to handle Operating direction
TO OPERATE Knob diameter
Knob length
Cleaning of the glass
while moving
Load to activate window regulator

Speed of Load to activate window regulator


operations Operating direction
Easy handling of Load to close security switch
G O O D

security switch Position of security switch


Load to introduce the key
Easy introduction
EASY BLOCK of the key Load to open by key
OF Load to open by key with low temperature
THE LOCK Load to introduce the key
Easy handling of
the key Load to open by key
Load to open by key with low temperature
It has not have to
Load to open by key with low temperature
freeze

NO SEEPAGES No water seepages Stiffness of weather strips

No dripping
Stiffness of weather strips
while opening

These slides can not be reproduced or distributed without permission in writing from the author. 53
1.2. VOC & QFD
TRANSLATING CUSTOMER WORDS INTO MEASURABLE
CHARACTERISTICS

TECHNICAL CHARACTERISTICS

armrest grip to
characteristics

characteristics

close the door


Position of
of opening

of opening
Dynamic
Customer's requirements Customer's requirements

the door

position

position
Static
close
Load
(expressed by its own words) converted in

to
generally not measurable measurable characteristics

Load to close the door


Easy to close
outside Dynamic characteristics of opening position
Static characteristics of opening position
EASY TO OPEN / CLOSE

Load to close the door


Easy to close Dynamic characteristics of opening position
inside Static characteristics of opening position
Position of armrest grip to close the door
Load to activate internal door handle
Load to activate external door handle
Easy to open
(inside) Dynamic characteristics of opening position
Shape of internal door handle
Position of internal door handle
It does not Static characteristics of opening position
rebound Dynamic characteristics of opening position
t must stay ope- Static characteristics of opening position
ned at a wanted
position Dynamic characteristics of opening position

These slides can not be reproduced or distributed without permission in writing from the author. 54
1.2. VOC & QFD

HOUSE OF QUALITY
BUILDING BLOCKS
(ROOMS)

These slides can not be reproduced or distributed without permission in writing from the author. 55
CAR DOOR

WHYs

IMPORTANCE FROM THE CUSTOMER POINT OF VIEW

CRITICAL REQUIREMENTS
COMPARISONS WITH COMPETITORS
REPAIRS BY CUSTOMER SERVICE

O P P O R T U N I T I E S
AND THEIR IMPORTANCE
CUSTOMER’S

Competitor 1
Competitor 2
REQUIREMENTS

We
CUSTOMER’S
REQUIREMENTS
(WHATs)
EASY TO CLOSE/OPEN
Easy to close outside
AND THEIR
“RELATIVE”
Easy to close inside
Easy to open (inside)
WHATs

It does not rebound


It must stay opened at a wanted position
WINDOW REGULATOR EASY TO OPERATE
Winder easy to get
IMPORTANCES
Winder easy to grasp
Easy to handle
Cleaning of the glass while moving
(WHYs)
Speed of operations
EASY BLOCK OF THE LOCK
Easy handling of security switch
Easy introduction of the key
Easy handling of the key
It does not have to freeze
NO SEEPAGES
No water seepages
No dripping while opening

These slides can not be reproduced or distributed without permission in writing from the author. 56
1.2. VOC & QFD

RATING CUSTOMER’S REQUIREMENTS


Main criteria:
a) Importance of each need from the Customer’s
point of view (e.g. through customer surveys)
b) Evaluation of bad quality aspects (from historical
data /defects / complaints related to the current
products)
c) Brand perception (which needs are usually
linked to the brand image we want to diffuse)
d) Objective comparison with competitors
(products)

These slides can not be reproduced or distributed without permission in writing from the author. 57
1.2. VOC & QFD

Customer Rating of the Competitors

Understanding how customers rate the


competitors products can be a
tremendous competitive advantage.

So, in this step of the QFD process, it is


also a good idea to ask customers how
your product or service rates in
relation to the competition.

These slides can not be reproduced or distributed without permission in writing from the author. 58
HOWs vs. WHATs
CAR DOOR

Voice of the
MATRIX PRODUCT’S TECHNICAL Company
CHARACTERISTICS
Symbols indicate the weight
(0, 1, 3, 9),
BY WHICH
THE COMPANY
(DESIGNERS)
PLANS TO MEET
i.e. the strength, THE CUSTOMER’S

Position of lever to set the external rear-view mirror

No water seepages (design, verifications and tests)


REQUIREMENTS

Dynamic characteristics of opening position


of each relationship

Load to open by key with low temperature


Static characteristics of opening position

Position of armrest grip to close the door


Load to activate external door handle
Load to activate internal door handle
among Customer’s

Load to activate window regulator


F A C T O R S

Position of internal door handle


Shape of internal door handle
Load to close security switch

WINDOW REGULATOR

WINDOW REGULATOR

BLOCK OF THE LOCK

Position of security switch


Load to introduce the key
Load to close the door
Requirements (WHATs)

Load to open by key

Operating direction
Winder position

Knob diameter
Knob length
and Technical Characteristics

OTHERS
H U M A N
L O A D S
DOOR

LOCK
(HOWs).

EASY TO CLOSE/OPEN
Easy to close outside
Easy to close inside
Easy to open (inside)
It does not rebound
It must stay opened at a wanted position
WINDOW REGULATOR EASY TO OPERATE

Voice of the Winder easy to get


Winder easy to grasp
Easy to handle

CUSTOMER Cleaning of the glass while moving


Speed of operations
EASY BLOCK OF THE LOCK
Easy handling of security switch
Easy introduction of the key
Easy handling of the key
It does not have to freeze
NO SEEPAGES
No water seepages
No dripping while opening

These slides can not be reproduced or distributed without permission in writing from the author. 59
1.2. VOC & QFD

The main purpose of this activity is to identify


the best opportunities to overcome competitors.

So a useful rule is to quantify the strength of each


relationship among Technical Characteristics (HOWs)
and Customer’s Requirements (WHATs),
not in term of “absolute” measurements
(of the phenomenon’s physical parameters),
but by evaluating
the variation from the current State of Art
(of the best manufacturers).

These slides can not be reproduced or distributed without permission in writing from the author. 60
1.2. VOC & QFD

Example

How strong is the relationship between:

• the kind of tire (HOW) and


• he speed required by the customer (WHAT) ?

Thinking that a normal car can not move without wheels, you
might be tempted to assess the strength of this relationship
equal to 9.

But this consideration is trivial. It is more appropriate to note


that, only beyond a certain speed, we have to change kind of
tire and therefore it is reasonable to quantify the strength of
the tire/speed relationship equal to 1 (or at most equal to 3).
These slides can not be reproduced or distributed without permission in writing from the author. 61
1.2. VOC & QFD
P R A C T I C A L R E C O M M E N D A T I O N S
Customer’s Requirements (WHATs) Technical Characteristics (HOWs)

 It is essential to ensure that all re-  In the selection of subsystems to be


quirements have been highlighted in the columns of Techni-
considered, as if some are cal Characteristics (HOWs), should
missing (perhaps because be kept in mind that the purpose of
obvious) might compromise the this activity is to identify priority
validity of the results. areas for action (from a customer
 On the other hand, it is important perspective).
to make sure not to repeat the sa-  With the help of experience, this can
me requirement twice (or even lead to some simplifications and/or
more), maybe with different to the introduction of some "virtual“
words. subsystem (e.g. a particular fit be-
tween mechanical parts).
 On the other hand, if some columns
show the same symbols in all their
cells, they are expressing the same
things and we can merge them.
These slides can not be reproduced or distributed without permission in writing from the author. 62
1.2. VOC & QFD

Customers Ranking

If the product is intended for different customers,


each with special needs/priorities,
it is often useful to assign
a “weight” to each kind of customer
(e.g. in terms of purchased quantities, image, etc.),
by which we can then revise the overall importance
of each requirement:
in this way, we can rationally decide
whether a single product can reasonably satisfy all customers
or whether it is preferable to differentiate it.

These slides can not be reproduced or distributed without permission in writing from the author. 63
1.2. VOC & QFD

Final, intermediate and internal customers


(economical and safety aspects)

Separately from those of the “final” customer, the QFD must


consider (usually at the bottom), also:
• “intermediate” customer requirements (e.g. speed of
assembly, special packaging, etc.): their meeting will give
us the trust of the customer and his preference;
• aspects (usually internal) generally related to safety (for
example the movement of bulky or dangerous items) and
also economical aspects: these are requirements so im-
portant to be “obvious”; however they must be highligh-
ted, otherwise some of the made choices may not be
justified.

