工業(yè)產(chǎn)品設(shè)計外文翻譯參考文獻(xiàn)

工業(yè)產(chǎn)品設(shè)計外文翻譯參考文獻(xiàn)工業(yè)產(chǎn)品設(shè)計外文翻譯參考文獻(xiàn)(文檔含中英文對照即英文原文和中文翻譯)DesignWithoutDesignersIwillalwaysremembermyfirstintroductiontothepowerofgoodproductdesign.IwasnewlyarrivedatApple,stilllearningthewaysofbusiness,whenIwasvisitedbyamemberofApple'sIndustrialDesignteam.Heshowedmeafoammockupofaproposedproduct."Wow,"Isaid,"Iwantone!Whatisit?"Thatexperiencebroughthomethepowerofdesign:IwasexcitedandenthusiasticevenbeforeIknewwhatitwas.Thistypeofvisceral"wow"responserequirescreativedesigners.Itissubjective,personal.Uhoh,thisisnotwhatengineersliketohear.Ifyoucan'tputanumbertoit,it'snotimportant.Asaresult,thereisatrendtoeliminatedesigners.Whoneedsthemwhenwecansimplytestourwaytosuccess?Theexcitementofpowerful,captivatingdesignisdefinedasirrelevant.Worse,thenatureofdesignisindanger.Don'tbelieveme?ConsiderGoogle.Inawell-publicizedmove,aseniordesigneratGooglerecentlyquit,statingthatGooglehadnointerestinorunderstandingofdesign.Google,itseems,reliesprimarilyupontestresults,nothumanskillorjudgment.Wanttoknowwhetheradesigniseffective?Tryitout.Googlecanquicklysubmitsamplestomillionsofpeopleinwell-controlledtrials,pittingonedesignagainstanother,selectingthewinnerbaseduponnumberofclicks,orsales,orwhateverobjectivemeasuretheywish.Whichcolorofblueisbest?Test.Itemplacement?Test.Webpagelayout?Test.ThisprocedureishardlyuniquetoGoogle.Ahaslongfollowedthispractice.YearsagoIwasproudlyinformedthattheynolongerhavedebatesaboutwhichdesignisbest:theysimplytestthemandusethedatatodecide.Andthis,ofcourse,istheapproachusedbythehuman-centerediterativedesignapproach:prototype,test,revise.Isthisthefutureofdesign?Certainlytherearemanywhobelieveso.Thisisahottopiconthetalkandseminarcircuit.Afterall,theproponentsaskreasonably,whocouldobjecttomakingdecisionsbasedupondata?

TwoTypesofInnovation:IncrementalImprovementsandNewConcepts

Indesign—andalmostallinnovation,forthatmatter—thereareatleasttwodistinctforms.Oneisincrementalimprovement.Inthemanufacturingofproducts,companiesassumethatunitcostswillcontinuallydecreasethroughcontinual,incrementalimprovements.Asteadychainofincrementalinnovationenhancesoperations,thesourcingofpartsandsupply-chainmanagement.Theproductdesigniscontinuallytinkeredwith,adjustingtheinterface,addingnewfeatures,changingsmallthingshereandthere.Newproductsareannouncedyearlythataresimplysmallmodificationstotheexistingplatformbyadifferentconstellationoffeatures.Sometimesfeaturesareremovedtoenableanew,low-costline.Sometimesfeaturesareenhancedoradded.Inincrementalimprovement,thebasicplatformisunchanged.Incrementaldesignandinnovationislessglamorousthanthedevelopmentofnewconceptsandideas,butitisbothfarmorefrequentandfarmoreimportant.Mostoftheseinnovationsaresmall,butmostarequitesuccessful.Thisiswhatcompaniescall"theircashcow":aproductlinethatrequiresverylittlenewdevelopmentcostwhilebeingprofitableyearafteryear.Thesecondformofdesigniswhatisgenerallytaughtindesign,engineeringandMBAcourseson"breakthroughproductinnovation."Hereiswherenewconceptsgetinvented,newproductsdefined,andnewbusinessesformed.Thisisthefunpartofinnovation.Asaresult,itisthearenathatmostdesignersandinventorswishtoinhabit.Buttherisksaregreat:mostnewinnovationsfail.Successfulinnovationscantakedecadestobecomeaccepted.Asaresult,thepeoplewhocreatetheinnovationarenotnecessarilythepeoplewhoprofitfromit.InmyAppleexample,thedesignersweredevisinganewconception.InthecaseofGoogleandAmazon,thecompaniesarepracticingincrementalenhancement.Theyaretwodifferentactivities.NotethattheAppleproduct,likemostnewinnovations,failed.Why?Ireturntothisexamplelater.Bothformsofinnovationarenecessary.Thefightoverdata-drivendesignismisleadinginthatitusesthepowerofonemethodtodenytheimportanceofthesecond.Data-drivendesignthroughtestingisindeedeffectiveatimprovingexistingproducts.Butwheredidtheideafortheproductcomefrominthefirstplace?Fromsomeone'screativemind.Testingiseffectiveatenhancinganidea,butcreativedesignersandinventorsarerequiredtocomeupwiththeidea.

