Ergo

Rebuilding a product design process (PDP) with creativity techniques can effectively assist designers in encouraging

1. Introduction management, research and development, produc-tion, marketing, and decision-making, and re-

ARTICLE IN PRESSNew product development (NPD) is a complexarea of ideas and innovations, involving strategy,

quires linking science and technology/inventionor innovation with the marketplace. The success ofnew products mostly depends on new productdevelopment process and management (Chaturve-di and Rajan, 2000), including effective attainment

Corresponding author. Tel.: +886-6-275-7575; fax: +886-6-274-6088

0169-8141/$ - see front matter r 2004 Elsevier B.V. All rights reserved.doi:10.1016/j.ergodesign creativity, further improving the overall performance of innovative product design. Accordingly, this research

provides a new approach to the design process for industries by taking advantage of a creativity-based design process to

achieve the goals of innovative product design.

r 2004 Elsevier B.V. All rights reserved.

Keywords: Creativity; Design process; Product design; Decision-making model; Systematic approachIn todays highly competitive and uncertain market environment with short product life cycles, product development

must not only satisfy the quality and speed of production, but it must ensure that products themselves have included

innovative values. As creativity plays an important role in new product development (NPD), it can be utilized in search

of novel ideas for innovative product design, and also can be regarded as a helpful tool in advancing NPD output.

In this paper, we developed a creativity method based on the naturally sensuous ability of human beings. We also

proposed a creativity-based design process integrating some systematic design methodologies with a developed

creativity tool. The essence of this proposed design process is not the sole performance of the creativity tool but the

coherent efforts among each involved process technique.

To prove the practicability of this design process, a case study was conducted according to various procedures that

were followed. By applying evolutionary thinking, the sensuous association method, and other systematic approaches

to the design process, a number of concept solutions were produced. An optimal solution was then determined by using

an operative decision-making model based on the weighted generalized mean method.

Relevance to industryAbstractShih-Wen Hsiao, Jyh-Rong ChouDepartment of Industrial Design, National Cheng Kung University, Tainan, Taiwan 70101, ROC

Received 21 February 2003; accepted 10 May 2004

Available online 12 August 2004A creativity-based design process for innovative product designInternational Journal of Industrialn.2004.05.005nomics 34 (2004) 421443

ARTICLE IN PRESS

nal oof knowledge and using it as part of the productdesign process (Poolton et al., 2000). Cooper(1996) indicated that new product development isa vital endeavor for modern industries. Companiesmust learn how to innovate effectively, over-hauling their new product processesfrom ideato launchand incorporating winning ideas forsuccessful new products. However, Grifn (1997)reported that, despite widespread acceptance in theliterature, almost 40% of rms surveyed still useno formalized product development process.Product design is a goal-directed problem-solving

activity that relies heavily on human experience,creative thinking, and related knowledge; it shouldbe done by integrating creativity and innovationtools with axiomatic design methodology for durableproduct development (Goel and Singh, 1998). Villa(1998) indicated that continuing innovation inindustries is dened as: promoting a frequent re-design of offered products as well as of theproduction processes required. Also, he proposedan innovation loop concept and introduced aconceptual model of multi-resolution design process.Majaro (1988) suggested the rst step in theinnovation process is the generation of ideas orcreativity, but an important minority of pioneeringdesign theorists has implied that the most valuablepart of the design process is that which occurs insidethe designers head and partly out of reach of hisconscious control (Jones, 1992, pp. 4647). Unfortu-nately, the black-box view of designing can not beclearly expressed in contemporary design theoriesbased on rational analytics. In fact, design work notonly consists of logical rules dealing with distinctprocedures but also innovative thinking producingintricate creativity.Product design process (PDP) is an essential

factor during the early phase of new productdevelopment, which can be considered a complexset of integrated efforts, including generatingideas, developing concepts, modifying details,and evaluating proper solutions. An inappropriateproduct design process not only affects productlife-cycle phases but also increases the possibilityof failure in new product development. Since the1960s, some design scholars, such as Hall (1968),Archer (1971), French (1971), Pahl and Beitz

S.-W. Hsiao, J.-R. Chou / International Jour422(1984), Hubka (1989), etc. have successivelydeveloped many design processes, taking advan-tage of denite methodology to eliminate theillusion of Designer as black boxes. Theconventional design process can offer designers ahelpful tool to product design and thus increasesNPD output. However, Stevens and Burley (1997)indicated that about 3000 raw ideas are required toproduce one substantially new commercially suc-cessful industrial product. They also found thatNPD requires breakthrough creativity because therst ideas for commercialization are almost nevercommercially viable until they have been substan-tially revised through a thought process involvingbranching (Stevens et al., 1999).Bullinger et al. (2000) once stated: the new

millennium has fostered a look towards the future,accompanied by both hopes and fears. Facingtodays global information society, companies mustfocus on the task of meeting competition andchallenges. A high level of complexity, dynamismand uncertainty characterizes the economic situationfor industry and its production facilities. In todayshighly competitive and uncertain market environ-ment with short product life cycles, product designmust not only satisfy the quality and speed ofproduction, but it must ensure that productsthemselves have included innovative values. Stevenset al. (1999) demonstrated that positive correlationswere found between prots resulting from NPDproject analyses and the degree of creativity of theanalysts evaluating those projects. McAdam andMcClelland (2002) also documented the perfor-mance results using new product ideas and creativitypractices in the UK textile industry. Therefore, thisresearch focuses on the integration of a creativitymethod and systematic design approaches, andproposes a creativity-based design process forinnovative product design. The proposed processcan effectively assist designers in encouraging designcreativity, further improving the overall perfor-mance of innovative product design.

2. Theoretical background

2.1. The use of creativity techniques

Creativity is still one of the most mysterious

f Industrial Ergonomics 34 (2004) 421443elements in human thinking. Behavior involves

oneare

(2)

(3)

(4)

listening and feelingcatalysis and outburstcourse.

The SAM is used for refreshment of sensuous-ness and association of inspiration, and it can beregarded as a creativity tool for encouragingdesigners potential to produce innovative ideasquickly. The operation of the SAM is describedbelow:

(1) Put together a design team from a group ofcarefully selected people for a product designproject.