These slides can not be reproduced or distributed without permission in writing from the author. 64
CAR DOOR

L O A D S
DOOR
Load to close the door
Load to activate internal door handle
Load to activate external door handle
Static characteristics of opening position
Dynamic characteristics of opening position
LOCK
Load to close security switch
Load to introduce the key
Load to open by key
Load to open by key with low temperature
WINDOW REGULATOR
Load to activate window regulator
H U M A N F A C T O R S
WINDOW REGULATOR
Winder position
Operating direction
Knob diameter
knob length
BLOCK OF THE LOCK
Shape of internal door handle
Position of internal door handle
Position of security switch
OTHERS

These slides can not be reproduced or distributed without permission in writing from the author.
Position of lever to set the external rear-view mirror
Position of armrest grip to close the door
No water seepages

them.
characteristics.
among technical
INTERACTIONS

(Positive or negative)
AMONG TECHNICAL

if there are interactions


POSITIVE AND NEGATIVE

defining priorities among


CHARACTERISTICS (HOWs)

In the “Roof” of the House

This could be useful when


of Quality we can visualize

65
AND
“LOGICAL”
APPROACHES
“MATHEMATICAL”

These slides can not be reproduced or distributed without permission in writing from the author.
NO SEEPAGES
Easy to handle

%
Winder easy to get

No water seepages
It does not rebound
Easy to close inside

Speed of operations
Easy to open (inside)
Easy to close outside

Winder easy to grasp


EASY TO CLOSE/OPEN
CAR DOOR

Easy handling of the key

Absolute
It does not have to freeze

No dripping while opening


Easy introduction of the key
EASY BLOCK OF THE LOCK
Easy handling of security switch

CRITICALITIES (*)
Cleaning of the glass while moving

Weight based priority

& OPPORTUNITIES (+)


CHARACTERISTICS
It must stay opened at a wanted position

WEIGHTS OF TECHNICAL
WINDOW REGULATOR EASY TO OPERATE

L O A D S
DOOR
Load to close the door
Load to activate internal door handle
Load to activate external door handle
Static characteristics of opening position
Dynamic characteristics of opening position
LOCK
Load to close security switch
Load to introduce the key
Load to open by key
Load to open by key with low temperature
WINDOW REGULATOR
Load to activate window regulator
H U M A N F A C T O R S
WINDOW REGULATOR
Winder position
Operating direction
Knob diameter
Knob length
BLOCK OF THE LOCK
Shape of internal door handle
Position of internal door handle
Position of security switch
OTHERS
Position of lever to set the external rear-view mirror
Position of armrest grip to close the door
No water seepages (design, verifications and tests)

IMPORTANCE FROM THE CUSTOMER POINT OF VIEW


REPAIRS BY CUSTOMER SERVICE
COMPARISONS WITH COMPETITORS
We
Competitor 1
Competitor 2

CRITICAL REQUIREMENTS
O P P O R T U N I T I E S
CRITICAL

(basic logic)
IDENTIFICATION
CHARACTERISTICS

66
1.2. VOC & QFD

QFD: mathematical and logical approaches

To quantify the importance of Technical Characteristics from the


customer’s point of view, three methods can be used:

 two different mathematical algorithms to "weight“ the


importance;
 a further approach simply based on “logic”.

These slides can not be reproduced or distributed without permission in writing from the author. 67
1.2. VOC & QFD
QFD: mathematical approaches
“Absolute approach”.
It is the easiest, the most intuitive and the most used.
For each Technical Characteristic, i.e. for each HOW column, all the
products of the importance value (WHYs) by the weight (or strength) of
the relationship HOW vs. WHAT (0, 1, 3, 9) are calculated and then added:
the greater the sum, the higher the importance of the corresponding
Technical Characteristic (HOWs column).
TECHNICAL CHARACTERISTICS [HOWs]

from the custo-mer's

Relationship strength
IMPORTANCE

ROW TOTAL
Characteristic, x1

Characteristic, x3

Characteristic, x4

Characteristic, x5

IMPORTANCE
Characteristic x2

ROW TOTAL
point of view
CUSTOMER'S
"Relative"

Technical

Technical

Technical

Technical

Technical
REQUIREMENTS
[ WHATs ]
4 x 2 = 12 +
Requirement y1 3 12 1
4 9 3 9 25
4 0,160
x 3 = 12 +
Requirement y2 4 3 12 9 12 0,333
24 =
WEIGHTS
OF TECHNICAL CHARACTERISTICS
[HOWs]
Absolute WEIGHTS (classical)
36 24 40 12 36
(1st mathematical algorithm)
Absolute PRIORITIES (classical)
(1st mathematical algorithm)
2 4 1 5 2
Standardized WEIGHTS
1,44 1,48 3,16 0,48 1,44
nd
(2 mathematical
These slides can not be reproduced or distributed without permission
algorithm)
in writing from the author. 68
1.2. VOC & QFD
QFD: mathematical approaches
“Relative Approach”.
It is an evolution of the previous algorithm.

If the number of Technical Characteristics (HOWs) related with a specific


Customer’s Requirement (WHAT) are much more or much less than the
average, the previous algorithm could lead to an incorrect "weighing" of
the importance of these Technical Characteristics.

This second algorithm makes a recalibration of the previous one. The


weight of the relationship HOW vs. WHAT is not used directly (as in the
above algorithm), but it is divided by the sum of all the weights of the
row.
So the "weights" are used in a “relative” (and not “absolute” ) way. Each
Customer’s Requirement (WHATs row) contributes to the importance of all
related Technical Characteristics (HOWs) for a total amount depending on
its own importance (from the customer’s point of view).

These slides can not be reproduced or distributed without permission in writing from the author. 69
1.2. VOC
TECHNICAL& QFD [HOWs]
CHARACTERISTICS

from the custo-mer's

Relationship strength
IMPORTANCE

ROW TOTALROW TOTAL


x1

x3

x4

x5

IMPORTANCEIMPORTANCE
Characteristic x2Characteristic x2

ROW TOTAL ROW TOTAL


point of view point of view
CUSTOMER'S

Characteristic, xCharacteristic,

Characteristic, xCharacteristic,

Characteristic, xCharacteristic,

Characteristic, xCharacteristic,
"Relative"

Technical

Technical

Technical

Technical

Technical
REQUIREMENTS TECHNICAL CHARACTERISTICS [HOWs]
[ WHATs ]

from the custo-mer's

Relationship strength
IMPORTANCE

5
CUSTOMER'S

"Relative"

Technical

Technical

Technical

Technical

Technical
Requirement
REQUIREMENTSy1 4 9 3 1 3 9 25 0,160
[ WHATs ] y2
Requirement 4 3 9 12 0,333
WEIGHTS
OF TECHNICAL CHARACTERISTICS
Requirement y [HOWs] 1 4 90,48 3 1 3 9 25 0,160
Absolute WEIGHTS
Requirement y2 (classical)
4 1,00 3 9 12 0,333
st 36 24 40 12 36
(1 mathematical algorithm)
WEIGHTS
Absolute PRIORITIES (classical)
OF TECHNICAL CHARACTERISTICS
(1st mathematical
[HOWs] algorithm)
2 4 1 5 2
Absolute WEIGHTS
Standardized (classical)
WEIGHTS
st
nd 36
1,44 24
1,48 40
3,16 12
0,48 36
1,44
(1 mathematical algorithm)
(2
Absolute PRIORITIES
Standardized (classical)
PRIORITIES
st
nd
(1 mathematical algorithm)
(2
2
3 4
2 1 5 2
3
Standardized WEIGHTS
1,44 1,48 3,16 0,48 1,44
(2nd mathematical algorithm)
Standardized PRIORITIES 9 + 3 + 1 + 3 + 9 = 25  4 / 25 = 0,16
(2nd mathematical algorithm) 33+ 9 2 =112  5
4 /312 = 0,33