WhyTestingIsBothEssentialandIncomplete

Data-drivendesignis"hill-climbing,"awell-knownalgorithmforoptimization.Imaginestandinginthedarkinanunknown,hillyterrain.Howdoyougettothetopofthehillwhenyoucan'tsee?Testtheimmediatesurroundingstodeterminewhichdirectiongoesupthemoststeeplyandtakeastepthatway.Repeatuntileverydirectionleadstoalowerlevel.Butwhatiftheterrainhasmanyhills?Howwouldyouknowwhetheryouareonthehighest?Answer:youcan'tknow.Thisiscalledthe"localmaximum"problem:youcan'ttellifyouareonhighesthill(aglobalmaximum)orjustatthetopofasmallone.Whenacomputerdoeshillclimbingonamathematicalspace,ittriestoavoidtheproblemoflocalmaximabyinitiatingclimbsfromnumerous,differentpartsofthespacebeingexplored,selectingthehighestoftheseparateattempts.Thisdoesn'tguaranteetheveryhighestpeak,butitcanavoidbeingstuckonalow-rankingone.Thisstrategyisseldomavailabletoadesigner:itisdifficultenoughtocomeupwithasinglestartingpoint,letalonemultiple,differentones.So,refinementthroughtestingintheworldofdesignisusuallyonlycapableofreachingthelocalmaximum.Isthereafarbettersolution(thatis,isthereadifferenthillwhichyieldsfarsuperiorresults)?Testingwillnevertellus.Hereiswherecreativepeoplecomein.Breakthroughsoccurwhenapersonrestructurestheproblem,therebyrecognizingthatoneisexploringthewrongspace.Thisisthecreativesideofdesignandinvention.Incrementalenhancementswillnotgetusthere.

