a study on the thinking path model

DESIGN RESEARCH AND TECHNOLOGY 217

INTERNATIONAL DESIGN CONFERENCE - DESIGN 2002Dubrovnik, May 14 - 17, 2002.

A STUDY ON THE THINKING PATH MODEL OFCREATIVE DESIGN PROCESS

Hisataka Noguchi and Yukari Nagai

Keywords: Design process, creative thinking, experiment,observation, tacit knowledge

1. IntroductionRecently, researches on the creative thinking have been increasing especially in the field of designscience.In this flow, several researches with experimental studies have been made remarkable results such asGoldschmidt, Purcell and Gero, Dorst and so on [Goldschmidt 1991][Purcell&Gero 1998][Dorst2001]. Before these researches, Finke, Ward and Smith made famous research on creative cognition,in which they had experiments of invention using several geometrical shaped parts [Finke, Ward &Smith 1996]. Those experiments had also been verifications of the “Geneproa model” which washypothetical model of inventory thinking proposed by Finke. In their experiment, the subjects wereobserved in the tasks of combining the parts that resulted some inventive discoveries. Finke, Ward andSmith did not focused on the subject’s drawings. On the other hand, Goldschmidt, Purcell, Gero andDorst focused on the subject's drawings. However, their experiments were put emphasis on theanalysis of subjects’ protocols, and drawings were treated as a reference for the analysis of thinkingprocess.We assert that, in analyzing the design thinking process, to know creative thinking paths in thesearching space is important [Maher, Poon & Boulanger 1996]. However, to do it, the observationshould be put focus on the subject’s drawing process and behaviors rather than the protocols. Because,in design experiment, the protocols often do not reflect the contents of subject’s thinking process. Inthat case, the experimenter should slip into the subject’s mind and should infer his/her thinkingprocess as if the experimenter were dwelled in the subject’s mind. The thinking path of creative designcould be represented as a model by using this standpoint. In following sections, we introduce thisexperimental study.

2. Experiment

2.1 Procedure of the experimentWe think that the most generalized representation of design thinking process is translation of designpurpose (expressed in linguistic mode) to visuo-spatial outcomes of it. Then, we had an experiment toknow the real time thinking process in which designer translate the design purpose (expressed bywords) to drawn sketches as the final ends.The experiment was held at Musasino Art University between June and July of 2001. The subjectswere 4 students of 3rd and 4th year of under graduate design course: 2 students from an engineeringoriented university and 2 students from an art oriented university.The each subject was, at first, assigned to design a “beautiful shaped” (we call this first keyword) tape

DESIGN RESEARCH AND TECHNOLOGY218

dispenser and after while (about 5~7 minutes after the start) he/she was assigned to make it “gives softimage"”(we call this second keyword) as a sub-concept of beautiful shape. Our intention of giving 2staged keywords was that it might make easier to see how designer changes his/her thinking modeduring the design task. The subject was given about 20 minutes to accomplish the assignment.The subject’s behaviors during design task were recorded by two video cameras: one camera recordeddrawing from over head of the subject and the other camera recorded the subject’s performances(including expressions and protocols) in front of the subject.After finishing the design task, the experimenter conducted the subject to speak his/her thinkingprocess under watching play backed video movie (retrospective protocols).

2.2 The results of experimentWe put the observed data together and analyzed each subject’s thinking process as follows.(1) The subject 1: A 4th year student of art oriented university (visual communication design).After started the task, he was sunk in thought about 40 seconds and fingered soft rubber eraser. Thenstarted drawing with thin and weak lines then, in a moment, stopped drawing and erased it. Hefrequently repeated this behavior. After 1 minute and 40 seconds has passed, he started his first sketchwith thin lines in not so quick movement. After 4 minutes have passed, his drawing was graduallychanged to bold lined and put shade. After about 7 minutes have passed, he finished the first drawingand turned page of sketchpad. Then he started perfectly different sketch than the first one.At this moment, the second keyword "soft imaged" was given from the experimenter and he stoppedall movements. After a while, he started drawing and said, "Just I think it resembles to a snail" andturned the page of sketchpad. He moved hand and took aim at the sketchpad with untouched line (wecall this “pre-drawing”). Again, he frequently repeated erasing and redrawing. After 13 minutes havepassed, he started his 3rd sketch. He gradually drew a whale-like shape in frequently tracing lines.After a while, he started to deform the whale-like shape and redrew curved lines with changingproportions of the shape. After fixed the shape, he put shades to his sketch. After 18 minutes and 30seconds have passed, he turned again the page and started to draw his 4th sketch. This time he drewquickly, elaborately and without erasing sketch. During the task, he did not make so many talkingexcept when he explained some metaphors emerged in drawings. In the retrospective protocols, hetook a lot of times to explain his thought on the functions of tape dispenser.(2) The subject 2: A 3rd year student of art oriented university (visual communication design).After started the task, she was sunk in thought a few minutes, and began drawing. After 4 minuteshave passed, she concentrated to drawing. After 5 minutes have passed, she began 2nd drawing andseemed to be in deep thought while drawing. After 7 minutes and 15 seconds have passed, theexperimenter gave her the second keyword "make soft imaged". After a few moments being sunk inthought, she started her 3rd sketch. After about 10 minutes have passed she started to draw her 4thsketch and at 40 seconds have passed she turned the page of sketchpad. Then, she started to drawperfectly different shape but suddenly stopped drawing and discarded it by cross lines. After a fewminutes thinking, she started to draw perfectly different shape, then modified it into a shape like liquiddrops. She concentrated to drawing this sketch until 15 minutes 40 seconds have passed. Then, sheturned the page of sketchpad. Her final drawing was started with "pre-drawings" and drew similarshaped sketch as the 3rd one but in large scale this time. She finished final sketch with elaborate linedrawings and shadings. She was almost very quiet during the task, but in the retrospective protocols,she turned to talkative. She made detailed explains about her drawings shown in video screen.(3) The subject 3: A 4th year student of engineering oriented university (industrial design).As soon as the first keyword was given, she started drawing. At right side in the page of sketchpad,she drew a lined frame, at first, and then began to draw basic shape of existing tape dispenser in it.She drew next to next cartoon like sketches in talkative attitudes with quickly moving hand. After 1and half minutes has passed, she gradually concentrated to drawing and became quiet. After 3 minuteshave passed, she explained her sketches in answering the experimenter's question. During theexplanation she did not stop drawing and sometimes added ideas into the drawing. After 7 minuteshave passed, she became concentrated again after a moment of pre-drawing, and then she starteddrawing at slow speed. After the second keyword, she sunk in thought a moment and made quick pre-


drawing, then immediately started drawing. She made a number of sketches one after another rapidly,and seemed to generate the next idea during the drawing. She often generated new shapes bycombining geometrical forms. At the final sketch, she used the eraser and finished it more elaborately.During the task, she had been talkative and seemed to be avoided stress by talking. Her attitudes werestable through the experiment and kept drawing continuously. As the result, she made the largestnumber of sketches (19) among 4 subjects.In retrospective protocols, she spoke with not so many words concerning metaphors but spoke withwords directly connected to geometrical forms in her 20 minutes talking.(4) The subject 4: A 4th year student of engineering oriented university (industrial design).After first keyword was given, several times, she made questions on the task, and then drew framelines at first. She repeatedly drew thin lines on the frame lines and put shadings with rhythmicallymoved hand. She scarcely talked during the task except several times talked about trivial things of notconcerning to the sketches. While she were in drawing, if the experimenter talked to her, she onlyresponded "mm..." and never turned her eyes from the sketches. Every time before she drew newsketch, she did pre-drawing and then drew frame lines. When she moved to next sketch, she took alittle thinking time, and then started to draw quite different form. She drew nicely varied sketches withbig stroke and quick moving of pencil. As drawing was progressed to some extent, she sometimestried to retouch and put brushing up to her sketches.After second keyword was given, she had a few thinking time, then repeated pre-drawings and drewcurved lines into an integrated shape. She scarcely erased her drawing but often reviewed herprecedent sketches when she moved to draw another sketch. She seemed to be interesting to the task.In the retrospective protocols, she talked a lot on her sketches connecting to some metaphors.

Figure 1. The subjects’ sketches (left to right: subject 1, subject 2, subject 3, and subject 4)

3. Discussions on the method of analysis

3.1 Protocols did not indicate thinking processIn this experiment, we intended to know the designer's real time thinking processes but the subjects’protocols were perhaps not available to know it. The subjects were reticent when they sunk in thoughtand generating image of target object. In contrast, the subjects became talkative when they explainedtheir sketches after finished the tasks. Usually, the protocol analysis put weight on the subject’s talksand explanations. In that case, retrospective protocols were adopted as the most significant clues toknow the creative thinking process. However, the retrospective protocols do not give real timethinking process They give only past explanations of the thinking process as the results of creativethinking. To know how creative idea was generated, the real time trace should be adopted based on thecommon basis of the subjects’ thinking situations. Therefore, we needed another method to know theircreative thinking process as follows.