0,16 x 3 = 0,48 +
0,33 x 3 = 1,00 +
1,48 =

These slides can not be reproduced or distributed without permission in writing from the author. 70
1.2. VOC & QFD
QFD: mathematical approaches
Literature says that, almost without exception, the 2nd mathematical algorithm
leads to more coherent and convincing results than the 1st one.
To clarify the difference between the two methods, the most simple and direct way is
to consider the case of two Customers' Requirements: for greater simplicity and
clarity, we have assumed that they have the same importance equal to 4 (see
Figure below that compares the two examples above).
TECHNICAL CHARACTERISTICS [HOWs]

from the custo-mer's

Relationship strength
IMPORTANCE

ROW TOTAL
Characteristic, x1

Characteristic, x3

Characteristic, x4

Characteristic, x5

IMPORTANCE
Characteristic x2

ROW TOTAL
point of view
CUSTOMER'S
"Relative"

Technical

Technical

Technical

Technical

Technical
REQUIREMENTS
[ WHATs ]

Requirement y1 4 9 3 1 3 9 25 0,160

Requirement y2 4 3 9 12 0,333
WEIGHTS
OF TECHNICAL CHARACTERISTICS
[HOWs]
Absolute WEIGHTS (classical)
36 24 40 12 36
(1st mathematical algorithm)
Absolute PRIORITIES (classical)
(1st mathematical algorithm)
2 4 1 5 2
Standardized WEIGHTS
nd 1,44 1,48 3,16 0,48 1,44
(2 mathematical algorithm)
Standardized PRIORITIES
(2nd mathematical algorithm)
3 2 1 5 3
These slides can not be reproduced or distributed without permission in writing from the author. 71
1.2. VOC & QFD
QFD: mathematical approaches

 The 1st customer’s requirement y1 depends on many (5) Technical


Characteristics;
 the 2nd customer’s requirement y2 depends on few (2) Technical Characteristics.
 In such conditions, the two algorithms lead to different results.
 The 1st requirement could be achieved only by few technical characteristics, while
the 2nd by many, so the contribution to the importance (criticality) of the 1st one
must be higher than the 2nd one
 In our example, the 2nd Technical Characteristic x2 goes from the 4th position
(according to the 1st algorithm) to the 2nd one (according to the 2nd algorithm): in
particular, the 2nd algorithm reduces the effects of the two 9-level relationships
between the 1st requirement y1 and the Technical Characteristics x1 and x5.

The above example shows that the 2nd algorithm tends to:
 reduce the weight of Technical Characteristics, if many of them are
influencing on the same customer’s requirement (see requirement y1);
 emphasize (in relative) the weight of Technical Characteristics , if only
few of them are influencing on the same customer’s requirement (see re-
quirement y2).

These slides can not be reproduced or distributed without permission in writing from the author. 72
1.2. VOC & QFD
QFD: mathematical approaches

For a proper use of the two above algorithms, we must also take into
account a further aspect.

In order to achieve customer satisfaction when there are several


Technical Characteristics influencing on the same Customer’s
Requirement, QFD is unable to distinguish whether:
 it is necessary that all (or at least some well-defined) Technical Charac-
teristics simultaneously assume their optimal values (AND logic): a typi-
cal example may be safety requirements;
 or it is sufficient that a few among all Technical Characteristics (or among a
well-defined set of them) assume its optimal value (OR logic): a typical
example may be a luxury sensation (it is sufficient that the product has a
few of the particular features that are able to "impress").

These slides can not be reproduced or distributed without permission in writing from the author. 73
1.2. VOC & QFD
QFD: mathematical approaches

It's easy, at this point, realizing that:

 the 1st algorithm is more suitable for a situation following an


AND logic, because it gives a high weight to all the
essential Technical Characteristics: what is particularly
important if the involved Technical Characteristics are of the
type on/off, rather than vary continuously;

 the 2nd algorithm is more suitable for a situation following an


OR logic, because it allocates the “importance” of the
Customer’s Requirement among all involved Technical
Characteristics, in proportion to the “weights” (of their
relationships with this requirement), through the sum of the
weights of the row: by the way literature recognizes this
situation as the most common.
These slides can not be reproduced or distributed without permission in writing from the author. 74
1.2. VOC & QFD
QFD: logical approach
Moreover, the identification of the importance
(weights)
of Technical Characteristics (HOWs)
through a purely mathematical approach
may not always be the best (most accurate) to identify
Technical Characteristics priorities.
As can be both accurate, the attributions of importance to cu-
stomer requirements (WHATS) and the definition (always sub-
jective) of the strength of their correlations with Technical Cha-
racteristics, we are always dealing with an estimate of priorities
at a “macro” level.
Moreover, you can never be entirely sure you have chosen the
best option between the first and second mathematical algo-
rithm.
These slides can not be reproduced or distributed without permission in writing from the author. 75
1.2. VOC & QFD
QFD: logical approach

So we can also follow a logical approach, by:

1. examining the graph of the comparisons with the best


competitors from the customer’s point of view,
2. identifying the (direct) critical aspects which to focus
on,
3. recognizing, through the Correlation Matrix, the more
involved Technical Characteristics,
4. and continuing with a conflict/synergy analysis on
the Roof of the House of Quality to identify any
indirect criticality (in addition to earlier direct) to be
monitored after improving Technical Characteristics
identified above.

These slides can not be reproduced or distributed without permission in writing from the author. 76
1.2. VOC & QFD

3 2

These slides can not be reproduced or distributed without


% permission in writing from the author. 77
1.2. VOC & QFD
Q.F.D.: mathematical and logical approaches

CONCLUSION
The comparison between the results
of the different types of approaches,
mathematical and logical,
allows a fine tuning
and ensures
the most effective final choice.

These slides can not be reproduced or distributed without permission in writing from the author. 78
1.2. VOC & QFD
In the building block "HOW MUCHes", the House of Quality may contain some (or all) of the
subjects (rows) listed below, divided into required and optional (plus any others).
Required WEIGHTS and PRIORITIES: based on one or both of the mathematical
algorithms discussed above.
Optional The FREQUENCY of REPAIRS carried out by Customer Service, sepa-
rately for each Technical Characteristic. This provides technical feed-
back on the present reliability of the considered Technical Characteristics
and, indirectly, monitors the inconvenience of customers having to use Cu-
stomer Service.
Optional The COST of repairs by Customer Service: provide the maximum eco-
nomic benefit achievable by reliability improvements (although difficult to
calculate, the benefits associated with greater customer satisfaction should
be added (at least a rough estimate): sales increasing, overcoming com-
petitors, etc.).
Optional The LEVEL of DIFFICULTY for improvement: technical difficulties and
required resources (economical/time). It is the logical integration (in or-
der to get the best trade-off) to the indications arising from Weights/Priori-
ties and Costs of repairs by Customer Service.
Continues 
These slides can not be reproduced or distributed without permission in writing from the author. 79
1.2. VOC & QFD
Continued 
Required TECHNICAL COMPARISON with best competitors on each Technical Charac-
teristic: each Technical Characteristic is "objectively“ evaluated, both for us and for
the best competitors, through direct measurements and/or derived indicators. In or-
der to compare our position with respect to best competitors evenly for all the Techni-
cal Characteristics (which are measured by different physical units), the physical u-
nits must be converted into levels of customer satisfaction, e.g. referring to a deci-
mal SAE scale. Comparing these results with the “performance” levels attributed by
the customers to us and to the best competitors (WHYs), we can understand, for
example, if we must act on some Technical Characteristics (not yet matching custo-
mer requiorements) or rather on the Company image (if our Technical Characteri-
stics are “objectively” better or equal to that of the best competitors, but are not ap-
preciated by customers according to their actual quality).
Required CRITICALITIES: taken from WHY-column and brought here, following the “Logical
approach” which also provides to complete their list by considering the antithesis
and the synergies reported on the Roof of the House of Quality.
Optional OPPORTUNITIES: taken from the WHY-column and brought here, as was done a-
bove for the criticalities.
Required TARGETS: to be defined in physical units, on the basis of both Critical Require-
ments objectives (specific WHY-column and use of the Logical approach) and all
the above HOW MUCHes rows.