BarrierstoGreatInnovation

Dramaticnewinnovationhassomefundamentalcharacteristicsthatmakeitinappropriateforjudgmentthroughtesting.Peopleresistnovelty.Behaviortendstobeconservative.Newtechnologiesandnewmethodsofdoingthingsusuallytakedecadestobeaccepted-sometimesmultipledecades.Butthetestingmethodsallassumethatonecanmakeachange,tryitout,andimmediatelydetermineifitisbetterthanwhatiscurrentlyavailable.Thereisnoknownwaytotellifaradicalnewideawilleventuallybesuccessful.Hereiswheregreatleadershipandcourageisrequired.Historytellsusofmanypeoplewhoperseveredforlongperiodsinthefaceofrepeatedrejectionbeforetheirideawasaccepted,oftentothepointthataftersuccess,peoplecouldnotimaginehowtheygotalongwithoutitbefore.Historyalsotellsusofmanypeoplewhoperseveredyetneverwereabletosucceed.Itispropertobeskepticalofradicalnewideas.Intheearlyyearsofanidea,itmightnotbeacceptedbecausethetechnologyisn'tready,orbecausethereisalotmoreoptimizationstilltobedone,orbecausetheaudienceisn'tready.Orbecauseitisabadidea.Itisdifficulttodeterminewhichofthosereasonsdominates.Thetaskonlybecomeseasyinhindsight,longafteritbecomesestablished.Theselongperiodsbetweenformationandinitialimplementationofanovelideaanditseventualdeterminationofsuccessorfailureinthemarketplaceiswhatdefeatsthosewhowishtouseevidenceasadecisioncriterionforfollowinganewdirection.Evenifasuperiorwayofdoingsomethinghasbeenfound,theautomatedtestprocesswillprobablyrejectit,notbecausetheideaisinferior,butbecauseitcannotwaitdecadesfortheanswer.Thosewholookonlyattestresultswillmissthelargepayoff.Ofcoursetherearesoundbusinessreasonswhyignoringpotentiallysuperiorapproachesmightbeawisedecision.Afterall,iftheaudienceisnotreadyforthenewapproach,itwouldinitiallyfailinthemarketplace.Thatistrue,intheshortrun.Buttoprosperinthefuture,thebestapproachwouldbetodevelopandcommercializethenewideatogetmarketplaceexperience,tobegintheoptimizationprocess,andtodevelopthecustomerbase.Atthesametimeoneispreparingthecompanyforthedaywhenthemethodtakesoff.Sure,keepdoingtheold,butgetreadyforthenew.Ifthecompanyfailstorecognizethenewlyemergingmethod,itscompetitorswilltakeover.Quiteoftenthesecompetitorswillbeastartupthatexistingcompaniesignoredbecausewhattheyweredoingwasnotwellaccepted,andinanyeventdidnotappeartochallengetheexistingbusiness:see"Theinnovator'sdilemma."Gestural,multi-touchinterfacesforscreen-drivendevicesandcomputergamesaregoodexamples.Aretheseabrilliantnewinnovation?Brilliant?Yes.New?Absolutelynot.Multi-touchdeviceswereinresearchlabsforalmostthreedecadesbeforethefirstsuccessfulmass-producedproducts.Isawgesturesdemonstratedovertwodecadesago.Newideastakeconsiderabletimetoreachsuccessinthemarketplace.Ifanideaiscommercializedtoosoon,theresultisusuallyfailure(andalargelossofmoney).ThisispreciselywhattheAppledesignerofmyopeningparagraphhaddone.WhatIwasshownwasaportablecomputerdesignedforschoolchildrenwithaformfactorunlikeanythingIhadeverseenbefore.Itwaswonderful,andeventomynormallycriticaleye,itlookedlikeaperfectfitforthepurposeandaudience.Alas,theproductgotcaughtinapoliticalfightbetweenwarringAppledivisions.Althoughitwaseventuallyreleasedintothemarketplace,thefightcrippleditsintegrityanditwasbadlyexecuted,badlysupported,andbadlymarketed.Theresistanceofacompanytonewinnovationsiswellfounded.Itisexpensivetodevelopanewproductlinewithunknownprofitability.Moreover,existingproductdivisionswillbeconcernedthatthenewproductwilldisruptexistingsales(thisiscalled"cannibalization").Thesefearsareoftencorrect.Thisisaclassiccaseofwhatisgoodforthecompanybeingbadforanexistingdivision,whichmeansbadforthepromotionandrewardopportunitiesfortheexistingdivision.Isitawondercompaniesresist?Thedataclearlyshowthatalthoughafewnewinnovationsaredramaticallysuccessful,mostfail,oftenatgreatexpense.Itisnowonderthatcompaniesarehesitant-resistant-toinnovationnomatterwhattheirpressreleasesandannualreportsclaim.Tobeconservativeistobesensible.