3.2 Dwell-in observationWhen we analyzed the subject’s creative thinking process, we had to infer it by way of as if we wereslipped into the subject’s mind. We named this way “dwell-in” observation depending on the MichaelPolanyi's words. Polanyi said that to understand tacit dimensions of one’s knowledge, we needed toput our thought into one’s mind as if we dwelled in it [Polanyi 1966]. To do this, the experimenterneeds to have a common ground with the subject’s situation. In this experiment, fortunately, theexperimenters have experiences of design works, so that they could understand the subject’s thinkingprocesses with the “dwell-in” observation. To do the “dwell-in” observation, the experimenter shouldhave similar experiences of the task, which is given to the subject.


4. The analyzed resultsOn the basis of the “dwell-in” observations, we analyzed the subjects’ creative thinking processes asfollows.

4.1 Two parts in the description of design goalWe inferred, as mentioned in another presentation, that designer usually recognizes the description ofdesign goal as two parts: for example in case of design goal described as “Design modern chair”,“chair” is recognized as the subjective word and “modern” is recognized as predicative word. Then,designer thinks separately of the subjective part and the predicative part of design goal.In this experiment, we observed the subject’s thinking process with keeping this thought in minds.The design goal of this experiment was “Design beautiful shaped (and gives soft image) tapedispenser”. The subjective part was “tape dispenser”, and the predicative part was “beautiful shaped”and then “gives soft image”.From the observation of the subjects’ drawing processes, all the subject of this experiment wereseemed to think separately the subjective part and the predicative part of the description of designgoal.

4.2 Searching modes observed in the experimentThrough whole tasks, we observed that the subject changed his/her thinking mode in several times.In another presentation, we pointed out that there were two kinds of thinking modes: F (Form making)mode and M (Metaphor) mode. However, we found, in this experiment, that both thinking modescould be decomposed into more numbers of searching modes. Just after the first keyword was given,some of the subjects confirmed the subjective part (tape dispenser) of the goal description by drawingtypical model of it. Then in next step, he/she searched metaphors by which he/she could interpret thefirst keyword (predicate part of the goal description). We can see it in the subject’s protocols. We callthis thinking process as “interpretative search”. In case of the subject 2, she searched things connectedwith the keyword “make soft imaged” in her memory. We call this “recollective search”. After this,the most of the subjects tried to decompose the parts of the prototype and recombined or modifiedthem into a fitted form to the keywords (beautiful shaped and makes soft imaged) under consideringthat it could perform the function of tape dispenser. We call this process as “semantic generationsearch”, because, translating keywords into image was the process of generating meanings. In case ofthe subject 3, she tried to make forms of tape dispenser by repeated recombination of the geometricforms. We call this “combination search”. At the latter process of the design task, most of the subjectstried to figure out their internal image of designed object onto the sketchpad. We call this process as“depictive search”. At the final stage, most of the subjects tried to refine and put shade on theirdrawings. This final process was not a searching process but the process of making up reality ofdrawings.

4.3 Pre-drawings as an important searching modeWe took notice on the pre-drawing actions of the subjects.The subject often moved his/her hand as if he/she was drawing something but did not touch pencil onthe paper. We named this action as “pre-drawing”. Almost every time when the pre-drawing wasemerged, the subject became reticent and seemed to be sunk into deep thought, and then he/she startedto draw new shapes with thin lines but soon erased it. This action was repeated until the new form wasfixed on paper.We thought that when the pre-drawing was emerged, the subject was generating his/her new image oftarget object in mind but it was not fixed yet as a figure of being able to draw with clear outlines.Therefore, we believed that the pre-drawing was the crucial key to know the creative thinking indesign process. While doing the pre-drawing, the subject must have been searching clear image of,which was interpreted from the description of design goal. Although an image was in his/her mind, itwas not so easy to decide the figure of the image. Then he/she repeated the pre-drawing and thin lineddrawing with frequent erasing. He/She had no confidence to draw by steady and bold outlines. We call


this behavior of the subject as “pre-drawing search”. The pre-drawing search might be included in thedepiction search, but it was important that the designer was searching a figure fitted to the design goalby moving his/her hand as if he/she were in drawing actions. It implied that the physical moving wasneeded to designer’s creative thinking process before getting visible form of design object.

4.4 Thinking path model as the trace map of thinking processAs the results, we tried to represent the subjects’ thinking processes as diagrams of thinking paths.To do this, thinking process must be decomposed into several behavioral units. We extracted thesubject’s behavioral unit not from observed outward appearances of them but from inferred processesof their inner sides.We adopted the standpoint that we thought as if we were doing the same task as the subjects, andtraced their thinking processes as possible as to have reality.As the result, we presented a thinking path model of creative design process as shown in figure 2.

Connectsubject andpredicate

Predicative part

Grasp basic typeof design object

Search metaphor ofpredicative part

Consider balanceof function andform

Expressionof reality

Final form ofdesign object

Try expressionof unfixed image

Expressionof fixedimage

Thinking path of creative design process

Extract andReconstruct basiccomponents

Bring metaphorto image

Description ofdesign goal

Subjective part

Stage of conceptual drawings Stage of form making drawings

High order semantic generation process

Recollective search + Combination search

Recollective search + Interpretative serach

Semantic generation search + Depictive search

( Pre-drawing search )

Finish drawing

Figure 2. Thinking path model of creative design process

5. Conclusion and perspectives

5.1 Creative thinking needs long searching pathFrom this thinking path model, we found that creative thinking needs long searching path. The reasonof this is as follows.In this experiment, the subject 4 seemed to take the most complicated thinking path and then shebrought the best results of design. The times taken in the thinking process are not related to thedistance of thinking path. For example, the subject 1 took long times in the stage of search of semanticgeneration, but his thinking path was not so complicated. The numbers of idea sketches do not relatedto the distance of thinking path. For example, the subject 3 generated the largest numbers of sketches,but her thinking path was simple. In case of the subject 2, she took many times in the stage ofinterpreting search, but her thinking path was also simple and the result was not so good in balance offunction and form.The distance of thinking path is related to the times of changing search modes and the numbers ofsearching modes used in the thinking process. We concluded that the designers made efforts to take asmany searching modes as he/she could for getting creative thinking. The creative result cannot be gotwithout taking effort to search the best clue of creation. In creative thinking process, designer musttake at least 3 stages of searching modes: interpretative search, semantic generation search anddepictive search. The best result can be got when the designer take adequate searching modes inadequate times of dynamic situations of thinking process.


After this experiment and analysis, many things are left for future researches. The most importantthing is the method of analysis. To make more objectify the dwell-in observation, we should have asystematic description tool of nonverbal behaviors in creative thinking process. To confirm therelations between searching modes and creativity, we might need much more experiments. To clarifythe inner process of pre-drawing, several results of recent brain science might be significant. Even ifall things mentioned above might be done nicely, still the approach to clarify the mechanisms ofhuman creative thinking would need long way.

ReferencesDorst, K. and Cross, N.: Creativity in the design process; co-evolution of problem-solution, Design Studies,22(5), 425-438, 2001.Finke, R.A., Ward, T.B. and Smith, S.M.: Creative Cognition, The MIT Press, 1996.Goldschmidt, G.: The dialectics of sketching, Design Studies, 4, 123-143, 1991.Maher, M. L., Poon, J. and Boulanger, S.: Formalizing designs exploration as co-evolution: a combined geneapproach, Advances in formal design methods for CAD, Chapman and Hall (UK), 1996.Polanyi, M.: The Tacit Dimension, Routledge & Kegan Paul Ltd., 1966.Purcell, A.T. and Gero, J.S.: Drawings and the design process, Design Studies, 19,4, 389-430, 1998.