These slides can not be reproduced or distributed without permission in writing from the author. 80
CAR DOOR

EXAMPLE OF DEFINITION
OF PRODUCT TARGETS
(FROM THE CUSTOMER’S

A COMPLETE HOW POINT OF VIEW)


FOR TECHNICAL
CHARACTERISTICS

Position of lever to set the external rear-view mirror

No water seepages (design, verifications and tests)


(HOWs)

MUCHes

Dynamic characteristics of opening position

Load to open by key with low temperature


Static characteristics of opening position

Position of armrest grip to close the door


Load to activate external door handle
Load to activate internal door handle

Load to activate window regulator


F A C T O R S

Position of internal door handle


Shape of internal door handle
Load to close security switch

WINDOW REGULATOR

WINDOW REGULATOR

BLOCK OF THE LOCK

Position of security switch


Load to introduce the key
“BUILDING BLOCK”

Load to close the door

Load to open by key

Operating direction
Winder position

Knob diameter
Knob length

OTHERS
H U M A N
L O A D S
DOOR

LOCK
WEIGHTS OF TECHNICAL
CHARACTERISTICS
Once measured TECHNICAL CHA- Absolute

RACTERISTICS by physical units %

(that are objective), the comparison Weight based priority

(see graph) has to be carried out re- TECHNICAL REPORTS

Customer Service repairs


per 1000 cars
ferring to a decimal scale (discus- Customer Service costs
[ m€ / car ]
sed later) which quantifies custo- Level of organizational
difficulty

mer’s satisfaction. TECHNICAL COMPARISONS


WITH COMPETITORS

Thus we are able to evaluate, from We


Competitor 1
Competitor 2

customer’s point of view:


 which are the Technical Characte-
ristics (HOWs) where we are not CRITICALITIES (*) AND
OPPORTUNITIES (+)

completely satisfied; TARGETS

 if best competitors have, for


them, a better or a worse position
than we have.
These slides can not be reproduced or distributed without permission in writing from the author. 81
Analyzing & Diagnosing the QFD

1. Empty columns
2. Empty rows
3. Rows with no strong relations
4. Multiple rows showing the same relations
5. Relations clusters
6. Rows with too many relations
7. Columns with too many relations
8. Correlations diagonal
9. Many weak relations

These slides can not be reproduced or distributed without permission in writing from the author. 82
Analyzing QFD – 1. Empty Columns

Product / process
requirements without
relations with the
customer needs

 The designers’ voice is


stronger than the
Customer voice

 This product /process


requirement must be
removed from the QFD
table (useless column).

These slides can not be reproduced or distributed without permission in writing from the author. 83
Analyzing QFD – 2. Empty Rows

Customer needs not


related to any Product /
process requirement

 No way to fulfill them.


Some Customer needs
have not been taken into
the proper consideration

 New Product /process


requirements must be
identified to fulfill all the
customer needs

These slides can not be reproduced or distributed without permission in writing from the author. 84
Analyzing QFD – 3. Rows with no strong relations
(see also the previous case)
It’s very difficult to satisfy the
Customer without at least
one Product /process
requirement with a strong
link

 A Customer need could


remain unsatisfied

 Involve the experts to


identify strong relations,
avoiding the risk that a
customer need may not be
fulfilled
These slides can not be reproduced or distributed without permission in writing from the author. 85
Analyzing QFD – 4. Multiple rows showing the same relations

Indicates a problem with


the Customer priorities

 Ensure that the level of


detail of the Customer
priorities is the proper
ones

These slides can not be reproduced or distributed without permission in writing from the author. 86
Analyzing QFD – 5. Relations clusters

It indicates possible
priority problems
either in the Customer
needs, or in the
product / process
requirements

These slides can not be reproduced or distributed without permission in writing from the author.
Analyzing QFD – 6. Rows with too many relations

Probably the Customer


need will have an high
risk to generate defects
and it will require a
deeper analysis

 The customer need


could be the cost, the
reliability or the safety

 Remove it from QFD


and perform a separate
analysis.

These slides can not be reproduced or distributed without permission in writing from the author.
Analyzing QFD – 7. Columns with too many relations

The product / process


requirement is highly
critical to the Customer
Satisfaction (the product
/process requirement could
be the cost, the reliability or
the safety)

 Remove it from QFD and


perform a separate
analysis.

These slides can not be reproduced or distributed without permission in writing from the author.
Analyzing QFD – 8. Correlations diagonal

The Customer needs could


be in reality the the product
/process requirements

 The real Voice Of the


Customer must be collected
and not the Voice Of the
Designers !!!

These slides can not be reproduced or distributed without permission in writing from the author. 90
Analyzing QFD – 9. Many weak relations

No way to take a clear


and defined decision

 The QFD is not so


useful

 Restart the QFD,


following carefully the
steps

These slides can not be reproduced or distributed without permission in writing from the author. 91
Q.F.D. example: Sony’s Walkman

Competitive
position

Moderate
Weak

Strong

These slides can not be reproduced or distributed without permission in writing from the author. 92
Q.F.D. example: Bumper

Low Materials Cost

Chemical stability
Stable dimensions

Camera check
Resistance

Visual Control

of color
Surface

Drop Test
Assembly easiness 3 
Uniformity of color 4 
Resistance to bumps 5  
Duration in time 2  
Few scraps 4 
Low costs 2  
Gap between bumper
and car body 5  
110 23 55 19 18 25 65
Results show that Critical Characteristics of the Process are: Dimensional Stability and Camera Checks

These slides can not be reproduced or distributed without permission in writing from the author. 93
1.2. VOC & QFD

1.2.1
Correlation between
physical measurement
and customer satisfaction
expressed in a SAE scale

These slides can not be reproduced or distributed without permission in writing from the author. 94
1.2.1. Correlation between phys. meas. and customer satisf.

We said earlier that it makes little sense to measure Quality


in "absolute“ terms, because what matters is its "relative" po-
sition in relation to the customer expectations.

However the measures that we are more familiar, because


more easily objectified, are those absolute, i.e. in physical
units.

In the following, is shown a method to correlate physical mea-


surements with the level of fulfillment of customer require-
ments.

These slides can not be reproduced or distributed without permission in writing from the author. 95
1.2.1. Correlation between phys. meas. and customer satisf.

Every customer expectation can be “deployed” into one or


more expected functions.

The responses to each expected function can always be


measured in "absolute" terms, i.e. in physical units (e.g.: kw
for engine power, dm3 for trunk capacity, pixel/mm2 for image
definition and so on): thus they are easy to measure and clo-
sely related to cost.

To make the conversion in terms of customer perception,


we need a correlation between physical values and custo-
mer satisfaction. This correlation can be expressed, for exam-
ple, using the values from 1 to 10 of a decimal SAE scale such
as that shown in the following slide.

These slides can not be reproduced or distributed without permission in writing from the author. 96
The quality level of each expected function may be measured
(both for us and for competitors) on a scale from 1 to 10:
1 being for complete dissatisfaction and 10 for full satisfaction.
Decimal SAE scale and its adjustments (like this below) are commonly used.
POSITIVE All
Middle customers Demanding customers Expert customers
OTHER CUSTOMER'S

APPRECIATION customers
POINTS OF VIEW

NEGATIVE
High Moderate Low Very low No one
PERCEPTION
Heavy Slight Slight
SENSATION Unbearable ness Irritation Little bother No bother
irritation irritation bother

REACTION Dispute Remonstrance Complaint Acceptability Appreciation Excitement


Close to the
JUDGMENT

LABELS Very bad Bad Very poor Poor acceptabi- Acceptable Adequate Good Very good Excellent
SCALE
MAIN

lity limit

GRADES 1 2 3 4 5 6 7 8 9 10

PRACTICAL RESULT NOT ACCEPTABLE A C C E P T A B L E


CORRECTIVE ACTIONS Radical and prompt
Quick corrective action To improve Attention is required No action
NEEDED intervention

Whenever it's possible, in engineering applications,


variables (suitable to numerically quantifying) are preferred to the word-descriptions (attributes).
In a SAE scale, words, expressing a variety of points of view,
help to better define the scores (numbers)
and so they contribute to reducing the dispersion of judgments,
because customers-judges, although each maintains his own "sensitivity",
should be able to refer to the same scale of values.
These slides can not be reproduced or distributed without permission in writing from the author. 97
Satisfying a customer’s need may require the contribution of
several (measurable) quantities.
Their combination in a mathematical expression
(to be defined),
becomes an indicator which represents an overall and indirect
physical measurement of the actual need.

Before assuming it as an indicator, the mathematical expres-


sion has to be verified by a “customer-jury test” pointing out
the level of correlation with the customer satisfaction (usual-
ly evaluated by a SAE scale).
 This approach is helpful because there are many situations in which customers
perceive overall sensations, but they are not able to judge each of the physical
quantities affecting these perceptions.
 If the pointed out level of correlation (customer-jury average satisfaction vs.
physical measurements or derived indicators) should be too low, we must try
again using more quantities and/or more fitting and sophisticated indicators,
while the sample used and the collected judgments may be left the same.