TheFuture

Automateddata-drivenprocesseswillslowlymakemoreandmoreinroadsintothespacenowoccupiedbyhumandesigners.Newapproachestocomputer-generatedcreativitysuchasgeneticalgorithms,knowledge-intensivesystems,andotherswillstarttakingoverthecreativeaspectofdesign.Thisishappeninginmanyotherfields,whetheritbemedicaldiagnosisorengineeringdesign.Wewillgetmoredesignwithoutdesigners,butprimarilyoftheenhancement,refinement,andoptimizationofexistingconcepts.Evenwherenewcreativeartificialsystemsaredeveloped,whetherbyneuralnetworks,geneticalgorithms,orsomeyetundiscoveredmethod,anynewconceptwillstillfacethehurdleofovercomingtheslowadoptionrateofpeopleandofovercomingthecomplexpsychological,social,andpoliticalneedsofpeople.Todothis,weneedcreativedesigners,creativebusinesspeople,andrisktakerswillingtopushtheboundaries.Newideaswillberesisted.Greatinnovationswillcomeatthecostofmultiplegreatfailures.Designwithoutdesigners?Thosewhodisliketheambiguityanduncertaintyofhumanjudgments,withitsuncertaintrackrecordandcontradictorystatementswilltrytoabolishthehumanelementinfavorofthecertaintythatnumbersanddataappeartooffer.Butthosewhowantthebiggainsthatcreativejudgmentcanproducewillfollowtheirownjudgment.Thefirstcasewillbringaboutthesmall,continualimprovementsthathavecontributedgreatlytotheincreasedproductivityandloweringofcostsofourtechnologies.Thesecondcasewillberewardedwithgreatfailuresandoccasionalgreatsuccess.Butthosegreatsuccesseswilltransformtheworld.不需要設(shè)計師的設(shè)計唐·諾曼我永遠(yuǎn)也不會忘記我第一次向人們介紹優(yōu)秀產(chǎn)品設(shè)計的魅力的經(jīng)歷，那時候我剛剛到蘋果公司，還在逐漸的學(xué)習(xí)工作上的事務(wù)。有一個蘋果工業(yè)設(shè)計小組的成員來我這里，向我展示了一個即將推出的產(chǎn)品的泡沫模型，“喔！”我說，“這是什么？我也想要個！”那次經(jīng)歷讓我體驗到了設(shè)計的原始力量：當(dāng)我還不知道他具體是什么之前我就已經(jīng)興奮不已，充滿熱情了。這種發(fā)自肺腑的回應(yīng)離不開很有創(chuàng)意的設(shè)計師。這種想法很主觀，也很有個人感情色彩。哦，不過工程師們可不愿意聽到這些。如果你不能提供和它有關(guān)的數(shù)據(jù)，它就沒什么了不起。這樣的結(jié)果是有一種不再需要設(shè)計師的趨勢。當(dāng)我們可以簡單的測試我們的成功之路時，誰還需要設(shè)計師呢？令人充滿激情興奮無比的設(shè)計被看得無足輕重。更嚴(yán)重的是設(shè)計的初衷也岌岌可危了。不相信吧？看看谷歌。最近谷歌的一位高級設(shè)計師有一次在公開場合宣稱，他們對設(shè)計不感興趣也不懂設(shè)計。據(jù)說，谷歌依靠最原始的測試結(jié)果而不是人類技巧和判斷。怎么知道一個設(shè)計是否成功呢？測試一下就可以了。谷歌會迅速地把樣品發(fā)送給對照試驗中數(shù)以萬計的用戶，與其他的設(shè)計做個對比，然后選出優(yōu)勝者。他們可以靠點擊量，銷售量以及其他任何他們想要采用的客觀依據(jù)。什么顏色的制服最好？測試一下；哪種項目布置最合理？