Hisataka Noguchi, Prof. of Design FundamentalsDepartment of Design, Faculty of EngineeringChiba University33, Yayoi-cho, Inage-ku, Chiba, Japan 263-8522E-mail: [email protected]/Fax: +81 43 290 3100



ETHICS IN INDUSTRIAL PRODUCT DESIGN(GOOD, GOODS AND GODS)

A. Can Özcan

Keywords: ethics

1. PrologueThe lack of ethical side and its possible reasons in our professional designer lives is the main theme ofthis presentation. For the basic principle is good for ethical existence, we usually avoid asking ethicalquestions in our professional design activities. We prefer principles of professionalism or codes ofconduct. Though the whole design culture of past, present and future have been and will beabstractions of human mind, yet we find ethics too abstract for design practice. We seem to preferdesigning for the sake of design’s own sake and never asking ethical questions like “is it good?” Ofcourse there comes a time for all of us to ask ethical questions and it usually happens when the goinggets tough. When somebody reproduces a sheep genetically, when a couple of designer-scientists getclose to apply human DNA into computer processors, when we start watching planes crashing intotowers just like a movie or a war just like a computer software, or when we see our children gettingmore violent in front of actual or fictional terror in our designed environment, then we start askingethical questions as it is nowadays.I have always been fascinated by the fact that the term “good” has been used for also utilitarianproducts. Though I could have preferred many of the designed objects never gone into productionthere was a spiritual side in calling things not only as things or objects, or products but also as“goods”. May the reason be lying in our ever forgotten design consciousness that designing meansbringing good things to life? Rather than introducing it as an ethical principal I would like to presentthe word good as a practical term in relation to design profession and its special kind of outcomescalled as “goods”.The most distinctive characteristics of goods are their anonymous design facilities. When we saygoods we mean anonymous, almost illegal, nameless, faceless design entities. We never name Starckdesigns or an Alesi kettle, or a Colani device as goods. Goods generally signify a negative category inour professional, or academical, or artistical design terminology. None of ourselves want to be knownas designing “goods”. We want to be product designers or a brand on products if we can. We do notcare about goods. They appear as non-qualified objects that we use in our non-conscious (or at leastsemi-conscious) everyday life.Giving birth to a baby and designing in industrial (or psychologically we may prefer post-industrial inthis new millennium) context are too much alike. Whilst only two are enough for the former action, acrowded mass of people participate in the latter one. Only two people (of different sexes of course)can start a life which seems to be the most miraculous event in the known limits of the universe. Onthe other hand for realizing a product, a group of authorized and qualified experts in their fields cometogether where it is not easy to distinguish who is doing what and to who. Engineers, planners,designers, medical experts, psychologists, sociologists, ecologists, technicians, administrators, marketresearchers, pr experts, and not only individuals but also institutions and associations join the designand production orgy, users and clients are no exception. This overintercourse ends with a baby called


product and designers seem to be the most insistent group of people that they are the real parents ofthe baby and they have the right to give it a name. Usually designers do not have any clear idea aboutthemselves whether they’re the “fathers” or “mothers”. In fact it is not possible to predict the nature ofeither the baby or the product from the way it has been created. Bad seeds may come out of lovemarriages, and fine characters may come out of prostitution in whatever the century is. And design is asignificant and unique concept that we can analyze the quality of our culture for its cooperativecharacter. I always tell my students that our designs reflect the character of our cooperativeconsciousness and mind. If anything goes wrong with our designs there must be a problem in ourminds.Two important trends under different names seem to be shaping the world of today and tomorrowwhich also bring out new concepts like “co-designing” to an ideal state of our practice. One is“globalization” while the other is “specialization”. Though I have definitely a positive approach to amulti-disciplinary design comprehension I do not share the same optimism for these two concepts andI remember a post-it sticker “trend is not destiny”.

2. Ethics of GlobalizationOne of the popular terms of our day is globalization. My personal belief is that if a word is beingpronounced too often there must be something wrong with it. Take AIDS for example. If we had acure for it we could have never talked about it that much. An AIDS with no cure creates a self-maintaining system of its own and a cure appears to be a thread for this existence, just like lack ofenemy makes military useless and meaningless. Though it means different things to different peoplethe term globalization is a little bit like AIDS I think. It is not a cure but a kind of disease for what itrepresents.Globalization does create two different designer characteristics depending on two differentunderstanding of globalization. According to Victor Margolin the term globalization represents anequilibrium model of the world to some people while an expansion model to the majority of its actors.The expansion model of the world under the term globalization moves designers work for marketsrather than people. Designers see themselves as strategic businessmen. The designed objects forget tobe goods and turn out to be “tokens of economic exchange” [Margolin 1998]. A brand car designed inItaly gets assembled in Turkey while the components arrive from the rest of the world. The word co-designing finds a practical area of use in this model and design becomes a manipulative term ofeconomic cost/benefit arguments and the concept of design management overshadows design as aresult. According to John Heskett , rather than designing goods, designers function as follows in thisglobal business:

• generating new product concepts• customer focus• speed to market• ease of manufacture• reducing product costs• reducing process costs• differentiating products• adding value to products• extending product life-cycles• innovation – opening new markets [Heskett 1998]

The equilibrium model of the world in globalization discourse appears as a counter ideal opposing thebasic dilemma in expansion model. In fact this paradoxial dilemma has been depicted in differentformats against industrial capitalism. The survival of industrial or global capitalism depends oncontinuous and ever increasing production and consumption without any limits which sounds like theexpansion model we have discussed. But as it has been expressed in many formats, “resources” arenot limitless in human reach. Equilibrium model depicts a world order which depends on the balanceof limited factors: resources, needs, population, environment, etc.. The equilibrium model representsthe awakening of subconscious conscience under serious threats. Widening gaps between economical


and social parties, destruction of nature, increasing social and psychological disabilities reallythreatens the whole system in the world. But the equilibrium model does not have the courage to spellthe underlying truth that it is the expansion model that creates the threats as well as the increasingwealth and status of designers under its wings. Designers trying hard to find solutions under termslike sustainability, or sustainable product development are the victims of equilibrium state of mindunder expansion and limitless economy. It is like fighting for peace, making love for virginity.One of the apparent efffects of globalism on design profession can be observed in the unification ofbig companies, that I call re-creation of gods. This unification can be in the form of happy marriagesor one company devouring another. Bosses, strategies, target markets, design understandings, brandcharacters change suddenly. A Scandinavian Volvo turns out to be a Ford-ian Volvo, a post-Volkswagen Skoda (e.g., Octavia model) is never the Skoda we have known for decades. Basicevolutionary principles of wild nature gets applied to the whole world: Those who are fit survives andevolves, those who are not become extinct (not virtually, but actually). The same rule applies not onlyto companies but also even to countries and even geographies. It reminds me of a softer racism indisguise. "If you are poor, crowded and technologically underdeveloped, then die. Because I will neveruse my wealth and technology to solve your problem, rather I will see you as a problem for me to getwealthier, and technologically much more developed.." [Özcan 1997]. Even the standarts beingemposed to design profession (ISO, ASA, DIN etc.,) in the majority of the world do not improvedesign ability but put an end to available and sufficient productivity and industrial design life.Designers are being forced to become not good designers but powerfully equipped strong competitorsas if in a wild jungle where a slight misjudgement turns out to be tragedy in chain reaction wheremeasure of everything is numerical. What about the design objects i.e., the goods?.. Whocares?...Production/Consumption Rates and Purchase Figures tell it all. We have stopped thinkingabout quality since the end of history [Fukuyama 1989].Robert Maynard Pirsig has spent most of his life thinking about quality without reducing it intoquantitative figures. He ends up explaining the evolution chain without quantitative terms. Inorganicquality, biological quality, social quality and intellectual quality are his terms to define theevolutionary process which can only be explained as development [Pirsig 1991]. But the last stage ofthe evolution, the intellect, is now threatining its pro-stages it can only exist on, i.e., the nature, thebiological human being and the society . In a personal letter, he has told me his appreciation that I’vebeen using his books in design courses. What he mentioned in that letter was also the fact that ...manypeople who talk about quality in Industrial Design define quality in terms of “serving an intendedpurpose”, but of course that says nothing about the quality of the purpose or whether there is qualityto be found outside the purpose. Goodness is a purpose and how can you figure out good in numbers,statistics, unless you do not have that basic sense of quality, unless you do not see the world ashardware and globalization as the software running on numerical codes but can only work well by thesacrifice of abstract, spiritual, ethical, aesthetical, logical, active and alive side of human side i.e., thesource of designing goods as well?

3. Ethics of SpecializationR. Buckminster Fuller warns us by giving two independent studies one from biology on the extinctionof species and one from anthropology on the extinction of human tribes, concluding that specializationis a way to extinction, and our society is thus organized [Papanek 1972]. As the world situationcontinues to inbreed specialization as it is now, general adaptability which is the basic advantage ofhuman beings will be lost.Human being refers to a general natural entity, a biological term and we rarely use this term forourselves. In social context we call ourselves as citizens, people, community, humanity, society,culture... For us the designers, we are classified as specialists in this world scale mechanism ofinteractions: producers,consumers, users, etc. The era we are living in can also be described as the eraof not professions but also specialists. “Who are you?”, “ I am a digital interface designer I am atheatre architect... I am a gynocologist... I am a divorce attorney”. They all also mean that I am adesigner designing for you and it’s my way of existence not everybodys’ or anybodys’ to design thegoods you use for your own sake, or, I am an architect who has taken the licence and knowledge of