(2) Before the SAM is performed, team membersmust be gathered into a discussion room, so

ARTICLE IN PRESS

Fig

nal of Indtwo different levels of creativity, namely personaland social-cultural (Liu, 2000). Heap (1989)explained that creativity is the synthesis of newideas and concepts through the radical restructur-ing and re-association of existing ones, whereasinnovation is the implementation of the results ofcreativity. Although creativity in the designprocess is regarded as a powerful tool, it is oftencharacterized by the occurrence of a creative leap,which is not easy to be described in logical rules ofconventional design theory. Research in creativitytechniques has been done in the past. For example,Geschka (1996) reviewed three creativity techni-ques: brainstorming, visual confrontation, andmorphological techniques, which were developedand have been used in Germany or German-speaking countries since the 1960s. Krohe (1996)presented a view on managing creativity as well asa summary of 22 creativity stimulating techniques,such as bug listing, goal/wish, manipulative verbs,nominal group technique, wildest idea, wishfulthinking, etc. Cross (1997) constructed a proce-dure based on a creative-leap example andgeneric descriptive models of creative design toprovide further insight into the example. Lugt(2000) described four creative problem solvingexperiments involving meetings in which graphicvariations on brainstorming technique were ex-plored. Dorst and Cross (2001) proposed rene-ments to the co-evolution model, and alsosuggested that creativity in the design processcan validly be compared to burst of develop-ment. These techniques are universally employedfor encouraging creativity and assisting people inproducing a wide variety of novel ideas.

2.2. Sensuous association method

Referring to the applicably existing creativitytechniques, we developed a creativity methodbased on the naturally sensuous ability of humanbeings. This method is what we call the SensuousAssociation Method (SAM), whose main aim isto produce creative ideas by both a designersindividual association and surrounding stimula-tion. As shown in Fig. 1, the method contains four

S.-W. Hsiao, J.-R. Chou / International Jourpersonal behaviors of human sensuousness andtive inspiration is increased through teaminteraction and surrounding atmosphere,which are also considered an input source of(5)

creativity output course.Stimulation: furthermore, the designers crea-structuring course.Describing: describe your mental imageformed after extraction and restructuringcourse.Comparing: compare what you look at andwhat you think aboutextraction and re-mation input course.Thinking: think about their origins and evolu-tionary trendsinference and re-association(1)extrinsic inuence of the environment. Theyexpressed as follows:

Looking: look at the involved thingsinfor-

. 1. Diagram of the sensuous association method (SAM).ustrial Ergonomics 34 (2004) 421443 423arranged that a large number of categorized

(4) During the SAM operation, the participant

(6)

Sofand

sphcanqua

3. O

Sprothecrea

nottoothe

lation matrix, interaction matrix/net, SAM in the

ARTICLE IN PRESS

nal o. The essence of this design process isthe sole performance of the creativity

l but the coherence among involved stages ofsomSAMutline of the creativity-based design process

ince creativity plays an important role in newduct development, it has to be integrated intoproduct design process. Hence, we propose ativity-based design process integratinge systematic design methodologies with theused. In such a natural and interactive atmo-ere, not only team members creative abilitiesbe enhanced, but a great deal of practical andlity ideas can be produced as well.ingaropotential hidden in each members mind.

ince the SAM includes four natural behaviorshuman beings: looking, thinking, comparingdescribing, as well as advantageous surround-stimulation, design creativity can be easilywritten down by the recorder.Consequently, members interactions will sti-mulate each others creative inspirations in ahighly conducive environment. The environ-ment is conducive to opening up the creativehas to compare his/her associations withinformation/pictures observation and contem-plation, extracting novel ideas regarding crea-tive concepts.

(5) While creative images appear in ones mindafter the comparison and extraction, they mustbe described in a sensuous phrase (e.g., lovelyand pliant like a little bird, slender waist of abeauty, as sharp-sighted as an eagle, etc.), andinformation/pictures are stuck on the wall-board to make a very creative environment.

(3) When discussion begins, one member shouldbe assigned as a recorder, the other membersshould look quietly at a section of productphotos on the wall-board. The sensuousassociating can be done by the design teammembers to help them think logically aboutthe origins and evolutionary trends of thetarget product.

S.-W. Hsiao, J.-R. Chou / International Jour424process.divergent stage, morphological analysis, analysisof interconnected decision areas (AIDA) in thetransforming stage, and then a decision-makingmodel based on the weighted generalized meanmethod used to evaluate an optimal designsolution in the convergent stage. The frameworkof this creativity-based design process is shown inFig. 2, and the detailed procedural method ofimplementation is illustrated step by step inSection 4.

4. Implementation methods

To illustrate the detailed procedure and to provethe practicability of the proposed design process,methods of implementation were used and a casestudy of appearance design for an electric scooterwas conducted in this section. According to thethree essential stages and framework of thiscreativity-based design process shown in Fig. 2,the procedures of implementation are described inthe following subsections.

4.1. Description of the target product

Electric scooters are considered a new greentechnology for the 21st century. They are alsoThe proposed design process includes threeessential stages: (1) divergence, (2) transformation,and (3) convergence (Jones, 1992, pp. 6369). Thedivergent stage is an analytic process for searchingthe problem space, which can be described asbreaking the design problem into pieces. Thetransformation stage is a synthetic process forgenerating the solution space, characterized asputting the pieces together in new ways. Theconvergent stage is an integration and evaluationprocess for nding applicable sub-solutions andoptimal design solutions, described as testing todiscover the results of putting the new arrange-ment into practice.There are many design methods and techniques

employed in the overall design process: evolution-ary thinking, systematic structure analysis, corre-

f Industrial Ergonomics 34 (2004) 421443regarded as a viable niche market and a potential

ARTICLE IN PRESS

y thin

arts tr

nal oSystematic structure analysis

Data collection & classification

Evolutionary thinking Evolutionar

Partial p

S.-W. Hsiao, J.-R. Chou / International Jourindustry for many countries, especially in Asianareas (Colella, 2000; Tso and Chang, 2003).In order to formulate product development

policy and establish intended goals of designing,related product information (e.g., product speci-cations, design regulations, technical documents,photos, patterns, etc.) should be collected rst, andthen categorized based on its characteristics andattributes.