These slides can not be reproduced or distributed without permission in writing from the author. 98
1.2.1. Correlation between phys. meas. and customer satisf.
INDICATOR generation using “CUSTOMER-JURY TEST”
1. Selecting a sample of people, who will evaluate a set of car mo-
dels appropriately chosen and with different characteristics;
2. Customer-jury test, conducted using an assessment question-
naire containing questions consistent with the performance tree
(which is an orderly and structured list of all customer requirements at a
whole car level);
3. Identifying specific assessment aspects mainly influencing
the overall performance perception;
4. Objective characterization, through the mathematical elabora-
tion of the signals acquired in appropriate tests/maneuvers on the
instrumented cars;
5. Definition of a correlation model between objective/subjective
parameters, in order to understand the link between the perfor-
mance level perceived by customers and the car objective cha-
racteristics.
These slides can not be reproduced or distributed without permission in writing from the author. 99
1.2.1. Correlation between phys. meas. and customer satisf.

Example: TAKE-OFF indicator

Take-off is mainly influenced by :


 weight of car in running conditions, W [t]
 torque delivered by engine, T [kgm]
 car speed at 1000 r.p.m. in the 1st gear, S [km/h].

The value of speed at 1000 r.p.m. in the 1st gear stands for the whole
kinematical chain, from engine to wheels (gears, differential unit, rolling
height, …). The higher the value of this speed, the lower the overall trans-
mission ratio (with low fuel consumption and good comfort, but poor per-
formance and poor take-off).

These slides can not be reproduced or distributed without permission in writing from the author. 100
Knowing that, the take-off is :
• proportional to the torque T delivered by the engine
• inversely proportional to the weight W of the car in running condi-
tions multiplied by the speed S at 1000 r.p.m. in the 1st gear,
a take-off indicator, TOI, may be assumed as:
TOI = T / (W . S)
The fitting must be validated on several car models (few but representati-
ve). Competitors models may also be employed.
Once car technical characteristics have been measured and take-off indica-
tors have been calculated, the customer satisfaction may be evaluated on a
SAE scale by a customer-jury, which of course has not been acquainted with
the technical characteristics and related indicators.
The output is a Table like the following:
Car weight Speed Average of
in running Torque at 1000 r.p.m. Take-off subjective
conditions [kgm] in the first gear index judgments
[t] [km/h] [SAE scale]

Model 1 1,20 10 9,5 0,88 5,0


Model 2 1,25 12 9,0 1,07 5,5
Model 3 1,20 11 8,0 1,15 6,0
Model 4 1,15
These slides can not be reproduced or distributed without permission in writing from the author.
13 9,0 1,26 7,0 101
The following diagram shows the relationship between physical measure-
ments (objective) and customer average perceptions (subjective) evaluated
in a SAE scale.
8,0
Averaged subjective judgments [SAE scale]

7,5
y = 5,0836x + 0,352
2
R = 0,9034; r = 95%
The good
7,0
correlation
6,5 (r = 95%)
6,0
experimentally
confirms the
5,5
fitting of the
5,0
assumed
take-off
4,5
indicator.
4,0
0,80 0,85 0,90 0,95 1,00 1,05 1,10 1,15 1,20 1,25 1,30

Take-off indexes (objective physical evaluation)

 Such a diagram is needed for each expected function.


 These diagrams have to be periodically updated (let’s say every 2÷5
years) and also updated every time it may be thought that something has
changed in the marked, in the product mission profile, in the customer de-
mands and so on. 102
These slides can not be reproduced or distributed without permission in writing from the author.
1.2. VOC & QFD

1.2.2
Practical impacts
of “House of Quality”

These slides can not be reproduced or distributed without permission in writing from the author. 103
1.2.2. Practical impacts of “House of Quality”

To sum up, targets are established on the basis of:


1. how much the customer perceives their requirements met
by us in comparison with the best competitors (WHY);
2. the Technical Characteristics priority from the customer
point of view (HOW MUCHes);
3. the feasibility of improving Technical Characteristics, taking
into account the budget availability (HOW MUCHes);
4. an objective assessment, based on physical measure-
ments (HOW MUCHes) of where each Technical Charac-
teristic (HOW) is placed, both in comparison with
competitors and in transverse (row) comparisons with the
other Technical Characteristics (what is made possible by
converting physical measurements into customer’s
perception through the Correlation Graphs).

These slides can not be reproduced or distributed without permission in writing from the author. 104
FINAL RESULTS
(example with
QFD in Italian)

PRIORITIES OF TECHNICAL CHARACTERISTICS


(from customer’s point of view)

SELECTION OF TECHNICAL CHARACTERISTICS TO BE DEVELOPED WEIGHTED


(by subsequent design-QFDs) TARGETS
for all
technical
characteristics
These slides can not be reproduced or distributed without permission in writing from the author. 105
1.2.2. Practical impacts of “House of Quality”

QFD main outputs

HOWs
Technical
Characteristics DESIGN CHOICES
Requirements

EXPECTED
Customer’s

FUNCTIONS
WHATs

WHYs

with their DESIGN


SPECIFICATIONS
TOLERABILITY
RANGES
PROCESS
MUCHEs

TARGETS CAPABILITIES
HOW

for each
Technical
Characteristics

These slides can not be reproduced or distributed without permission in writing from the author. 106
1.2.2. Practical impacts of “House of Quality”

 The case of the car door, whose exposure has just


been completed, is an example of detail-QFD.

 But it may also be useful to implement a QFD at a ma-


cro level, macro-QFD, for example to evaluate the dif-
ferences in terms of Customer Requirements (WHATs)
and Technical Characteristics (HOWs) between car
models of class A and car models of class D (see next
slide).

These slides can not be reproduced or distributed without permission in writing from the author. 107
1.2.2. Practical impacts of “House of Quality”
Comparison
between
2 macro-QFD

Of course, in macro-QFD like these, Technical Characteristics (HOWs) are only an


indication guidance for basic choices: for more details, we need detail-QFDs.
These slides can not be reproduced or distributed without permission in writing from the author. 108
Q.F.D. carries the “Voice Of Customer”
PRODUCT
CHARACTERISTICS through the product development phases

IMPORTANCE
CUSTOMER
CUSTOMER

NEEDS
NEEDS

PRODUCT SYSTEM/COMPON.
CHARACTERISTICS CHARACTERISTICS
TARGETS

CHARACTERISTICS
CHARACTERISTICS

PRODUCT
PRODUCT

TARGETS
Phase 1: Product definition

PROD. PROCESS
Product QFD (Marketing) SYSTEM/COMPON. CHARACTERISTICS
CHARACTERISTICS
TARGETS
identifies all main in-

SYSTEM/COMPON.C

CHARACTERISTICS
SYSTEM/COMPON.
HARACTERISTICS
terventions to satisfy

TARGETS
customer needs.
Phase 2: Design
PROCESS CONTROL
PARAMETERS

CHARACTERISTICS
CHARACTERISTICS

PROD. PROCESS.
PROD. PROCESS
PROD. PROCESS
CHARACTERISTICS

TARGETS
TARGETS

Design QFD
For every point to be developed, Phase 3: Industrialization
2 QFD have to be realized:
PROCESS CONTROL
• the first one to choose the more ap- PARAMETERS
TARGETS
propriate design solution;
Technology QFD
• the second one to optimize all de- Phase 4: Production
Targets related to the
tails of the selected design solution.
Company strategies
may also be added
(i.e. deadline). Production QFD
• operation sheets

McCabe Method • visual management


• etc.