building your own living spaces out of your hands and mind, or, I am the doctor and you do not knowwhat to do when something happens in your body, or, I am the one who can defend you and yourrights better than you, and we all need another specialist to change the taps in our bathrooms. Thisera, in contradiction with the fact that over-specialization leads to extinction and generalization is theadvantage of human beings, shows specialization as the only way of existence, but it starts to create itscontradictions. For everybody is a specialist in something, it becomes impossible to realize the wholeas in the case of designing products, especially when everything is bound in global scale with writtenand non-written rules.I will always ask the question if being a designer is a professional status or more or less a humaneability.It’s no surprise that the Rolling Stones song Satisfaction has been awarded by the authorities as thesong of the last century. The lyrics “I can’t get no satisfaction” not only defines the beat of our lifestyles, but also motive the designers to fulfill this sense of dissatisfaction while keeping and expandingit. If we loose the perception of our lives as a whole we can get no satisfaction out of it, but try tosubstitute it with a consumerism out of control. My people in Turkey can not help changing theirmobile phones frequently, a five years old car is an old car, last year’s model is out of trend, and thesong of the century says “ I can’t get no, satisfaction!”. Satisfaction of a need with a design shouldbring material satisfaction along with spiritual satisfaction. But any satisfaction appears as a threat forthe system being executed all over the world.It’s been a couple of years that I’ve been studying bicycles as a professional designer. Though themain design features of a bicycle have slightly changed for about 150 years It’s getting harder andharder for bicycle manufacturers and designers each year to design another model for it’s turning outto be an iron bound business and inevitable dependance on experts of any detailing on more than 300years old bicycle. This company has been specialized on front fork systems while that company is anexpert on rear suspension. You find the best chain wheels and cranks in Taiwan while the fittestsaddles have been designed in Italy. You find the same brand on the majority of bicycle gearcomponents and I can count not more than five professional manufacturers for wheels. It would bepossible to be a bicycle designer twenty years ago, but you can only be a saddle designer and a bicycleco-designer nowadays.Now think about the satisfaction of an independent designer who is witnessing the realization of hisdesign turning into a physical entity under total control on every detailing, and think about the otherone feeling like a compulsory contributor into an inevitable out of control activity. What’s good for adesigner?.. Freedom or dependance? I am sure that a bicycle getting assembled at the garage is muchmore satisfactory than depending on international trade and production unions. Same for the users:It’s being estimated that there will be justa a few companies producing automobiles for the whole. Itmeans that there will be less designers and design activity for more and more consumers in the nearfuture. And for designing will be a professional activity, one of the main features of human brain,designing through abstraction will no longer be a basic and general humane activity.

4. EpilogueAs Victor Papanek puts it as the first sentences of its first chapter of his first book, all men aredesigners and all we do almost all the time is design, for design is basic to all human activity [Papanek1972]. Victor Margolin classifies the active engagement of human beings with design in four basicways [Margolin 1995]:

1. they design products for others2. they design products for themselves3. they use products designed by others; and4. they use products they design for themselves

Years have been passed since this classification and it’s still but a diminishing hope to consider designactivity as a practice of human beings. Globalization and specialization hand in hand are trying hard toconquer and take it away this basic human ability from inside, with the help of profeesional designers.The basic principle for ethical abstraction lies in the term "good". If we do not have a sense of what isgood and what is not then it means that we've lost our ethical abstraction ability. You can not replace


the qualitative answer for good and not good with quantitative factors. Ethical human existence cannot get satisfied with quantitative answers. We must have a basic ethical abstracion for our designpractices. We must ask ourselves if it is good to depend on nuclear energy, and must find an answerapart from quantitative reasoning. We need ethical answers for questions like "is it good that thetechnology I create can make a majority suffer in favour of a minority's wealth?" or "is it good thatthe technology I am creating is making a lot of human being useless, or just giving them a sense ofuselessness? or, "is it good that I lost all my communication with my neighbour but I cancommunicate with the other side of the world through my computer?", or "is it good that while mykitchen is getting technologically well equipped my food is getting tasteless, and I am getting fatter?"I always think that semi-god Prometheus was a designer and the fire he brought to humans was designability. Now he is embarrassed and bringing the fire back to gods with a job application in his hands.By sacrificing the human side he thinks that he will be accepted among gods.Can I be considered to be pessimistic about design activity today. No. I enjoy practicing and teachingindustrial design.

ReferencesFukuyama, Francis., The End of History, The National Interest, 16, 3-18, 1989Heskett, John., The Economic Role of Industrial Design, In: Balcıoğlu, Tevfik (ed) The Role of Product Designin Post-Industrial Society, Middle East Technical University Press, METU-Kent Institute, Ankara, 1998, 77-92Hoffman, K., "Technological Change in Telecommunications; Implications for Industrial Policy in DevelopingCountries", New Technologies and Global Industrialization Prospects for Developing Countries, PPD, 142, 13November 1989Illich, Ivan., Tools for Conviviality, Fontana/Collins, England, 1975Margolin, Victor., The Product Milieu and Social Action, In: Buchanan Richard, Margolin Victor (eds.)Discovering Design, The University of Chicago Press, Chicago and London, 1995, 121-146Margolin, Victor., Design and The World Situation,. In: Balcıoğlu, Tevfik (ed) The Role of Product Design inPost-Industrial Society, Middle East Technical University Press, METU-Kent Institute, Ankara, 1998, 15-34Özcan, A. Can., Progress Depends on Unbalance and Disorder, ISTAS 97, International Symposium onTechnology and Society, Institution of Electrical Engineers, University of Strathclyde in Glasgow, Scotland,1997Papanek, Victor., Design For The Real World, Granada Publ., New York, 1972Papanek, Victor ., The Green Imperative, Thames and Hudson, 1995Pirsig Robert M., Zen and The Art of Motorcycle Maintenance, An Inquiry into Values, Vintage, 1972Pirsig Robert M., Lila, Bantam Books, USA and Canada, 1991Rams, Dieter, The Responsibility of Design in The Future, In: Balcıoğlu, Tevfik (ed) The Role of Product Designin Post-Industrial Society, Middle East Technical University Press, METU-Kent Institute, Ankara, 1998, 15-34

Can Özcan,PhD., Assistant Professor of Industrial Product DesignIzmir Institute of Technology, Department of Industrial DesignİYTE Kampusu, Gulbahce mevkiiUrla 35430, Izmir,TurkeyTel: + 90 232 4987066, Fax: + 90 232 4987012Email: [email protected]



TOWARDS A FLEXIBLE AND ADEQUATE USE OFMETHODS IN PRODUCT DEVELOPMENT

U. Pulm and U. Lindemann

Keywords: Design research, design theory, design science, designmethodology

1. IntroductionDesign methodology has become a major part of research in the field of engineering and offers a largevariety of methods in order to establish effective and efficient product development processes [Ehrlen-spiel 1995]. The main problem yet is that methods are barely used in practice or used in a wrong way.Corresponding to observations in industry, a survey resulted that there is no continuous use of methodsin practice [Grabowski & Geiger 1997]. One of the main reasons for the insufficient use of methodsmay be that there is a wrong understanding of methods and their meaning, which again may bedetermined by a wrong design of methods. With this contribution we would like to present a moreappropriate view on methods and the consequences for their design, teaching, and industrial use, notjust to save design methodology in itself but to optimise development processes by effective tools andstrategies. In this context, a method is understood as any tool, strategy, or proceeding that supports thesolution of a problem or the achievement of an aim in general.

2. Current situation in design methodology

2.1 Reasons for the inadequate use of methodsThe main reason for the insufficient use of methods seems to be that there is a wrong understanding ofmethods and their meaning, and by that wrong expectations. Often enough, designers think that thereis or should be a method that directly leads to the desired result. If the desired result fails to appear, itwill lead to disappointment, to a general questioning of methods and the non using of them. Theresults will not appear, if a method is understood as some kind of development process algorithm orautomation. This in fact cannot be due to the necessary creativity and innovation, which somehowdefine design. By this, methods are used wrongly, both strategically or efficiently, i. e. which methodto use, and operationally or effectively, i. e. how to use it.The wrong understanding of methods may be determined by a wrong design and description. It isoften discussed if methods prescribe a special procedure or describe a recommended or logical way ofproblem solving, while actually these differentiation between description and prescription must not beseen to narrow. The intention of describing phenomena in engineering design implies some kind ofprescription by showing best – or worse – practices. Vice versa, the way to prescribe processes,methods, etc. is done by the description of an exemplary proceeding. To understand this is importantfor understanding methods and their description itself.Another reason for the wrong understanding is that methods claim to be complete and generally valid.On one hand, these claims may be necessary in order to propagate the method and to gain a broaderacceptance. One the other hand, these claims lead to the fact that designers expect completeness andgeneral validity, which, again, cannot be fulfilled and leads to the above mentioned disappointment.