(Convergent stage)

Evaluation for an optimal solution

Conclusive design

Integration for concept solutions

FFiinnddiinngg aapppplliiccaabbllee ssuubb--ssoolluuttiioonnss

Determining the dimension parameters

Developing fundamental concept solutions

Defining the boundary conditions

Integration

Decision-

SSeeaarrcchhiinngg tthhee pprroobblleemm ssppaaccee

Sensuou

Building

Constru

Fig. 2. Framework of the creati(Divergent stage)Target product

king diagram

ee

f Industrial Ergonomics 34 (2004) 421443 4254.2. Divergent stage of the design process

As product design is considered a problem-solving activity, the problem spacethe locus inwhich problem solving activities take place (Gold-schmidt, 1997), should be dened by expandingthe boundary of design criteria so as to have alarge and fruitful search space in which to look forproblem solutions.

(Transforming stage)

New product

GGeenneerraattiinngg tthhee ssoolluuttiioonn ssppaaccee

Defining dimension parameters

Developing analogous shape parameters

Assembly/Layout analysis

of the sub-solutions among each subsystem

making model based on the weighted generalized means

s association & conceptual description

a correlation matrix

cting an interaction matrix and an interaction net

vity-based design process.

4.2.1. Evolutionary thinking of the target product

Basically, the wheel was a vital invention forhuman civilization that changed the way ofmoving on land and expanded the domain ofhuman activity. To understand the developmentalhistory and evolutionary pathway toward thetarget product, as well as anticipating its possibledevelopment in the future, we collected a large

number of related product photos and arrangedthem in an evolutionary thinking diagram asshown in Fig. 3. By contemplating the diagram,various kinds of two-wheel vehicles were distin-guished, including the bicycle, motorcycle, motorbike, electric bike, scooter, and electric scooter. Wesubsequently assumed the position of electricscooters is different from traditional fuel scooters.

ARTICLE IN PRESS

S.-W. Hsiao, J.-R. Chou / International Journal of Industrial Ergonomics 34 (2004) 421443426Fig. 3. Evolutionary thinking diagram for the target product.

ARTICLE IN PRESS

derl

nal oA. Basic framework

(Un

(Subsystem )

(System )

S.-W. Hsiao, J.-R. Chou / International JourElectrically powered two-wheel scooters howeverwill become a signicant means of transportationfor many urban areas. Moreover, we also identi-ed some design objectives for developing anelegant, harmonious, and humanistically designedelectric scooter.

4.2.2. Systematic structure analysis of the products

components

For reducing complex processes of designactivity and simplifying design objects, the targetproduct should be analyzed based on a systematicstructure. In this appearance design case, anelectric scooter was separated into three subsys-tems: the basic framework, the outer covering, andfunctional accessories, and their involved compo-

Appearance design of an electric scooter B. Outer covering

C. Functional accessories

Fig. 4. Partial parts tree for appearana.Handlebars (p1)b.Front absorber (p2)c.Main frame (p3)

d.Rear absorber (p4)e.Kick stand (p5) f.Back holder (p6) g.Front wheel (p7)h.Rear wheel (p8)

a.Headset (p9) b.Front fender (p10)

ying part of subsystems )

f Industrial Ergonomics 34 (2004) 421443 427nents were grouped into a partial parts tree asshown in Fig. 4. Due to the underlyingparts of subsystems being more subject to appear-ance designing, they should be consideredmajor units in the partial parts tree of this case.The subsystem component set can be expressed asbelow:

P fp1; p2; p3;y; png; 1

where pj is an underlying part of subsystems,j 1; 2; 3;y; n:

4.2.3. Searching the problem space

Following procedures of evolutionary thinkingand systematic structure analysis, critical problemssuch as product evolution, design objectives, and

c.Mudguard (p11)d.Footrest board (p12)e.Body-board (p13)f.Seat (p14)

a.Rear view mirror (p15)b.Instrument panel (p16) c.Headlight (p17)d.Taillight (p18) e.Front turn signal (p19)f.Rear turn signal (p20)

Speedometer

Odometer

Remaining energy indicator

Power/Charge indicator

Turn indicator

Lamps indicator

ce design of an electric scooter.

paronsear

(1)

ARTICLE IN PRESS

Tabl

List

Item I

s1s2

nal of Ind428phrases can be produced. By deleting thesimilar, preposterous, or indeterminatephrases, the selected conceptual descriptionS

wheIn

theboa22 ccreaitem

(2)ts analysis and grouping were identied. Basedinitial identication, the problem space can beched as follows.

Sensuous association and conceptualdescription.Using the SAM described in Section 2.2 toexpand the boundary of design creativity, alarge number of conceptual descriptions5 A bald-headed guy

s6 To obstruct or hinder in the middle

s7 To be deeply attached to each other

s8 To turn upward

s9 Flimsy wings, stacked up

s10 Looking for a needle in a haystack

s11 No need to cover withs3s4e 1

of the conceptual description phrases

no. Conceptual description phrase

Bulging front and raised rear

Round rear end, like a semi-sphere

Incline at 45

A at platereal level

S.-W. Hsiao, J.-R. Chou / International Jourphrases can be arranged into a set as below:

fs1; s2; s3;y; sqg; 2

re si is a single phrase, i 1; 2; 3;y; q:this case study, nearly 100 photos relevant to

scooters were classied and stuck on the wall-rd. Through the use of the SAM, we chose theonceptual description phrases characterized astive ideas of the problem space. They areized as shown in Table 1.

Building a correlation matrix between con-ceptual description phrases and underlyingparts.In order to properly apply the creative ideas tothe development of subsystem components, apracticable correlation between conceptualdescription phrases and each subsystem partshould be identied.Based on Eqs. (1) and (2), the correlationmatrix, R; can be represented as

R ST P s1; s2; s3;y; sqT p1; p2; p3;y; pn

rijqn; 3

where i=1,2,3,y,q; j=1,2,3,y,n.The entries rij are dened by the following entry

rules:Rule 1: If si correlates with pj ; then rij 1; i.e.,

the conceptual description phrase is able to applyto the subsystem part(s).

s13 A drop of tears

s14 To promote step by step

s15 To treat another as ones equal

s16 An unemployed locksmith

s17 To have edges and corners of view

s18 A drooping buttock

s19 A partial lunar eclipse

s20 A guy with big feet

s21 Round as pearls and smooth as jade

s22 Slender waist of a beautys1R

rij theeachU

casewhiandThedescthedesi

(3)tem no. Conceptual description phrase

2 A y with a huge head

ustrial Ergonomics 34 (2004) 421443ule 2: If si does not correlate with pj ; then0; i.e., the conceptual description phrase andsubsystem part(s) are icompatible withother.sing Eq. (3), a correlation matrix for thestudy was constructed as shown in Table 2,

ch helps us to clarify which pairs of partsphrases are compatible with each other.question of how to turn the conceptualription phrases into real design elements forsubsystem parts was the problem space in thisgn case.