These slides can not be reproduced or distributed without permission in writing from the author. 109
1.2.2. Practical impacts of “House of Quality”
McCabe Method
It is worth emphasizing that … Benefits
• From a QFD to the next, only the few most criti-
cal or most important aspects have to be deve- • Rational approach to be followed at least in
loped. these early activities, to prevent dealing with
• Usually, for the Technical Characteristics on emergencies.
which we are worse than the best competitors,
we must try to seize this opportunity to overco- • Agreement of all Business Functions: it is
me competitors, becoming "best in class" (in- essential to avoid later conflicts and it is
stead of "follower“). also the main way to make optimized choi-
• In the design phase, there are 2 categories of ces throughout the Product Development
QFD : the first one is aimed at finding the best Process.
concept solution (e.g. the type of an air
conditioning system: manual/automatic, single- • VOC up to the Production: it is clear that
zone/dual-zone, etc.) and the second one is the process management benefits from ex-
aimed at optimizing some details of the chosen plicit knowledge of what the customer per-
concept solution (e.g. selecting the best shape ceives as a result of the occurrence of po-
for air vents). tential defects.
• In the Industrialization phase, may be neces-
sary to take into account objectives insignificant • Opportunity for refinement of rules and
to the final customer, but important for the com- procedures: in a way consistent with the
pany: for example, if the time to upgrade the different Business Functions and always
production process is reasonable, it may be aimed to the customer’s requirements.
convenient to contact a best in class supplier,
while, if time is pressing, it may be better to pre-
fer the physically closest one.
These slides can not be reproduced or distributed without permission in writing from the author. 110
1.2.2. Practical impacts of “House of Quality”

These slides can not be reproduced or distributed without permission in writing from the author. 111
1.2.2. Practical impacts of “House of Quality”
QFD requires the involvement of:
 TOP MANAGEMENT: without his full support, the project will not produce effective re-
sults;
 ALL BUSINESS FUNCTIONS: their collaboration (if not the physical presence of a re-
presentative) must be warranted for all Inter-function Working Groups which alternate
at the successive stages.
 Marketing
Because of the need to make decisions that
 Research & Development take into account objectives and constraints
of all Business Functions :
 Product Planning
 The Representatives of Business Functions
 Design must possess sufficient competence and
authority to ensure a relatively rapid achie-
 Industrialization vement of final decisions (the "I will report
(design of the production process to my head and then refer" should be ad-
and related technological aspects ) mitted with extreme parsimony);

 Production  With the progress of activities, the Groups


(production process management: composition may vary (but not too much
and related human aspects ) because of the need for constant agree-
ment among all Business Functions ); but
 Testing surely will change their leadership, accor-
ding to the needs of each stage/substage.
 Sales

 Customer Service

These slides can not be reproduced or distributed without permission in writing from the author. 112
1.2.2. Practical impacts of “House of Quality”

NUMBER
OF CORRECTIVE ACTIONS

CLASSICAL
APPROACH

QFD
METHOD

PRODUCT DESIGN PRODUCTION AFTER SALES


DEVELOPMENT

These slides can not be reproduced or distributed without permission in writing from the author. 113
1.2.2. Practical impacts of “House of Quality”

NUMBER OF CORRECTIVE ACTIONS


comparison between a Japanese and an American manufacturer
JAPANESE AMERICAN
manufacturer (*) manufacturer (*)

N° OF CORRECTIVE ACTIONS PER MONTH

START OF PRODUCTION

MONTHS TO/FROM PRODUCTION STARTING


(July and August count as one month alone)

(*) Source: American Supplier Institute

These slides can not be reproduced or distributed without permission in writing from the author. 114
1.2.2. Practical impacts of “House of Quality”

Strong points of QFD

 Systematic focus on the customer;

 Involvement, right from the start, of all Business


Functions;

 Synthetic and officially recognized reference for the


priorities according to the customer’s point of view;
 Connected controls on all subsequent stages of pro-
duct development;

 Help in the consolidation of the obtained results,


through the definition of operational standards.

These slides can not be reproduced or distributed without permission in writing from the author. 115
1.2.2. Practical impacts of “House of Quality”

D e c i s i o n s t r a t e g y

COMPETITIVE POSITION

than competitors
Better

customer’s requirement
COST STRONG POINTS

of the examined
IMPORTANCE
REDUCTION (to advertise)

high
low

POSTPONABLE PRIORITY
ACTIONS ACTIONS
than competitors
Worst

These slides can not be reproduced or distributed without permission in writing from the author. 116
1.2.2. Practical impacts of “House of Quality”

Customers simulation

 In field marketing surveys are often very expensive;


 hence, to “simulate” customers and predict their requirements, arises the
interest in “homemade” inquiries: usually, company employees are also
users (i.e. customers) of company’s products;
 but we must exercise caution, because the company’s Specialists are often
oriented to the innovations more than the average customer, so that, in such
cases, we could get biased results;
 however, when the customer’s requirements are known or relatively easy to
predict (e.g. when adjusting a car model for a specific foreign Market), the
“homemade” QFD applies and, compared to traditional methods, offer two
advantages:
 a cross-checking to make sure we have not forgotten anything;
 and the consideration of the relative importance of customer’s require-
ments which leads to a better “trade-off".

These slides can not be reproduced or distributed without permission in writing from the author. 117
1. Design for Quality

1.3
Quality Measurements

These slides can not be reproduced or distributed without permission in writing from the author. 118
1.3. Quality Measurements

CUSTOMER MANUFACTURER

EXPECTED QUALITY Q.F.D.


PLANNED QUALITY
(IDEAL) Qex Qpl

Design
&
Experimental Process
evaluation

PERCEIVED QUALITY
ATTAINED QUALITY
(ACTUAL) Qpe Qat

Field feedback

Differences greater than what planned


are due to Q.F.D. inadequacy.

These slides can not be reproduced or distributed without permission in writing from the author. 119
1.3. Quality Measurements

Based on the QFD results, Expected Quality Qex


can be detailed by the mathematical expressions
of the expected functions Fi
and/or of the derived indicators,
(both measured on a SAE scale).

The SAE scale value of the highest levels


of customer satisfaction,
for both indicators and expected functions Fi,
is always, by definition, equal to 10.

These slides can not be reproduced or distributed without permission in writing from the author. 120
1.3. Quality Measurements
An estimate (in relative terms) of the Planned Quality Qpl
can be obtained from its ratio with the Expected Quality Qex.
To do this, targets must be assumed
for each expected function Fi identified with the help of QFD.
The Planned Quality Qpl is almost always less than or equal to the
Expected Quality Qex.
Denoting by Wi the weight (= importance, WHYs)
that customers assign to each customer’s requirement
and consequently to all the related expected functions Fi,
we can write:
Fpl and Fex can be evaluated both by physical units
Fpl or by a SAE scale, correspondently we have:
S Wi . ( ----- )i approximate results if Fpl and Fex are expressed in
Fex physical units (assuming that each correlation graph
is linear);
Qpl / Qex = --------------------
S Wi
correct results if Fpl and Fex are expressed in a
SAE scale: in this case Fex is always equal to 10.

where Fex is the maximum of satisfaction that the customer is able to per-
ceive.
These slides can not be reproduced or distributed without permission in writing from the author. 121
1.3. Quality Measurements

In a similar way, we can evaluate the ratio Qat/Qpl


between Attained Quality Qat and Planned Quality Qpl,
for example by experimental tests:
Fat and Fpl can be evaluated both by physical units or
Fat by a SAE scale, correspondently we have:
S Wi . ( ----- )i approximate results if Fat and Fpl are expressed in
Fpl physical units (assuming that each correlation graph
is linear);
Qat / Qpl = --------------------
S Wi correct results if Fat and Fpl are expressed in a SAE
scale.

Of course, the Attained Quality Qat


is rarely greater than the Planned Quality Qpl.

These slides can not be reproduced or distributed without permission in writing from the author. 122
1.3. Quality Measurements
At this point, we can also directly relate the
Attained Quality Qat with the Expected Quality Qex:
Fal and Fex can be evaluated both by physical units
Fat or by a SAE scale, correspondently we have:
S Wi . ( ----- )i approximate results if Fal and Fex are expressed in
Fex physical units (assuming that each correlation graph
is linear);
Qat / Qex = --------------------
S Wi correct results if Fal and Fex are expressed in a SAE
scale: in this case Fex is always equal to 10.

The accuracy of this evaluation will be confirmed


by the Perceived Quality Qpe
estimated on the basis of feedbacks from the field.

If the value of Perceived Quality Qpe


is significantly different from that of Attained Quality Qat,
it means that
QFD has not been effectively developed.
These slides can not be reproduced or distributed without permission in writing from the author. 123
1.3. Quality Measurements

Perceived Quality Qpe is the ratio between


functions perceived and expected by the customers.