By the way, completeness means that each aspect of the regarded topic is comprised within themethod; this is strongly connected to the general validity and signifies the above criticised because notpossible automation of design processes. It somehow also negates other, contradictory methods, whichmay be also valid by regarding the problem from another point of view.In the same way completeness and generality are delusive, the negligence of the necessity of technicalknowledge is precarious. Methods do not claim to overcome technical knowledge, but they often lackto explicitly refer to it. Methodology is not a substitute for specialised or technical knowledge nor analternative way of problem solving. It is the systematic general way to solve a problem that shall leadtechnical knowledge in the right direction. In this sense, the aim of a method is not primary the finalresult, but a structured and systematic proceeding to get there. When asked for methodical support,this is not possible without the integration of technical experts. Methodical and technical experts eitherhave to work together, or there has to be a comprehensive methodical education.Though most methods somehow attend to handle complexity, the problem of complexity itself isbarely regarded. Methods may be the theoretical right way to handle complexity, e. g. by showing allrelevant relations between the elements in a recommended representation. But how to regard reallycomplex systems, e. g. with myriads of elements and relations, is not described, or the method evenfails completely by handling them.Talking of complexity, the amount of methods and problem solving approaches itself has become aproblem, since there is no standard method or solution for a specific problem. Here, methods and toolsin order to estimate the efforts of a method and to choose the adequate one are still missing. "Methodsto estimate methods" shows the redundancy of the topic and by that what methodology means: it ismainly the self-reflection of one's acting. This means that there is always a superior level that allows amore general, abstract regard of the problem. So e. g. there is the description of technical knowledge,then the description of methods to handle this knowledge, and then the description of the flexibleadaptation of methods or their implementation. It is one aspect of the incompleteness or aninterpretation of the respective theorem, that this reflection will not end and by that no (meta) methodcan be complete. This is another reason why there has to be unconscious and intuitive elements withindesign processes [Wach 1994], either integrated in a general reasonable proceeding or not.These and other reasons for an insufficient use of methods [Zanker 2000] can be rephrased as thefollowing requirements on methods:

• concrete and specific formulation, not theoretical and complicated, practical• user friendly and up to date• no prescriptive character, regarding individual working stylea• regarding distributed development processes explicitly• regard flexible use in their description and offer methodical concept for their adaptation• clear connection and integration of methods, superior structure, continuous description• implementation strategies, as well as education, training, and coaching• clear target and way it works• reasonable and detailed classification of methods for a proper choice• reduced efforts for the use of methods with obvious benefits• no pressure to use methods, understanding their meaning.

Regarding projects in collaboration with industry and universities, it often seems that industry cannotuse the methods research has to offer while they have problems research cannot solve. Thisdiscrepancy may be on one hand justified in research's objective to think far ahead. On the other hand,it somehow shows the main problem of industry, that the will to solve a problem is more importantthan the ability. A method is not the solution, it is just a tool.

2.2 First approaches for a better use of methods and the positioning of design methodologyThere are different approaches concerning an abstract regard of design methodology. First, there is thesituative adaptation of methods. It bases on the model that methods consist of elementary methodssupporting elementary activities of the design process and that their character is specified according tothe boundary conditions, e. g. team or individual. The elementary activities are e. g. collecting, struc-


turing, separating, comparing, creating, or deciding. A complex method can be assigned to these ele-mentary activities, elementary methods directly support these activities. Elementary methods are e. g.matrices, portfolios, hierarchical trees, or flow-charts. Another basis of this approach is theclassification of methods concerning the criteria structure (used elements and classes like work steps,priorities, hierarchies, etc.), formulation (principle, dialog, rule, etc.), representation (language,diagrams, symbols), tools (hardware, software), media (paper, computer), and user group (individual,team). These criteria can be regarded directly, and as superior criteria, where the mentionedspecifications are the actual criteria that have to be specified, e. g. individual can be a manager, adesigner, or a controller. The situative adaptation of methods now consists of the approaches: use ofelementary methods for elementary activities, adaptation of methods on boundary conditions, andrecombination of methods in large scale. The choice between these depends on the task's complexity,its importance and the available capacities.Another pragmatic approach is the method implementation in integrated product development [Stetter2000]. The described proceeding consists of the steps initiation of a method implementation process(identification of strengths and improvement potential, establishment of objectives, moderation andteam building, etc.), analysis of the product development system (collection of information, intensiveanalyses of details, etc.), choice and adaptation of methods (choice of the appropriate method, adap-tation of methods to individual and group prerequisites, etc.), the actual implementation of methods(method teaching, coaching, etc.), and the evaluation of the impact (quantitatively and qualitatively).There are, of course, approaches to define a generally valid, comprehensive methodology [Grabowski,Rude & Grein 1998]. Such a design theory may serve as a discussion platform or a portal, but it cannotbe complete or even consistent, i. e. free of contradictories. This is due to the fact that each process isdifferent ant cannot be totally decomposed. This chaos or flexibility within the design processes isnecessary for continuous improvement and creativity or innovation at all. It may be also what has beendescribed as the discourse within the design process or the action orientation in design methodology[Lindemann & Wulf 2001]. This means that new solutions are generated within a dialog – an innerdialog of one person or a dialog between a few individuals – where analysis and solution are notstrictly separated and boundary conditions always change arbitrarily. There also has to be analternation of systematical (level of results) and associative (level of action) procedures. Though thereare both action oriented methods such as brainstorming or synetics (for teams), TRIZ and mindmaps(for individuals), and description and documentation oriented methods such as functional structuring,morphological matrices, evaluation methods, etc., it seems that design methodology focuses onplanning, directing, and controlling instead of thinking and acting.For a clear understanding of design methodology, there shall be a description of its positioning withinproduct development as shown in Figure 1. Objective of each development is the product, placed hereon top and represented by different product models. Each model, requirement specification, functionalstructure, CAD-data, etc., is a representation of the complete product, though of course not regardingeach aspect. The product is result and outcome of the development process, which is very fuzzy on theabove mentioned action oriented level, but can be described as a reference on the abstract level ofresults by building blocks, tasks, milestones, etc. These processes run within specific resources andboundary conditions represented in the bottom and influencing one another. These are the organizationof the enterprise, its culture and markets, its employees with their characteristics, money, time, as wellas methods and tools. These methods and tools can support single process steps concerning differentlevels of the product representation. They cannot completely describe the whole process and do notwork without appropriate boundary conditions such as the company organization, experienced,qualified, and motivated employees, and respective capacities [Pahl & Beitz 1996]. The results of theprocess or single process steps are again the input of new processes or process steps and also influencethe boundary conditions and resources. The borders within this model are of course fuzzy, e. g. theproduct representations are again steps or milestones of the process while the abstractly describedprocess can be understood as a resource of the enterprise. This model may give an impression of whatis covered by design methodology and shall help to understand the aim and integration of methods.The emphasis is on a flexible description of methods for different design processes and methodologyknow-how regarding employees as the companies' most important resource.


ProductProduct

ProductModels

ProcessModels

Process

• industrialengineering

• companyorganization

• industrialengineering

• companyorganization

• enterprise• product type• culture• market

• enterprise• product type• culture• market

• employeesqualificationetc.

• employeesqualificationetc.

• methods• tools• media

• methods• tools• media

Boundary Conditions & Resources

Process steps &building blocks,Actions, Tasks

Process steps &building blocks,Actions, Tasks

OutputOutputInputInput

supported bymethods etc.

different productrepresentations

• money • time • etc.

action

results

Figure 1. Model of product development

3. Further implications and conclusions on methods of product development

3.1 Elementary methods and their integrationTo have a methodology that is both continuous throughout the design process and manageable withreasonable efforts, it is recommended to develop more elementary methods on one hand, and on theother hand show their connection and integration in the overall concept. A lot of methods seem to beself-contained within the design process, such as benchmarks, requirement lists, functional structures,etc. Of course they are located within the design process, but the connection of the single methods andtheir interaction are not formulated clear enough. Furthermore, depending on the scope of the method,there are many redundancies, overlaps, and even contradictions within the methods, e. g. comparingvalue analysis and the problem solving cycle, or e. g. matrices within the design structure matrix orwithin quality function deployment. There are also many comprehensive methods or tools that areonly described as a whole – barely manageable – unit. Objective for design methodology as well aseach single method of any complexity is to offer single, manageable methods comparable to theelementary methods described above but in a more concrete form (matrices, tables, diagrams, proce-dures, etc., together with their dimensions, parameters, characteristics, etc.). All these clear methodshave to be integrated in an overall context, i. e. it is to show how they work together and can be putinto a workflow. This may be possible by explicitly described connections, but not only input andoutput. It shall not mean that there is one clear described design process, but it shall bring transparencyinto design methodology and enable to find a reasonable way in development processes. Or in otherwords, this shall not overcome single-standing methods nor complex methodologies, but it shallemphasize and demand to always regard both aspects – manageable methods and the overall context.

3.2 Constituents of a method or a toolNext to the differentiation of elementary methods and their integration stands the aspect of theconstituents of any method (Figure 2). According to this, a comprehensive method shall have adescribed process containing the respective work steps in their right arrangement, the language withwhich it describes the product or the regarded aspects of the systems (graphical symbols such as inUML or functional structures, matrices in QFD, etc.), tools that can be of any kind, either based onpaper or on computers, and finally rules both valid for the method and valid for the system regardedwith that method. The borders between these constituents may be sometimes blurred, but this modelshows the aspects that are to regard by building up a method.