Constructing an interaction matrix and aninteraction net.An interaction matrix is similar to that of acorrelation matrix, which helps us toanalyze the interaction between any two partswithin the underlying parts. The interaction

ARTIC

LEIN

PRESS

Table 2

The correlation matrix between conceptual description phrases and underlying parts for the design

Description

phrases

Basic frameworks Outer covering Functional accessories Others

Handle-

bars

Front

absorber

Main

frame

Rear

absorber

Kick

stand

Back

holder

Front

wheel

Rear

wheel

Headset Front

fender

Mudguard Footrest

board

Body-

board

Seat Rear

view

mirror

Instrument

panel

Headlight Taillight Front

turn

signal

Rear

turn

signal

p1 p2 p3 p4 p5 p6 p7 p8 p9 p10 p11 p12 p13 p14 p15 p16 p17 p18 p19 p20

s1 0 0 0 0 0 0 0 0 0 1 0 0 1 0 0 0 0 0 0 0

s2 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 1 0 0

s3 0 0 0 0 0 0 0 0 0 1 0 0 1 0 0 0 0 0 0 0

s4 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0

s5 1 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 1 0 0 0

s6 0 0 0 0 0 0 0 0 0 1 0 1 0 0 0 0 0 0 0 0

s7 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0

s8 1 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0

s9 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0

s10 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0

s11 0 0 0 0 0 0 0 0 1 1 1 0 0 0 0 0 0 0 0 0

s12 0 0 0 0 0 0 0 0 1 0 0 0 0 0 1 0 1 0 0 0

s13 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 1 1

s14 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0

s15 0 0 0 0 0 0 0 0 0 0 0 1 0 1 0 0 0 0 0 0

s16 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1

s17 0 0 0 0 0 0 0 0 0 1 1 0 1 0 0 0 0 0 0 0

s18 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0

s19 0 0 0 0 0 0 0 0 0 0 1 0 1 0 0 0 0 0 0 0

s20 0 0 0 0 0 0 1 1 0 0 0 0 0 0 0 0 0 0 0 0

s21 0 0 0 0 0 0 0 0 1 1 0 0 1 0 0 0 1 1 0 1

s22 0 0 0 0 0 0 0 0 0 1 1 0 1 0 0 0 0 0 0 0

S.-W

.H

siao

,J

.-R.

Ch

ou

/In

terna

tion

al

Jo

urn

al

of

Ind

ustria

lE

rgo

no

mics

34

(2

00

4)

42

1

44

3429

B

wheT

beloR

1:R

bij R

bij R

comU

scooF

searactishointethepair

ARTICLE IN PRESS

ying

Bb

1

1

0

0

0

0

1

0

1

nal omatrix, B; can also be presented asTable 3

The interaction matrix between any two parts within the underl

Aa Ab Ac Ad Ae Af Ag Ah Ba

Aa 1 1 0 0 0 1 0 1

Ab 1 1 0 0 1 0 0Ac 1 1 1 1 1 0Ad 0 0 1 1 0Ae 0 1 1 0

Af 0 0 0

Ag 1 0

Ah 0

Ba

Bb

Bc

Bd

Be

Bf

Ca

Cb

Cc

Cd

Ce

Cf

S.-W. Hsiao, J.-R. Chou / International Jour430PT P p1; p2; p3;y; pnT p1; p2; p3;y; pn

bijnn 4

re i 1; 2; 3;y; n; j 1; 2; 3;y; nhe entry rules of entries bij are dened asw:ule 1: If pi directly interacts with pj ; then bij

ule 2: If pi indirectly interacts with pj ; then1:

ule 3: If pi has no interaction with pj ; then0:

ule 4: pi and pi itself (i.e. i j) no need to bepared.sing Eq. (4), an interaction matrix for theter design was carried out as shown in Table 3.or extending the use of interaction matrix andching the problem space concerning the inter-ons of underlying parts, an interaction netuld be constructed based on the result of theraction matrix operation. As shown in Fig. 5,constructed network shows not only whichs of parts are connected, but also distinctlyillustrates the relative importance or interdepen-

parts

Bc Bd Be Bf Ca Cb Cc Cd Ce Cf

0 0 0 0 1 0 0 0 0 01 0 0 0 0 0 0 0 0 00 1 1 1 0 0 0 0 0 00 0 1 0 0 0 0 0 0 0

0 0 0 0 0 0 0 0 0 0

0 0 1 1 0 0 0 0 0 01 0 0 0 0 0 0 0 0 0

0 0 1 0 0 0 0 0 0 0

0 0 0 0 1 1 1 0 1 01 1 1 0 0 0 1 0 1 0

0 0 0 0 0 0 0 0 0

1 1 0 0 0 0 0 01 0 0 0 1 0 1

0 0 0 0 0 0

0 0 0 0 0

1 0 0 0

0 1 0

0 1

1

f Industrial Ergonomics 34 (2004) 421443dence among the underlying parts.Analyzing Fig. 5, we found that interaction was

closer between the basic framework and the outercovering, whereas the functional accessories weremore independent than the other subsystems.Considering all parts, the main frame of thescooter (Ac(6+4)) interacted with 10 parts, includ-ing six direct interactions and four indirectinteractions. The body-board (Be(6+3)) interactedwith nine parts, including six direct interactionsand three indirect interactions. The front fender(Bb(4+5)) interacted with nine parts, including fourdirect interactions and ve indirect interactions.The analysis demonstrates the appearance designof an electric scooter should focus on the mainframe for the basic framework design, and thebody-board and the front fender for the outercover design. Most of the functional accessoriesdirectly interacted with the headset (Ba(5+1)),including the rear view mirror (Ca(2+0)), theinstrument panel (Cb(1+1)), the headlight(Cc(2+2)), and the front turn signal (Ce(2+2));hence, they could be considered for collocationwith the headset in the design.