It is usual to measure the responses to the expected functions


on a SAE scale, with the advantage of obtaining an evaluation
oriented to the customer’s requirements/perceptions. In this
situation, of course, the ratios between planned and expected, or
between attained and expected, hardly ever exceed 1.
But, of course, nothing prevents us from measuring them in ab-
solute terms, i.e. in physical units. Physical units are different for
each quantity, but it is always possible to calculate relations be-
tween homogeneous quantities (measured by the same units): in
particular, the ratios between planned or realized and expected. In
this situation, it is possible that some of these ratios exceeds 1.
So we can see again, from a new interesting point of view, the
above mentioned examples.
These slides can not be reproduced or distributed without permission in writing from the author. 124
1.3. Quality Measurements
To avoid confusion, we used the character PosterBodoni BT
for Quality functions and indicators when expressed in absolute terms.

Attained Quality = Qat = Fat / Fex


If this ratio is < 1: the attained quality is lower than the expected one
and it is likely that the product is considered poor.
If this ratio is > 1: the attained quality is higher than the expected one
and it is likely that the manufacturer is supporting unnecessary costs,
as in the example of women's fashion magazine.

Perceived Quality = Qpe = Fpe / Fex


To hope that the perceived quality is higher than the expected one may
be considered overly optimistic (ratio > 1).
If this ratio is < 1, i.e. if the perceived quality is lower than the expected
one, it might have been caused by two different situations:
 attained quality is really lower than the expected one and customers
realize this;
 perceived functions do not match the expected ones and therefore it is
likely that there was an error in the assessment of customer’s require-
ments, due to a QFD absent or inefficient, as in the case of the dolls.
These slides can not be reproduced or distributed without permission in writing from the author. 125
1.3. Quality Measurements

1.3.1
Summing-up
on the Quality measurements

These slides can not be reproduced or distributed without permission in writing from the author. 126
1.3.1. Summing-up on the Quality measurements
 What matters is to assess relative quality,
 but it is difficult to measure, especially when forecasting.
 Instead, absolute quality is easy to measure, because it uses physical
units.
 It is then convenient, especially in the forecast, try to estimate the relative
quality as a function of absolute quality, i.e. as a function of measurable
quantities or as a function of “indicators” derived from them (see example
of take-off indicator).
 For that, we must have available the appropriate correlation diagrams
(one for every customer’s requirement); these diagrams have in the ab-
scissa the physical measurements or the indicators derived from them and
in ordinate the customer’s satisfaction expressed in a SAE scale.
 These diagrams are obtained by providing a number (small but represen-
tative) of different products (e.g. car models) and then making them jud-
ged by a properly chosen jury: customer-jury tests.
 If the correlation between the average satisfaction levels (of the judges)
and the physical measurements shows a too low correlation index, we
have to try again with measurements and/or indicators more suitable (and
probably more sophisticated); in this second approximation, tested pro-
ducts and related jury-assessments remain the same, what changes is
only the indicator’s algorithm.
These slides can not be reproduced or distributed without permission in writing from the author. 127
1.3.1. Summing-up on the Quality measurements

The true Attained Quality level is mainly expressed, in relative


terms, as the ratio between what has been “made” by the company
and what has been “perceived” by the customer; namely the ratio
between the actually Perceived Quality, Qpe, and the maximum
that customers are able to perceive, i.e. the Expected Quality, Qex :
Qpe
Attained Quality = Qat = ---------
Qex

Perceived Quality, Qpe, is measured on a decimal SAE scale,


in which Expected Quality, Qex, is always, by definition, equal to
10 (full scale) and therefore this ratio can never exceed 1. In prac-
tice, for each characteristic/function, there must be a diagram that
converts physical measurements into levels of customer’s satisfac-
tion expressed in a SAE scale.
These slides can not be reproduced or distributed without permission in writing from the author. 128
1.3.1. Summing-up on the Quality measurements

To express Attained Quality, Qat, in absolute terms, we can


refer to the functions expressed in physical units and calculate the
ratio between the measure of the Attained Quality (Fat) and the
measure of the Expected Function (Fex), which, in this case, is no
longer equal to 10, but equal to the physical value corresponding
to a SAE value of 10 in the correlation diagram described above:
Fat
Attained Quality = Qat = ---------
Fex

The term Function indicates a characteristic/performance of the product.


Hence the term Expected Function indicates the (physical) value of this
characteristic which is the maximum that customers are able to percei-
ve/enjoy, instead the term Perceived Function indicates the amount of this
characteristic actually perceived by the customer on the manufactured
product.

These slides can not be reproduced or distributed without permission in writing from the author. 129
1.3.1. Summing-up on the Quality measurements

In conclusion:

 the ratio between Attained Quality and Expected Quality


can be > 1 only if both Qualities are expressed in absolu-
te terms (physical characteristics, almost always associa-
ted with economic aspects = money);
 If expressed in relative terms, Attained Quality can not e-
ver exceed the Expected one, because it could not be per-
ceived by customer and then the ratio will always be  1.

These slides can not be reproduced or distributed without permission in writing from the author. 130
1.3.1. Summing-up on the Quality measurements

What really matters is Perceived Quality, expressing the appreciation from the
customer. But it can be measured only by interviewing the customers after the
sales of the new product have begun. In this situation, we can ask the customer
an overall judgment ("relative" measurement) and also, with his help, we can build
the "satisfaction / physical_units correlations diagrams" for each expected func-
tion, through which it will be possible (as a verification) rebuild the previous overall
judgment through "absolute“ measurements. Perceived Quality, as measured by
the customers, represents the true value of Attained Quality actually achieved by
the product.

So Attained Quality (where "attained" means obtained/realized at the end of the


production process) can be regarded as an estimate of Perceived Quality during
the final experimental tests before starting to sell the product (see also Page 97).
Before this time, we know neither Qpe (Page 102) nor Fpe (Page 103) and so we
must settle for an estimate, of course in "absolute" terms. Instead of using the
theoretical relationship Qpe = Fpe / Fex , we can write:
^
Qpe = Qat = Fat / Fex
Of course, this estimate is truthful only if both the initial QFD and all correlation
diagrams are correct.

These slides can not be reproduced or distributed without permission in writing from the author. 131
1.3.1. Summing-up on the Quality measurements

In conclusion, the expression of Page 103 is "theoretical" before the commercial


launch and then becomes "practice", and ideally equal to that of Perceived Quality,
after the commercial launch and the acquirement of customer’s evaluations. In other
words, as companies strategies are clearly aimed to the customer’s satisfaction,
Attained Quality should be fairly close to the Perceived Quality (and this latter to the
Expected Quality). However, we can not be sure that we have detected all the cu-
stomer’s expectations (error of QFD), or that customers are able to appreciate all
the ("absolute") qualities of our product (absolute Attained Quality too high).

In practice, the Attained Quality (measured at the end of the Product Development
Process) and the Perceived Quality (which constitutes a check on the field for the
actual Attained Quality) may differ for errors in the initial QFD and/or in the "corre-
lation diagrams”.

These slides can not be reproduced or distributed without permission in writing from the author. 132
1. Design for Quality

1.4
CAR “CLASSES”

These slides can not be reproduced or distributed without permission in writing from the author. 133
1.4. Car “classes”

Each new car model, in advance of its commercial launch, is


associated with a car class on the basis of the following fea-
tures:
• body type;
• dimensions (length, width, height, wheel base);
• range of offered engines;
• placement price;
• source (= derivation) model (if any).
If these characteristics give rise to contradictory results, are u-
sed as “drivers”: placement-price and source-model.

These slides can not be reproduced or distributed without permission in writing from the author. 134
1.4. Car “classes”
Examples of detailed criteria for allocation

Range
Range of gasoline engines
motorizzazioni bz - Italia
offered on the80Italian
(volumi - 90%) Market

3,2

3,0

2,8

2,6

2,4

2,2
Cilindrata [c.c.]