Customerwishes

Deg

rees

of

Fre

ed

om

Ra

ng

eo

fV

alu

es

Rea

liza

tio

n

Cri

tica

lP

rop

erti

es

Pro

du

ctP

rop

ert

ies

Ev

alu

atio

n

Re

leva

nce

for

Cu

sto

mer

Criteria

Kind ofDegree Process

Kind ofRealization

Parts/Assemblies

CriticalUnits

Adaptation

Need forEvaluation

Correlations

Correlations

Correlations

aa

ee

bb cc

dd ff gghh ii

kk

ll mmjj

nn

oo

pp

Customerwishes

Deg

rees

of

Fre

ed

om

Ran

ge

of

Va

lue

s

Rea

liza

tio

n

Cri

tica

lP

rop

ert

ies

Pro

du

ctP

rop

ert

ies

Ev

alu

ati

on

Rel

eva

nce

for

Cu

sto

mer

Criteria

Kind ofDegree Process

Kind ofRealization

Parts/Assemblies

CriticalUnits

Adaptation

Need forEvaluation

Correlations

Correlations

Correlations

aa

ee

bb cc

dd ff gghh ii

kk

ll mmjj

nn

oo

pp

Baugruppe ABaugruppe A

Freiheitsgrad 17Freiheitsgrad 17

Real.: ModulReal.: Modul

xyz

xyz

43,5742

43,5742


Freiheitsgrad 17Freiheitsgrad 17


xyz

xyz

43,5742

43,5742

BB

XXXX

AAAA

FFFF


Freihei tsgrad 17Freiheitsgrad 17


xyz

xyz

43,5742

43,5742


Freihei tsgrad 17Freiheitsgrad 17


xyz

xyz

43,5742

43,5742

BB

XXXX

AAAA

FFFF

ProcessProcess

RulesRulesI. 1

I. 2

I. 3

I. 4

I. 5

Te i l a ufg a b e n (Pr o d uk t u n d Pro ze ß ) fo rm u l ie re n

u n d se q u en ti e l l o de r p a ra l l e l a ba r be i ten .

Z u er st d a s W ic h ti gs te (v o m z u m

, v om zu m ).

Vo m z um .

Vo m zu m .

Z u er st d a s D r i n g l ic h st e .

W E SEN TLI C H EN

W EN IG E R W ES EN TL IC H EN G R O B EN F EIN EN

RIS IKO RE IC H E N RISIK O A R M EN

C H AN C EN REI C H E N C H A N C EN A RM E N

LanguageLanguage

ToolsTools

Figure 2. Constituents of methods and tools

It also helps to classify existing methods and identify potentials to extend them (e. g. QFD is focusedon the language, CAD on tools). This is represented by the graph in the middle of Figure 2, i. e. that allfour aspects shall be regarded nearly equally. This also implies that the 'method' behind a computertool has to be clear [Ambrosy 1997]. For a classification, there are of course more than these fourcriteria (e. g. efforts, user group, targets). The aspects are iteratively connected to each other, e. g. onestep of the process is supported by a specific representation (language) or even again a method, whilea process also describes how to fill in the language, e. g. in form of a matrix.

3.3 Regard of methods as productsA pragmatic approach to improve methods is to regard them as products themselves. By that, a methodhas to be developed, tested, offered, and sold, as well as there has to be a service and training for it. Italso implies that there may be different competing or alternative methods, that methods can havesuccess or not, which will only be shown by the market, that there is a life cycle of the method withgrowing and declining, as well as also a final end of the method, to which another method may follow.Comparing methods to products means also to refer to similar requirements, as there are e. g. an easeof use, also if the product itself is complex, safety and reliability in its use, modularity and customers'variability, fulfilling strategies like product families or platforms, etc.

3.4 Collaboration of industry, research, and educationBoth the implementation of methods in industry and the closely connected need for methodologyknow-how lead to a strong emphasis of the collaboration between industry, research, and education.The need for knowledge in methodology shows the importance of the individual and the individualeducation. This means in contrast to the before described regard of methods as products, thatmethodology is closely bound to persons, as knowledge of it is by definition. This is on one hand themain way of transferring methods into industry, when those individuals go into practice. On the otherhand it is one of the main that principles of methodology to instruct individuals to both dynamic andsystematic acting, just supported by methods. This is in student education as well as in the furthereducation integrated in graduations or dissertations regarding design methodology.This collaboration benefits in the three dimensions learning, testing, and thinking ahead. Industrylearns new methods, proceedings, and techniques, i. e. method implementation, students learn aboutpraxis in early phases, and research learns about management and general problem solving. Thetesting concerns new methods and proceedings, also for all three of those groups. This also refers tosome freedom in research and progress by trial and error, i. e. that not all results need to be economicalmeaningful which otherwise could also be topic in industry. The thinking ahead is important for theprogress in industry as well as in research and in the personal development. Talking of collaboration,research has a main part in the link between industry designers and software designers: industry can


often only show the need while software designers only see the implementation; the underlying andlinking methodology is part of research, which unfortunately sometimes seem to be competitor ofcommercial IT-companies.

3.5 The narrow meaning of science regarding design methodologyTalking of flexibility in design methodology in order to cope with ever changing processes somehowcontradicts the main principles of design science or science itself: just to find the consistencies withindifferent systems, to 'define' in the meaning of differentiating things that actually belong together, allin order to handle complexity and allow a precise acting. This is what methodology represents,generally valid procedures or tools. But due to the fact that design processes are highly complexsystems, where an accurate comparison is nearly impossible, a method again will never be completeand all someone can come up with. It is some kind of tightrope walk or dialectics between strictlydescribing and prescribing methods for industrial product development and advocating flexibilitywithin design methodology. So both aspects are right, there is a need for a strict methodology in orderto have a platform for discussion and education as well as a reference for design processes, and thereis an even stronger need of understanding that design methodology is a very flexible and dynamicinstrument. This dialectics may show an aspect where design science has to go explicitly.

4. SummaryThough design methodology has become a major part of engineering science and industry is aware ofits need, the use of methods in practice still has to be optimized. It may be reasoned by a wrong designof methods, that they are not understood und used properly. It is important to offer flexible andmanageable methods that are connected to one another along with clarifying that a method is just asupporting tool, that it needs specialized knowledge, and that it is no design automatism. A strictmethodology – e. g. in form of a design theory – may be necessary for discussion and education,which is in fact one of the major aspects of this topic since methodology is closely bound toindividuals. The dialectics between clear methods and their flexible use seems to become a main topicof design science and shall be regarded in theses describing methods explicitly.

1.1 ReferencesAmbrosy, S., "Methoden und Werkzeuge für die integrierte Produktentwicklung", Shaker Aachen, 1997.Ehrlenspiel, K., "Integrierte Produktentwicklung", Hanser Muenchen, 1995.Wach, J., "Problemspezifische Hilfsmittel für die integrierte Produktentwicklung", Hanser Muenchen, 1994.Zanker, W., "Situative Anpassung und Neukombination von Entwicklungsmethoden", Shaker Aachen, 1999.Stetter, R., "Method Implementation in Integrated Product Development", Dr.Hut Muenchen, 2000.Lindemann, U., Wulf, J., "Action orientation in design methodology", Proceedings of the 13th Intern. Conferenceon Engineering Design 2001. ImechE Glasgow, 2001, Vol. "Design Applications", pp. 131-138.Grabowski, H., Geiger, K., "Neue Wege zur Produktentwicklung", Raabe Stuttgart, 1997.Grabowski, H., Rude, S., Grein, G., "Universal Design Theory", Shaker Aachen, 1998.Pahl, G., Beitz, W., "Engineering Design – A Systematic Approach", Springer London, 1996.

Udo Pulm, Dipl.-Ing.Technische Universitaet Muenchen, Institute of Product DevelopmentBoltzmannstr. 15, 85748 Garching, GermanyPhone: +49 (0) 89 289 15155Fax: +49 (0) 89 289 15144Email: [email protected]



REDESIGNING PRODUCT PROGRAMSFOCUSING ON VARIETY, COMPLEXITY ANDCOMMONALITY

J. Reitan, S. Stormo and K. Aasland

Keywords: Product development methodology, Product programs,Multiproduct development

1. IntroductionDuring the last couple of decades there has been a change towards more and more customizedproducts within many industries. Since customers are different, with different needs and criteria forchoice, it has become more and more necessary for companies to offer a spectrum of models andvariants to satisfy various customers. Market demands have increased, and products’ market life spansdecrease. New products or new variants must be introduced more frequently than they used to. Productassortments in companies are growing, which often leads to increased internal complexity.On the other hand production and other internal stakeholders want less complexity and moreeconomies of scale. It can be a challenge to satisfy marked demand and at the same time make money.To survive many companies are shifting from mass production strategy to a variety strategy. The resultof this conflict is increased turnover, but unfortunately reduced profit margin. [Schuh 1999]Continuous product development activities, is a necessity for companies that want to be in theforefront of development, and focus on multiproduct development instead of development of singleproduct is vital. One of the challenges for companies with high internal complexity is how they canachieve mass production benefits and at the same time offer a variety of products. Managingcomplexity with product modelling or configuration is one way another solution is reducingcomplexity through means like standardisation, product platforms and modularization.We are focusing on what we call a product program. A product program is a planned product portfoliofor a company, which considers both internal and external impacts of the products. The model forimproving product program complexity presented in this paper contains a framework, methods andtools for reducing complexity, increasing commonality and increasing external variety.