ARTICLE IN PRESS

6+3)

(5+1)

(2+3)d

among the

nal of Ind(6+4)

(1+3) ((3+1)

(1+4)

(1+7)

(2+1)

(1+2) Ac

AAe

Ag

Bd

AfAh

Bf Be

Fig. 5. Interaction net4.3.

Mprodesisolubetakimak

4.3.

Fticamussub

(1)(1+2)

(2+2)

(1+1)

(4+5)

Cd

BbCf

Ce

S.-W. Hsiao, J.-R. Chou / International JourTransforming stage of the design process

any design concepts were created from theblem space through the divergent process. Thegn concepts regarded as radical problemtions are not concrete enough. They shouldtransformed into a denite solution space,ng advantage of positive design elements toe the concepts a reality.

1. Generating the solution space

or converting the design concepts into prac-l design elements, the denite solution spacet be generated corresponding to the threesystems we designated before.

Dening dimension parameters for the basicframework design.According to the analysis of Fig. 5, the basicframework design should be based on themain frame, whose key points are not only toensure the structural strength but also to meetthe dimensional parameters, such as center-of-gravity position (CG), wheelbase, overall

(2): Direct relation between parts : Indirect relation between parts

(4+2)Ab

underlying parts.(1+1)(2+2)

(2+0)

(2+1)(3+3)

Ba Ca

Cc Cb

BcAa

ustrial Ergonomics 34 (2004) 421443 431length, overall width, overall height, etc. Sincethe CG position of an electric scooter isdifferent from that of a conventional fuelscooter (Lai et al., 2003), it must be preciselydetermined. In this design, over-sized and no-inner-tube tires with front-disc and rear-drumtype of brakes were used. The signicantdimension parameters of the basic frameworkdesign are shown in Fig. 6, and they can beconsidered a solution space in teams of thebasic framework subsystem.Developing analogous shape parameters forthe outer cover design.Since outer covering is the most importantfactor in appearance design of an electricscooter, it must be considerably well-thought-out. Based on the result of Table 2,if a conceptual description phrase correlateswith any underlying part (i.e. rij 1), itshould be turned into an analogous shapeparameter by image outlining of the phraseand parts feature, and the image outlines alsoshould be represented in a morphologicalchart.

ARTICLE IN PRESS

nal o432Aconideachaof tlogiparfendtwobodThedesipermpar

(3)S.-W. Hsiao, J.-R. Chou / International Jours shown in Table 4, the morphological charttains some concrete design elements of creatives originating from the SAM operation. Thert can be considered a solution space in teamshe outer covering subsystem. In the morpho-cal chart, we have four pieces of shapeameters for the headset (Ba), six for the fronter (Bb), four for the mudguard (Bc),for the footrest board (Bd), six for the

y-board (Be), and three for the seat (Bf).oretically, there are 3456 sets of conceptualgn possibilities in the solution space byuting and composing these pieces of shape

ameters.

Assembly/layout analysis of the functionalaccessories.

Acleasetstion

Fig. 6. The signicant dimension parameters of Industrial Ergonomics 34 (2004) 421443As we found in Fig. 5, the functionalaccessories are more independent of butnecessary for driving safety because they mustbe appropriately collocated with the othercomponents. The analysis of interconnecteddecision areas (AIDA), an analyticalmethod for dealing with the determinationsof related decision-making concerning designproblems (Luckman, 1967), can be applied tofacilitate layout design of the functionalaccessories.

s shown in Fig. 7, the AIDA method canrly help us to understand all the compatibleregarding the layout conditions of the func-al accessories. There are 15 total pairs of

f the basic framework design.

ARTICLE IN PRESS

r des

nal oTable 4

Morphological chart of the shape parameters for the outer cove

B 1 2 3

S.-W. Hsiao, J.-R. Chou / International Jourcompatible solutions shown in Table 5, which wereidentied by the analysis of the decision areas.These solutions also can be considered a solutionspace in terms of the functional accessory sub-system.

a. Headset

s5 s11 s12

b. Front fender

s21 s3 s1

c. Mudguard

s17 s19 s11

d. Footrest board

s15 s 6

e. Body-board

s1 s13 s18 s19 s3 s22` ` `

f. Seat

s 7 s15 s14ign

4 5 6

f Industrial Ergonomics 34 (2004) 421443 4334.4. Convergent stage of the design process

Because of a large number of problem solutionswithin the solution space, they must be convergedand integrated into formal solutions, not only for

s 21

s 4 s 6 s 11 s17 s22

s 22

s 2 s 18 s 21 s17 s 9` ` `

s 21

` `

ARTICLE IN PRESS

Ceurn signal

Bbront fender

tiona

nal of Ind434Cb Instrument panel

Cc Headlight

Fig. 7. AIDA for the funcndan o

4.4.

Capp

(1)

Tabl

The

Func

Rear

Instr

Head

Taill

Fron

RearCa (Rear view mirror ) Front t

Ba (Headset) (F

Aa(Handlebar)

S.-W. Hsiao, J.-R. Chou / International Jouring applicable sub-solutions but for evaluatingptimal design solution as well.

1. Finding applicable sub-solutions

orresponding to the transforming process, thelicable sub-solutions can be found as follows.