2,0

1,8
Displacement

min
1,6 max

1,4 modale

1,2

1,0

0,8

0,6

0,4

0,2
A B C D E
0,0

These slides can not be reproduced or distributed without permission in writing from the author. 135
1.4. Car “classes”
Examples of detailed criteria for allocation

Forbice
Range prezzi per segmento
of prices - Italia
- Italian Market
(80%-90% volumi)
60000

55000

50000

45000

40000

35000
min
min
[€]
Euro

30000 max
max
Euro

weighted
medio average
pesato
25000

20000

15000

10000

5000
A B C D E
0

These slides can not be reproduced or distributed without permission in writing from the author. 136
1.4. Car “classes”
TABLE OF CAR CLASSES
CAR CLASSES EXAMPLES
# Symbol D e f i n i t i o n OF FIAT-LANCIA-ALFA CAR MODELS
 FIAT Seicento
A Small cars: compatible with general use, but especially suitable for urban use.  FIAT Panda
 FIAT 500

 LANCIA Ypsilon
B Utility cars for general use.  FIAT Punto & Grande Punto
 ALFA ROMEO MiTo
M a i n

 ALFA ROMEO Giulietta


C Low-medium cars  FIAT Bravo
 LANCIA Delta

D Medium cars  FIAT Croma  ALFA ROMEO 159

E Medium-high cars  LANCIA Thesis

 FIAT Panda Van


 FIAT Scudo
 FIAT Punto & Grande Punto Van
Comm. Commercial vehicles
 FIAT Fiorino
 FIAT Strada
 FIAT Ducato
 FIAT Doblò Cargo

Vehicles with high image and performances: the best suitable variable to define
G them is their price.
 FERRARI all models  MASERATI all models

Body coupé & spider, characterized (among other things) by a car height < 1370  ALFA ROMEO GT
H  FIAT Barchetta  ALFA ROMEO Brera
O t h e r s

mm.  ALFA ROMEO Spider


Sport Utility Vehicles (SUV), defined (among other things) by a ground clearance
I > 170 mm and by the kind of wheel drive.
 FIAT Sedici

Multi purpose vehicles, characterized by the modularity of the interior


L0  FIAT Idea

L rooms and by a vehicle height > 1520 mm (the class L is further divided into L1  FIAT Multipla
subclasses L0, L1 and L2).
L2  FIAT Ulysse  LANCIA Phedra

Vehicles for transporting people, derived from commercial vehicles, in turn


P originated from class B cars.
 FIAT Qubo  FIAT Doblò

These slides can not be reproduced or distributed without permission in writing from the author. 137
1. Design for Quality

1.5
QUALITY, RELIABILITY
AND NEED OF
A STATISTICAL APPROACH

These slides can not be reproduced or distributed without permission in writing from the author. 138
1.5. Quality, Reliability and need of a Statistical Approach

1.5.1
Quantities related to Quality

These slides can not be reproduced or distributed without permission in writing from the author. 139
1.5.1. Quantities related to Quality

QUALITY consists of many aspects,


one of them is RELIABILITY.

In turn, the RELIABILITY discipline


includes several other quantities
(in addition to the strict reliability),
which help to complete the outline
of the product performances.

These slides can not be reproduced or distributed without permission in writing from the author. 140
1.5.1. Quantities related to Quality

These slides can not be reproduced or distributed without permission in writing from the author. 141
1.5. Quality, Reliability and need of a Statistical Approach

1.5.2
Reliability definition

These slides can not be reproduced or distributed without permission in writing from the author. 142
1.5.2. Reliability definition

RELIABILITY
ability of a product/service to fulfill its required
mission in specified operating conditions and
for a stated period of time.

These slides can not be reproduced or distributed without permission in writing from the author. 143
1.5. Quality, Reliability and need of a Statistical Approach

1.5.3
Other important quantities
related to Reliability

These slides can not be reproduced or distributed without permission in writing from the author. 144
1.5.3. Other important quantities related to Reliability

MANTAINABILITY
ability to enable a fast recovery of product functions, after a failure (when mainte-
nance is carried out following the prescribed standards); its typical indicator is the
average time to repair a failure trep: the shorter the mean time to repair, the higher
the maintainability.

AVAILABILITY
ability of a product/service to be ready to function when the operation is requested:
measured as a ratio between the sum of operating times and the total time (t1 - t0):
t1 - t0 - trep . Pf - S tproactive_maintenance(1)
Availability = ------------------------------------------------------
t1 - t0
where Pf is the mean probability of failure.

(1) Said also “servicing”.

These slides can not be reproduced or distributed without permission in writing from the author. 145
1.5.3. Other important quantities related to Reliability
DURABILITY
ability of a product/system to be used for a long period of time without having to be
replaced, under definite conditions of use and maintenance:
 for a non-repairable system (e.g. a light bulb), durability coincides with reliability,
 while, for a repairable system, durability ends when repair is not technically
feasible or economically profitable: with this definition, the typologies number of
durability-failures may increase over time, with the decrease of the economic value
of the vehicle.
 By convention, durability is considered finished when a product/system (e.g. a vehi-
cle) undergoes a complete overhaul.
 The time interval that expresses durability can be measured in miles, hours, cycles,
etc..
 In its lifetime, each (repairable) system/component may have several reliability-fai-
lures, but only a single durability-failure.

E X A M P L E S
 Car with low reliability and high durability  SIMCA 1000.
 Car with high reliability and low durability  FERRARI F1.
The deformation of an air vent (which is to be replaced) is:
 a durability-failure for the air vent;
 a reliability-failure for the dashboard (facia).
These slides can not be reproduced or distributed without permission in writing from the author. 146
1.5.3. Other important quantities related to Reliability

Operational differences between RELIABILITY and DURABILITY

FAILURES EXPERIMENTAL VERIFICATIONS


Close
Predictability
Number of correlation
of the Use of Tests
Types of different with the
potential specific tests acceleration
failure types of customer's
types of coefficient
failure mission
failures
profile
every type of low
Reliability failure
many difficult necessary rare
(1÷3)
catastrophic high
Durability failures only
few easy unnecessary common
(2÷10)

These slides can not be reproduced or distributed without permission in writing from the author. 147
1.5. Quality, Reliability and need of a Statistical Approach

1.5.4
Dependability definition

These slides can not be reproduced or distributed without permission in writing from the author. 148
1.5.4. Dependability definition

DEPENDABILITY
Dependability is a list of the main features of a product; these features have to be
selected (for example among those defined above) for each product category and,
put together, can give the user justified confidence in the service for which the pro-
duct is intended.

So Dependability is a list and not a numerical indicator


(such as Reliability, Durability, etc.).

E X A M P L E S
Dependability could consist in:
 reliability and maintainability for a car;
 availability and durability for a commercial vehicle.

These slides can not be reproduced or distributed without permission in writing from the author. 149
1.5. Quality, Reliability and need of a Statistical Approach

1.5.5
Need of a statistical approach

These slides can not be reproduced or distributed without permission in writing from the author. 150
1.5.5. Need of a statistical approach
D E T E R M I N I S T I C A P P R O A C H (safety margin and safety factor)

MATERIAL

ACTUAL SIZE

SURFACE CONDITIONS
(roughness, hardening, etc.)
ASSEMBLING
Safety factor = 4 / 2 = 2

safety margin
COMPONENT
RATED LOAD

STRENGTH

0 1 2 3 4

These slides can not be reproduced or distributed without permission in writing from the author. 151
1.5.5. Need of a statistical approach

P R O B A B I L I S T I C A P P R O A C H ( closer to reality )

MATERIAL

ACTUAL SIZE

SURFACE CONDITIONS

LOAD CONDITIONS
Peak load distribution VARIABILITY
(load spectrum) COMPONENT STRENGTH
In operation VARIABILITY ASSEMBLING

Overlapping warning about possible failures

These slides can not be reproduced or distributed without permission in writing from the author. 152
1.5.5. Need of a statistical approach
In most thrillers

the guilty

is the butler

In companies, (wrongly or rightly) often happens that,


to be blamed for the low quality, is the supplier.
Let us imagine that a bad supplier
has lost control of its production process,
which now produces
items whose main characteristic has
the same mean as before
but with a “variation” (measured by standard deviation)
much larger …
These slides can not be reproduced or distributed without permission in writing from the author. 153
1.5.5. Need of a statistical approach

OPTIMIZATION BY REDUCING STRENGTH VARIABILITY

MATERIAL

ACTUAL SIZE
REDUCED VARIABILITY
OF COMPONENT STRENGTH
SURFACE CONDITIONS

LOAD CONDITIONS
VARIABILITY
ASSEMBLING

Indication of remaining possibility of failure

These slides can not be reproduced or distributed without permission in writing from the author. 154
1.5.5. Need of a statistical approach

Note that if the strength distribution, without any change


in the average value, expand (or narrows) because its di-
spersion increases (or decreases), the failure probability
changes (even dramatically), while the safety factor re-
mains unchanged.
This refers to Probabilistic Design that will be discussed a little later.

But it is already clear that,


to effectively address these problems ,
it is necessary to use a statistical approach.

These slides can not be reproduced or distributed without permission in writing from the author. 155

S-ar putea să vă placă și