2. Product Program CharacteristicsOne of the reasons why product variety has been a problem for many companies is that the newproducts are not planned and considered in relation to the rest of the product assortment. This is whythe product assortment should be planned and coherent with the strategies of the company. Theproduct program should have positive effects on both the internal and external relations. Only byfulfilling and co-ordinating these criteria consider it to be a product program. Product programcharacteristics have been identified as variety, commonality, complexity and architecture [Andreasen2001].

2.1 External varietyIs it necessary for the company to offer “everything”? We do not think so. But it is necessary to offer


exactly what the customers want. The company needs to focus on differentiating attributes in order tooffer distinct products and avoid cannibalisation. Variety should be related to the market strategy andmarket segments.

Figure 1. Product program characteristics [Andreasen 2001]

Positive variety is defined as “variation that is customer- driven and, as such, directly linked toverifiable customer interest or demand; adds value and increases sales but does not add unwarrantedcosts.”[Galsworth 1994] What companies should strive for is not just product variety, but positivevariety to increase sales and profit.

2.2 Internal CommonalityCommonality can be introduced in the product itself, but also in activities or processes connected tothe company. Reuse, for example in the manufacturing process, can lead to reduced complexity andreduced costs. Not all components should be reused if it is important for the customers that they aredistinct.A platform is defined as a collection of values that share something common for a product. This couldbe commonalty in components, processes, knowledge or staff and relationship. [Robertson and Ulrich1998]

2.3 Organisational and product complexityBecause of the increasing variety and product complexity, a lot of companies experience increasingorganisational complexity. This complexity results in extra, unwarranted costs, both indirect anddirect, which is difficult to spot in traditional accounting principles.Reducing organisational complexity can be done through standardisation and commonality principlessuch as product platforms.

2.4 Product ArchitectureProduct architecture is defined as “(1) the arrangement of functional elements; (2) the mapping fromfunctional elements to physical components; (3) the specification of the interfaces among interactingphysical components.” [Ulrich 1998] An architecture can be modular (one to one mapping betweenfunction and components) or integral.Modularization is a way of organising the product architecture in order to achieve positive effects bothinternal and external.

3. The redesign processWe have realised a need for a systematic approach for redesign processes, focus on internalcommonality and external variety. A model to improve and develop an existing product program hastherefore been developed. This model is both guidelines for development or redesign projects,


containing simple solutions and methods. The intention for this model is through a methodical wayincrease internal commonality and at the same time increase external variety.Compared with traditional models for integrated product development, these model focuses especiallyon product program characteristics. This means according to the model, focus externally on variety,internally on commonality and complexity and focus on the product architecture. These differentfocuses are separated in the first phases, but should be considered throughout the process.

Figure 2. Model for redesigning product programs

The model is divided into 6 different phases, each containing different tools, and depending on theperformance. The tools itself is nothing new, just a collection of related methods. The methods areinputs in a specific phase, considered from our view and needs, not considering the main purpose forthe methods.

3.1 ObserveThe first step in this process is to observe and visualize/describe the existing product program. Wewant to observe the products itself (components, modules, architecture etc.) but also the internal andexternal circumstances (production activities, market segments etc.).The observe phase is considered as a state of the art, where only available/written information is ofinterest. This information, which is part of the strategies, is available in annual- and progress reports,bill of materials and marketing material like product assortment lists and catalogues.The outcome should be big sheets of paper, which visualises and describes existing product program,but also emphasises the aspects which needs further analysis.

3.2 AnalyseIn the same way as the phase above, this phase is divided into internal, product and externalconsiderations. The intention is to go thoroughly into aspects which is seen as defective from theprevious phase. It is important to mention that exactness in these two phases, is essential to makecorrect recommendations.

3.3 RelateThe three considerations that were separated in the previous phases are related in this phase.According to figure 1, product program characteristics [Andreasen 2001], there is a relationshipbetween product program characteristics. To focus on e.g. external variety itself, without concerningon other product program characteristics, do not consider it to be a “product program”. All relevantinformation from previous phases are input.


3.4 RestrictIn addition to product program characteristics, a product program is attached to a timeplan.Requirements which regards product program development, should contain top level strategies (e.g.corporate identity) but also future-oriented guidelines due to product program characteristics, inaddition to more traditional project depending specifications. [Figure 3] The guidelines should havelimited duration of e.g. 5-10 years.

Figure 3. Future-oriented plan for synchronising new products

3.5 GenerateThe outcomes of the generation phase are new concepts of products or improved processes based onthe guidelines and strategies developed in previous phases. Due to product programs characteristicthese concepts do not need to be noticeable amendments for the customers, but e.g. merely internalimprovements.

3.6 EvaluateThis phase contains the same measures and visualising methods as in the first stages, and are used tocompare situations before and after. An evaluation, through fulfilled restrictions or not, it isdetermined whether the concepts are ready for implementation or have to go one more lap.

3.7 Tools and methodsIn table 1 different tools and methods are presented. We have implemented and tested some of thesetools in connection to the model. This is not a complete list, but some of the methods that can be foundin the literature.

Table 1. Examples of different tools applied to each phase

Phase: Tools: References:internal Variant tree (“Variantenbaum”) [Schuh 1999]product Part index [Galsworth 1994]

Observe

external Product characteristic “Merkmalbaum” [Schuh 1999]internal Commonality plan

VAT (Variety effectiveness process analysing tools)[Robertson ea. 1998 ][Galsworth 1994]

product Product family master plan (PFMP) [Mortensen ea. 2000]

Analyse

external Differentiation planCompetitor analysis

[Robertson ea. 1998]

Relate Relationship matrices [Robertson ea. 1998]RestrictGenerate Traditional concept generating tools

MFD (Modular Function Deployment) [Ericson ea. 1999]Evaluate QFD, evaluating indexes


4. Conclusion and further researchIn this paper we have presented a model for improving existing product programs. The model itself isjust a guideline, including tools and methods for support. To develop product program and achieve thebenefits explained in the paper, a product program mindset (according to product programcharacteristics) is needed.Our intention for this paper was to present a case story from a kitchen manufacturer, where we haveused this model to support a process of improving the product program for this company. Due toconfidentiality issues the experiences of this project can not be presented in this paper.The model seems like a plausible approach to develop and improve product programs. This research isonly implemented in one case study and need to be implemented on several projects to get some moregeneral results.

ReferencesAndreasen, M.M,”Hva påstår vi forskere at vi ved om multiproduktutvikling?” Proceedings ofProdukutviklingsdagen, Multiproduktutvikling, DTU Copenhagen, 2001Ericsson, A and G. Erixon “Controlling design variants : modular product platforms“ Dearborn Mich 1999Galsworth, G.D.” Smart, simple design” Oliver Wight Publications Vermont US, 1994Mortensen, N.H, U. Harlou, M.P Nielsen and M..M. Andreasen, “Procedure for modelling product families inconfiguration systems”, proceedings of Design for Configuration, Tampere, 2000Robertson, D. and K. Ulrich ”Platform Product Development”, Sloan Management Review, Vol.39 No. 4, 1998,pp 19-31Schuh, G. “GPS Komplexitätsmanagement GmbH”, Variantenmanagement für die Praxis, seminar, 1999.Ulrich, K. and S. Eppinger “Product Design and Development”, 2000Ulrich, K., “Managing Product Variety: A study of the Bicycle Industry” to be published in Managing ProductVariety, Ho T. and C. Tang (editors) Kluwer, 1998

Jarl ReitanSINTEF, Industrial ManagementS.P.Andersensvei 5, N-7465 Trondheim, Norway+47 73 55 06 59, +47 73 59 36 70Email: [email protected]



DEVELOPMENT AND APPLICATION OFMODULAR FUNCTION DEPLOYMENT

Ctibor Stadler and Stanislav Hosnedl

Keywords: Modularisation, Design Science, Technical System,Customer Requirement, Evaluation, Design Process, KnowledgeSupport, Frame of Machine Tool, FEM

1. IntroductionModular Function Deployment (MFD) is a method [Erixon 1998], which is based on the well-knownQuality Function Deployment (QFD) method. The QFD is more and more used because of itsorientation on customer requirements on technical products. It is assumed as one of the best TotalQuality Management (TQM) method. The goal of QFD is to ensure optimal fulfillment of customerrequirements on a single product only. The MFD enables above to ensure the same task by optimalmodular product or maybe its part composed only from the selected modules if required. This isconsiderably more effective both for the producer and the customer.According to the analyses performed MFD has, in spite of its very careful and detailed form, one weakpoint. This is insufficient interconnection to a designed product. These relationships are largelygenerally described in Design Science based on the Theory of Technical Systems (TTS) [Hubka &Eder 1996]. One of the most significant results of this approach is that all phases of the EngineeringDesign Process and their operations are systematically linked with general knowledge on any designedtechnical product/system (TS).