Determining the dimension parameters for thebasic framework design. One vital purpose ofproduct design is production. Thus, theinvolved dimension parameters must be deter-mined. According to the dened dimensionparameters shown in Fig. 6, the signicantdimensions of the basic framework designwere determined as follows: L1=1200mm,L2=1240mm, H1=235mm, H2=520mm,H3=645mm, H4=610mm, H5=745mm,H6=870mm, H7=1020mm, W1=120mm

(2)

(3)

e 5

compatible solutions identied by the analysis of the decision areas

tional accessory Pair(s)

view mirror 3

ument panel 2

light 3

ight 1

t turn signal 4

turn signal 2l acce Taillight

Be(Body-board)

)

ssory layout design.Cd

Cf(Rea r turn signal)

ustrial Ergonomics 34 (2004) 421443(rear wheel width: 130mm), W2=555mm,W3=320mm, u=28, and v=38.Developing fundamental concept solutions forthe outer cover design.Based on morphological analysis shown inTable 4, we can theoretically produce 3,456sets of conceptual possibilities; however, sub-sequent to concept assessment, there were justsix sets of fundamental solutions chosen byrudimentary evaluation of 102 idea sketches.The six sets of fundamental concept solutionsare shown in Fig. 8.Dening the boundary conditions for thefunctional accessory layout design.For rational and practical options, boundaryconditions of the compatible solutionsshould be dened. As shown in Table 6, thedenition of boundary conditions can help us

Compatible sets

CaAa, CaBa, CaBb

CbAa, CbBa

CcAa, CcBa, CcBb

CdBe

CeAa, CeBa, CeBb, CeCa

CfBe, CfCd

4.4.

con

Bcon

ARTICLE IN PRESS

nal oS.-W. Hsiao, J.-R. Chou / International Jourto clarify the layout condition of each compa-tible solution, as well as to make an appro-priate selection within the compatible sets.

2. Integration of sub-solutions for developing

cept solutions

ased on the six sets of fundamentalcept solutions shown in Fig. 8, the concept

Fig. 8. Six sets of fundamental concepf Industrial Ergonomics 34 (2004) 421443 435design for the electric scooter must befurther developed by integrating the involvedsub-solutions of the basic framework and thefunctional accessory subsystems. Throughcomprehensive evaluation and integration,we developed three sets of concept solutionsfor the electric scooter design as shown inFig. 9.

t solutions in outer cover design.

ARTICLE IN PRESS

Table 6

Denition of boundary conditions for the functional accessory layout design

C Compatible solution Compatible

set

Description of the condition

Rear view mirror (Ca) To set on the Handlebar (CaAa) If the shape parameter of Headset is Ba1

or Ba2, the Rear view mirror can be

directly set on the Handlebar.

To set on the Headset (CaBa) If the shape parameter of Headset is Ba3

or Ba4, the Rear view mirror can be

easily set on the Headset.

To set on the Front fender (CaBb) It seems untting for our target product

due to ergonomic considerations, but it

can be considered to be used in a larger

type of electric scooter design.

Instrument panel (Cb) To set alone on the Handlebar (CbAa) If the shape parameter of Headset is

Ba2, the Instrument panel can be

independently attached to the

Handlebar.

To integrate with the Headset (CbBa) If the shape parameter of Headset is

Ba1, Ba3 or Ba4, the Instrument panel

can be integrated into the Headset.

Headlight (Cc) To set alone on the Handlebar (CcAa) If the shape parameter of Headset is Ba1

or Ba2, the Headlight can be set alone

on the Handlebar.

To integrate into the Headset (CcBa) If the shape parameter of Headset is Ba3

or Ba4, the Headlight can be integrated

into the Headset.

To combine with the Front fender (CcBb) If the shape parameter of Headset is

Ba2, Ba4 or the Front fender shape is

Bb2, Bb3 or Bb5, the Headlight can be

combined with the Front fender.

Taillight (Cd) To set alone on the Body-board (CdBe) This is a common use in current design.

Besides, it also can be set on the Seat

(CdBf) or the Back holder (CdAf),

but both of the locations are not

considered feasible because of their

impracticality for mass production.

Front turn signal (Ce) To x on both sides of Handlebar (CeAa) This position provides for good signal

visibility because both sides of the

Handlebar represent the maximum

width of the electric scooter frame.

To attach to the Headset/combine with

the Front fender

(CeBa) /

(CeBb)

This is a common design for todays

fuel-powered scooters.

To integrate into the Rear view mirror (CeCa) The pair of CeCa provides more

signal visibility than CeAa, but the

design may be expensive because of the

need for standardization of parts.

Rear turn signal (Cf) To integrate in the Body-board (CfBe) This design gives a smooth and aesthetic

look.

To be combined with the Taillight (CfCd) This is an easy assembly but provides

poor functional performance.

S.-W. Hsiao, J.-R. Chou / International Journal of Industrial Ergonomics 34 (2004) 421443436

ARTICLE IN PRESS

nal oS.-W. Hsiao, J.-R. Chou / International Jour4.4.3. Evaluation of concept solutions for an

optimal solution

For further conclusive design, an optimalsolution must be determined among the three setsof concept solutions. In this study, we developedan operative decision-making model based on theweighted generalized mean method. The formulaof the weighted generalized means is dened asbelow (Klir and Folger, 1988, pp. 5861):

haa; w Xni1

wi aai

!1=a; 5

Fig. 9. Three sets of concept solutionf Industrial Ergonomics 34 (2004) 421443 437where wi is the parameter weight; wiX0;Pni1 wi 1; ai the relative value of evaluation

factors, the a the variable parameter, NoaoN:The major difference between the weighted

generalized means and the conventional weightedmeans is the involvement of a parameter. By varyingthe a parameter within the interval of N;N; wecan derive a homologous ha value based on a multi-factor analysis, and then draw the a ha curves.The a ha curves can be used to rank thealternatives harmoniously so as to accuratelyidentify an optimal solution among them.

s for the electric scooter design.

ARTICLE IN PRESS

A

6

9

7

6

7

9

7

4

7

9

6

8

Inn

nal oTable 7

Statistical results of the quantitative judgments

Alternative Evaluation factor Good (0.5)

Solution 1 Humanistic 51

Innovative 80

Stylish 62

Aesthetic 58

Solution 2 Humanistic 69

Innovative 87

Stylish 66

Aesthetic 50

Solution 3 Humanistic 66

Innovative 84

Stylish 56

Aesthetic 70

Table 8

The weighted parameters of the evaluation factors

Score Factor

Humanistic

S.-W. Hsiao, J.-R. Chou / International Jour438The decision-making model for evaluating anoptimal concept solution was developed andillustrated in the following steps:

Step 1: Establishing evaluation parametersIn accordance with product design objectives,

we identied four evaluation factors: humanistic,innovative, stylish, and aesthetic, and threeevaluation grades: good (score: 0.5), acceptable(score: 0.3), and ordinary (score: 0.2). Based onquantitative judgments by design team members,the statistical results are shown in Table 7.The relative values for each evaluation factor

involved in the design solutions are summarized ina matrix below:

A aji

340:52 0:83 0:64 0:59

0:68 0:88 0:67 0:50

0:67 0:85 0:58 0:71

264

375Solution 1

Solution 2

Solution 3

:

6

Step 2: Setting up the weighted function

Sum 174 168

Average 4.35 4.2

Naturalization 0.29 0.2cceptable (0.3) Ordinary (0.2) Relative value

2 40 0.52

2 77 0.83

5 53 0.64

3 55 0.59

6 54 0.68

5 80 0.88

0 65 0.67

8 53 0.50

5 58 0.67

3 75 0.85

2 57 0.58

1 58 0.71

ovative Stylish Aesthetic

f Industrial Ergonomics 34 (2004) 421443In order to distinguish the relative importanceof the evaluation factors, we dened the weightedfunction as follows:

wi I1=humanistic I2=innovative I3=stylish

I4=aesthetic; 7

where Ii is the weight for each evaluation factor;P4i1 Ii 1:Further, we used an interaction matrix to

analyze the relatively important degree amongeach pair of evaluation factors, and arranged theminto a weighted parameter list shown in Table 8.With the results, we constructed the weighted

function and converted it into a matrix as below:

wi 0:29=humanistic 0:28=innovative

0:26=stylish 0:17=aesthetic; 8

wi 0:29 0:28 0:26 0:17 : 9

Step 3: The operation of weighted generalizedmeans

156 102

0 3.90 2.55

8 0.26 0.17

ARTICLE IN PRESS

0.5

1.0

nal oh

S.-W. Hsiao, J.-R. Chou / International JourLet i 4; substituting the data of matrixes (6)and (9) into formula (5),i.e.

hja w1aaj1 w2aaj2 w3a

aj3 w4a

aj4

1=a;

NoaoN; j 1; 2; 3; 10

where j represents the jth solution.After calculation, we obtained an a ha relation

diagram as shown in Fig. 10.Step 4: Analysis of the results

-50-100 - 0

SSS

Fig. 10. The a ha relation diagram of the

Y

f(c)

X1 X2X0 a X

0

Fig. 11. Using Simpsons rule and the mean vaf Industrial Ergonomics 34 (2004) 421443 439As shown in Fig. 10, if a > 0; the maximum of hais Solution 2; but if ao0; the maximum of ha isSolution 3. For a more precise comparison, theaverage ha of each solution must be solved.

Step 5: Solving the averages of ha:It can be observed in Fig. 10 that the curve

a ha levels out as the curve approacheseither side of positive innity and negativeinnity. Hence, we applied Simpsons rule andthe mean value theorem (see Fig. 11) to solve theaverages of ha:

10050

olution 1olution 2olution 3

weighted generalized mean solutions.

X

f(x)

iX Xi-1 Xi+1 X2nb

lue theorem to obtain the averages of ha:

Simpsons rule:Z ba

f x dxDDxff x0 4f x1 2f x2

? 2f x2n2 4f x2n1 f x2ng=3: 11

Let a=100, b=100; f x hj a w1aaj1 w2a

aj2 w3a

aj3 w4a

aj4

1=a:Z

characteristics: (see Fig. 13)

new

ARTICLE IN PRESS

S.-W. Hsiao, J.-R. Chou / International Journal o440Areaj 100

100hja da

Dhj100 4hj99 2hj98

? 2hj98 4hj99 hj100=3: 12

Mean value theorem:

_Z b

a

f x dx f c b a; 13

Z 100100

hja da hja100 100; 14

i.e.

hja Areaj=200; 15

where hja represents the ha of the jth solution, j

1; 2; 3:By applying formulas (12) and (15), we obtained

results as shown in Table 9.Step 6: Determining the optimal solutionAccording to the comparison of Table 9, we

found that Solution 3 was superior to Solution 2 interms of the conventional weighted means, buttheir averages were very close. For a more credibleevaluation, the weighted generalized means wasdetermined, and the result showed that Solution 3was better than Solution 2. Accordingly, we choseSolution 3 as the optimal solution in this casestudy.

Table 9

Operational results of the decision-making model

Area ha E

Solution 1 134 0.67 0.6499

Solution 2 138 0.69 0.7028

Solution 3 142 0.71 0.7038

E Pn

wi ai; wiX0;Pn

wi 1i1 i16. Conclusions

Creativity as a tool plays an important role innew product development. It should be integratedinto a product design process. In this research, wedeveloped a creativity method called the sensuousassociation method (SAM), which can be em-ployed to encourage designers potential to pro-duce innovative ideas quickly. Subsequently, weproposed a creativity-based design process inte-Flyidesigratwan, R.O.C. in 2002. We have authorized theng electric motor company to use the patentedgn in future production.ertyTaihown in Fig. 14. In addition, we patented thisstyling design through the Intellectual Prop-Ofce of Ministry of Economic Affairs,desias sased on engineering drawings of the conclusivegn, a prototype of the electric scooter was builtBand detachable battery module provides foroperational stability.(5)metal accessories in chromate cladding revealits modern technical aesthetics.The combination of a one-piece running-board(1) No front fender design gives people a newimpression of an electric scooter.

(2) The body outline designed with the goldensection proportion yields a scooter with aclassical and elegant appearance.

(3) A rounded and smooth conguration gives thelook of harmony and gracefulness.

(4) The digital luminescent instrument panel and5. Results and discussions

With the determination of the optimal solution,we made a 1/20 scale mock-up of the scooter frompolyurethane foam for previewing the 3D appear-ance. Additionally, the conclusive design wascompleted by constructing and modifying thedetailed contours according to the aestheticprinciple of golden section proportion (Elam,2001, pp. 542). The outcome of the appearancedesign is shown in Fig. 12, and has the following

f Industrial Ergonomics 34 (2004) 421443ing various systematic design methodologies

ARTICLE IN PRESS

nal oS.-W. Hsiao, J.-R. Chou / International Jourwith the SAM. The proposed design processincludes three essential stages: divergence, trans-formation, and convergence, whose essence is notthe sole performance of the stages and thedeveloped creativity tool but the coherent effortsamong each involved technique of the process.

Fig. 13. Illustration of the conclu

Fig. 12. Appearance design according to the aesthf Industrial Ergonomics 34 (2004) 421443 441To illustrate the procedures of implementingand verifying the practicality of the proposeddesign process, a case studyappearance designfor an electric scooterwas conducted step bystep. By applying evolutionary thinking, systema-tic structure analysis, the sensuous association

sive design characteristics.

etic principle of golden section proportion.

ARTICLE IN PRESS

w elec

nal omethod, and other systematic approaches to the

Fig. 14. The prototype of the ne

S.-W. Hsiao, J.-R. Chou / International Jour442design process, not only the problem space wassearched and identied, but also a large number ofsub-solutions involved in the solution space weregenerated, particularly the six sets of fundamentalconcept solutions in outer cover design.By integrating the applicable sub-solutions

among each subsystem, three sets of conceptsolutions for the electric scooter design weredeveloped, and then an optimal solution wasdetermined by using the decision-making modelbased on the weighted generalized mean method.With the determination of the optimal solution,the conclusive design was completed by construct-ing and modifying the detailed contours accordingto the aesthetic principle of golden sectionproportion. A prototype of the electric scooterwas also built.In conclusion, the nal work of the electric

scooter design has met our initial product devel-opment objectives. We hope that the research canprovide a new approach to the design process forindustries, and allow companies to take advantageof the creativity-based design process to achievethe goals of innovative product design.Acknowledgements

tric scooter for this design case.f Industrial Ergonomics 34 (2004) 421443The authors are graceful to the editor andreviewers for their helpful suggestions on revisingthis paper.

References

Archer, L.B., 1971. Technological Innovation: a methodology.

Inforlink, Frimley.

Bullinger, H.J., Lentes, H.P., Scholtz, O.H., 2000. Challenges

and chances for innovative companies in a global informa-

tion society. International Journal of Production Research

38 (7), 14691500.

Chaturvedi, K.J., Rajan, Y.S., 2000. New product develop-

ment: challenges of globalization. International Journal of

Technology Management 19 (7/8), 788805.

Colella, W.G., 2000. Market prospects, design features, and

performance of a fuel cell-powered scooter. Journal of

Power Source 86, 255260.

Cooper, R.G., 1996. Overhauling the new product process.

Industrial Marketing Management 25, 465482.

Cross, N., 1997. Descriptive models of creative design:

application to an example. Design Studies 18 (4), 427440.

Dorst, K., Cross, N., 2001. Creativity in the design process: co-

evolution of problemsolution. Design Studies 22 (5),

425437.

Elam, K., 2001. Geometry of Design: studies in proportion and

composition. Princeton Architectural Press, New York.

French, M.J., 1971. Engineering Design: the conceptual stage.

Heinemann, London.

Geschka, H., 1996. Creativity techniques in Germany. Creativ-

ity and Innovation Management (June), 8792.

Goel, P.S., Singh, N., 1998. Creativity and innovation in

durable product development. Computers & Industrial

Engineering 35 (12), 58.

Goldschmidt, G., 1997. Capturing indeterminism: representation

in the design problem space. Design Studies 18 (4), 441445.

Grifn, A., 1997. PDMA research on new product development

practices: updating trends, and benchmarking best prac-

tices. The Journal of Product Innovation Management 14,

429458.

Hall, A.D., 1968. A Methodology for Systems Engineering, 6th

Edition. Van Nostrand, Princeton, NJ, pp. 8591.

Heap, J., 1989. The Management of Innovation and Design.

Cassell, London.

Hubka, V., 1989. Principles of Engineering Design. Springer,

New York.

Jones, J.C., 1992. Design Methods. Van Nostrand Reinhold,

New York.

Liu, Y.T., 2000. Creativity or novelty? Design Studies 21 (3),

261276.

Luckman, J., 1967. An approach to the management of design.

Operational Research Quarterly 18 (4), 345358.

Lugt, R., 2000. Developing a graphic tool for creative problem

solving in design groups. Design Studies 21 (5), 505522.

Majaro, S., 1988. Managing Ideas for Prot. McGraw-Hill,

Maidenhead.

McAdam, R., McClelland, J., 2002. Sources of new product

ideas and creativity practices in the UK textile industry.

Technovation 22, 113121.

Pahl, G., Beitz, W., 1984. Engineering Design: a Systematic

Approach. Chapter 3: The design process, Design Council,

London.

Poolton, J., Ismail, H.S., Shahidipour, M., 2000. The new

products process: effective knowledge capture and utiliza-

tion. Concurrent Engineering: Research and Applications 8

(2), 133143.

Stevens, G., Burley, J., 1997. 3000 Raw ideas=1 commercial

success. Research Technology Management 40 (3),

1627.

Stevens, G., Burley, J., Divine, R., 1999. Creativity+Business

Discipline=Higher Prots from New Product Develop-

ARTICLE IN PRESS

S.-W. Hsiao, J.-R. Chou / International Journal of Industrial Ergonomics 34 (2004) 421443 443Information. Prentice-Hall International, London.

Krohe, J., 1996. Managing creativity. Across the Board

(September), 1622.

Lai, H.C., Liu, J.S., Lee, D.T., Wang, L.S., 2003. Design

parameters study on the stability and perception of riding

comfort of the electrical motorcycles under rider leaning.

Mechatronics 13, 4976.468.

Tso, C., Chang, S.Y., 2003. A viable niche marketfuel cell

scooters in Taiwan. International Journal of Hydrogen

Energy 28, 757762.

Villa, A., 1998. Product-process design and continuing innova-

tion. Robotics and Computer-Integrated Manufacturing 14,

393401.Klir, G.J., Folger, T.A., 1988. Fuzzy Sets, Uncertainty, and ment. Journal of Product Innovation Management 16, 455

A creativity-based design process for innovative product designRelevance to industryIntroductionTheoretical backgroundThe use of creativity techniquesSensuous association method

Outline of the creativity-based design processImplementation methodsDescription of the target productDivergent stage of the design processEvolutionary thinking of the target productSystematic structure analysis of the products componentsSearching the problem space

Transforming stage of the design processGenerating the solution space

Convergent stage of the design processFinding applicable sub-solutionsIntegration of sub-solutions for developing concept solutionsEvaluation of concept solutions for an optimal solution

Results and discussionsConclusionsAcknowledgementsReferences