2. Enhancement of MFD

2.1 Knowledge support of engineering design processIn general, knowledge support of engineering design process can be hierarchically structured into thefollowing three levels: intuitive, methodical and systematic one [Hosnedl, Vanek & Borusikova 2001].The intuitive approach: We have “no” theories, “no” methods, only previously acquired general andspecialized knowledge and practical experience. In this case e.g. the scope of considered properties ofthe TS depends only on intuitive decisions of engineering designers. The methodical approach: Wehave prescriptive and normative instructions, which restrict space for creative design process, e.g. BS2000, VDI 2221. According to our analyses MFD has been used up to now on this level. The flexiblesystematic approach: We have moreover a system of pieces of engineering design knowledge and theirrelations, i.e. Design Science, which enables to manage and optimally support creative engineeringdesign process. The MFD has been enhanced to this level with the use of the above-mentionedtheories.The general structure of the engineering design process is shown in the Fig. 1. It represents synthesisof the known design process models from Eekels, Koller, Hubka & Eder, Pahl & Beitz, VDI 2221, etc.[Hosnedl 1997].


Figure 1. Structure of the engineering design process of the technical system

2.2 What is modularisation?The modularisation is a decomposition of technical system into building blocks (modules). Themodules must have specified interfaces. We can distinguish three levels of modularisation accordingto the structures of technical systems as follows: function, organ and constructional one. Themodularisation on the higher level always includes modularisation on all lower levels, i.e. organmodularisation includes constructional modularisation and function modularisation includes organ andconstructional modularisation. A typical example of the function modularisation is a personalcomputer (PC). The PCs can have e.g. CD-ROM device for reading of CD-ROM (one module),graphical card for displaying of picture (another module) and sound card for reproduction of sound


(another module). These modules fulfil different functions. An example of the organ modularisation isa car, which has relatively uniform function structure. The cars can have e.g. engine (module) workingon different principles (explosion or Diesel one). This module fulfils the same function. An exampleof constructional modularisation is a rolling bearing, which has the same function and organ structure.Rolling bearings can differ in sizes, shapes, etc., which fulfil the same functions performed by thesame organs.

2.3 Engineering design process with the use of enhanced MFDWe will proceed step-by-step according to the structure of the engineering design process (Fig. 1.)with the use of MFD for organ modularisation of the designed TS.2.3.1 Elaborating the assigned problem - design specification on properties of technical systemWe use the theory of properties from the Theory of Technical Systems [Hubka & Eder 1996].According to this theory the properties are separated into two basic groups as follows: external andinternal one. The external properties are required and judged by customer. The internal properties aregenerated and controlled by engineering designer. Sometimes the customer has also requirements onsome internal properties. The external properties causally depend on the internal properties. Bothgroups of these properties are separated into twelve classes as shown in the Fig. 2.

Figure 2. Classes of properties of technical systems (TS) and their basic relationships

To include all life cycle requirements on the designed TS the design specification should be createdwith the use of these classes of properties. The enhanced MFD is an effective way, how to perform it.This step is using the central organ of MFD, called The House of Modularity, as is shown in the Fig.3.


Figure 3. Design specification in The House of Modularity

2.3.2 Establishing the function and organ structure – concept of technical systemBefore establishing the function structure the main and assistant operational technical processesrelated to the designed TS should be analysed. For example if we design a carriage for lathe we mustestablish the operational processes for the lathe because in this case the technical process of turning isrunning “in the lathe”. The carriage is a technical sub-system of the lathe only. The establishedfunction structure can be represented e.g. in a form of hierarchical function tree. For establishing theorgan structure we can use e.g. the method of morphological matrix and/or TRIZ system. Theabbreviation TRIZ in the Russian language means “Theory of Solution of Invention Tasks” and TRIZsystem in form of a software product is called Tech Optimizer.

Figure 4. The House of Modularity

We then put organs for each variant of organ structure to The House of Modularity (Fig. 4) andevaluate suitableness of their properties both for fulfilment of the given requirements and modulardesign of the designed TS. The optimal organ structure from the viewpoints of design specificationand modular design can be then evaluated.

3. Concept and support of applicationThe application has been focused into the area of engineering design of frames for machine tools,specifically on the development and evaluation of their optimal modules. The first problem here deals


with the identification of functions ensured by organs of the constructional structure. This has beenperformed by evaluation of frame modules with the use of Finite Element Method (using softwareMARC, ANSYS and I-DEAS).A straight constructional module with rectangular cross-section has been designed because it hasdifferent properties in all directions and is simple (Fig. 5). It is suitable for mutual comparing of singleproperties of constructional variants of this module. The sizes of the cross-section have beendetermined thus that the height is double of the breadth. The length of the module has been calculatedon the base of a theory of buckling thus that in case of compressive loading the module will bestrained by buckling. The material of module for determination of its length is isotropic steel becauseit has the highest limiting depth-thickness ratio from current frame materials. The module for statingits length is a plate one. The module is rigidly fixed at one end and loaded by forces and moments inall three directions (Fx, Fy, Fz, Mx, My, Mz). Orientation of axial force has been chosen thus that themodule will be strained by tension.

Figure 5. Module for evaluation with loading

The sizes of cross-section: b = 100 mmh = 2 b = 2 . 100 = 200 mm

The limiting depth-thickness ratio: λm = 100The limiting length of module (from the theory of buckling):

100 1001443

48 48m

m

bl mm

λ ⋅= = = (1)

The chosen length of module: l = 1700 mmThe values of loading:Fx = 500 kN Mx = 12 000 NmFy = 10 kN My = 7 000 NmFz = 15 kN Mz = 10 000 Nm

These values have been chosen thus that they would cause real strains within module.

Figure 6. Several types of modules for evaluation

Each variant of the module for evaluation have the same external sizes (breadth, height and length),gripping and loading. The material or the internal structure of modules can be different. The solution


of the module will be performed by FEM software. The displacements of point in which the loadinghas been will be read after the solution. Six stiffnesses (kx, ky, kz, ktx, kty, ktz) will be then computedfrom loading and displacements as follows:

x

Fk x

x ∆=

x

xtx

Mk

ϕ∆= (2, 3)

The answer to the question: “Which organ of the module fulfils required function at the best?” will befound according to the computed stiffnesses of all the solved modules. If we would like to get otherresults we will take the same modules and we can solve e.g. buckling, dynamics or durability. Theseresults will serve as a support for engineering design of optimal frame module for machine toolsaccording to the procedure, which has been described, in the previous chapter.

4. ConclusionThe paper presents flexible systematic procedure of engineering design of modular technical system.The modular technical system is more and more important because joints together the process ofstandardisation with multifarious solutions, which fulfil customer requirements at the best. Themodular solution also simplifies reconfiguration of TS if the customer requirements are changed. It isachieved with the use of modules with clearly defined interfaces. The next advantage of a modularsolution is that producer or customer can accomplish the reconfiguration himself. The paper proves theimportance of application of the Finite Element Method for evaluation of different concepts ofmodules for the design of frames for machine tools before the concrete constructional structure of itsmodules is achieved. This evaluation answers the question which organ of the constructional structurefulfils a required function at the best. These results can serve for the re-design of the developedmodular structure and its modules to fulfil the given requirements optimally. It will contribute to thehigher quality of engineering design of frames for machine tools and to shortening of the design time.

ReferencesAndreasen, M., M. and Blessing, L., Proceedings of “European PhD-Course “Engineering Design Research2001”, Part I – II, Vasteras SE and Magdeburg D, 2001.Erixon, G., “Modular Function Deployment – A Method for Product Modularisation”, ISSN 1104-2141, TheRoyal Institute of Technology, Stockholm SE, 1998.Hosnedl, S., Vanek, V. and Borusikova, I., “Design Science for Engineering Design Practice”, Proceedings ofICED 2001. Glasgow UK, 2001, pp 363-370.Hosnedl, S., “The structure of diploma design projects”, University of West Bohemia in Pilsen CR, 1997.Housa, J., “ Design of highly dynamic machine tools and usage of laser in cutting”, CTU Prague CR, 2001.Hubka, V. and Eder, W., E., “Design Science”, Springer-Verlag London UK, 1996.

Ctibor Stadler M.Sc.University of West Bohemia in Pilsen, Faculty of Mechanical Engineering, Department of Machine DesignUniverzitni 8, P.O. Box 314, 306 14 Pilsen, Czech RepublicTelephone: +420-19-7491556Telefax: +420-19-7423185Email: [email protected]

a study on the thinking path model

Documents