測試一下；哪種網(wǎng)頁排版最好呢？測試一下。這可不是谷歌的專利，亞馬遜早就也這么做了。幾年前我很榮幸的被告知它們不再為哪個設(shè)計最好而爭論不休了，他們會測試一下然后用數(shù)據(jù)來決定。當(dāng)然，這個也是以人為本的迭代設(shè)計法采用的途徑：原型，試驗和修改。這是設(shè)計的未來嗎？有很多人會真么認(rèn)為。這是一個人們談?wù)摵脱芯拷涣鞯臒衢T話題，畢竟，支持者也有理有據(jù)：誰不想靠數(shù)據(jù)來做決定？兩種類型的創(chuàng)新：不斷改善和全新的概念在設(shè)計和幾乎所有改革中，其實都至少有兩種不同的類型。第一種是持續(xù)改進(jìn)現(xiàn)有產(chǎn)品，在產(chǎn)品制造業(yè)中企業(yè)認(rèn)為通過不斷地改善和優(yōu)化單位成本也會持續(xù)的降低。不斷改善的帶來穩(wěn)定的利益鏈條又強(qiáng)化了操作，資源部門和產(chǎn)業(yè)鏈管理。產(chǎn)品的設(shè)計并沒有停止，改變一下外表，增加一些新的功能，不時的做些小的改動。新的產(chǎn)品都是對現(xiàn)有平臺很小的改動，每年都宣稱有了與眾不同的特征。有時候一些功能被去掉以用來支持一條新的，低成本的生產(chǎn)線，有時候很多功能又被組合或被添加上。產(chǎn)品不斷地改善，但基礎(chǔ)的平臺一直沒有改變。持續(xù)的設(shè)計和改進(jìn)可沒有開發(fā)新概念或新理念那樣的引人矚目，但是它們很常見也很重要。很多這樣的創(chuàng)新都是小規(guī)模的，但大多數(shù)都很成功。這就是企業(yè)們所說的“搖錢樹”：一條只需要很小改進(jìn)的生產(chǎn)線，但是卻可以年復(fù)一年的有利可圖。第二種類型的設(shè)計就是在設(shè)計，工程和MBA課程中經(jīng)常談?wù)摰降摹坝型黄菩缘膭?chuàng)新設(shè)計”。這里提出了全新的概念，新穎的產(chǎn)品定義和新型的商業(yè)模式，而且這些正是設(shè)計的樂趣所在。因此，這也是大多數(shù)的設(shè)計師和發(fā)明家樂意為之的地方。但是風(fēng)險也很大：絕大多數(shù)的新發(fā)明都以失敗告終。那些成功的設(shè)計發(fā)明往往需要數(shù)十年才得到了人們的認(rèn)可，這樣的后果就是發(fā)明者不一定就是以它們獲利的人。在我剛才提到的蘋果公司的例子中，設(shè)計者正在開發(fā)一種新概念產(chǎn)品。在谷歌和亞馬遜的例子中，這些公司在不斷地實踐著不斷的優(yōu)化。它們是兩種不同的行為，看看蘋果的產(chǎn)品，像大多數(shù)的新發(fā)明設(shè)計一樣失敗了。為什么呢？我一會兒再回到這個案例中。這兩種設(shè)計都是很有必要的。對數(shù)據(jù)主導(dǎo)型設(shè)計的激烈爭論誤導(dǎo)了人們，我們用前者的力量否定了后者的重要性。通過測試數(shù)據(jù)主導(dǎo)型設(shè)計對改進(jìn)現(xiàn)有的產(chǎn)品很有效果。但是新產(chǎn)品最初的觀念有從何而來？一些人創(chuàng)造性的想法。測試可以高效的優(yōu)化一個想法，但是創(chuàng)造性的設(shè)計者和發(fā)明家卻需要有自己的想法。為什么說測試既很有必要又不太完美數(shù)據(jù)主導(dǎo)型的設(shè)計就是“爬山策略”，我們熟知的一種追求最優(yōu)化的算法。假設(shè)你在黑夜里站在一個連綿起伏的山坡上，你什么也看不到，你怎么知道你就站在山坡的最高處？檢驗一下你周圍的環(huán)境，判斷哪個方向最陡峭，然后向這個方向邁進(jìn)。這樣不斷的重復(fù)而知道每個方向就找到了最低的地方。但是如果山坡上有很多的山峰又該怎么做呢？你怎么知道你是否已經(jīng)在最高的地方了？答案是你會不知道。這就是所謂的“局部最大值”問題：你不能區(qū)分你是在最高處呢還是只在一個小山坡的最高點。當(dāng)計算機(jī)在數(shù)學(xué)空間里攀登時，它可以通過無數(shù)次的嘗試來探索不同的空間以避免局部最大化的難題。雖然這不能保證可以找到真正的最高點，但至少可以避免掉入低層次的行列中。對設(shè)計師來說這種戰(zhàn)略幾乎毫無用處。解決一個單一的起點就夠困難了，更不用說錯綜復(fù)雜的問題了。通過測試了改良設(shè)計通常能夠達(dá)到局部的最大利益。還有更好的解決辦法嗎（就是說，有沒有受益大于測試結(jié)果的情況）？測試不能告訴我們。這時候就得靠有創(chuàng)意的人了，他對問題的重新組合，于是就決定去看似錯誤的地方探索一下，新的突破就是這樣產(chǎn)生的。這正是設(shè)計發(fā)明創(chuàng)造性的一面，不斷地改良和完善不能讓我們擁有這樣的效果。偉大發(fā)明的障礙激動人心的新發(fā)明往往有一些基本的特點讓它們不適應(yīng)由測試所做出的判斷。人們往往也不太喜歡獵奇，行動也很保守。新的科技發(fā)明和方法往往經(jīng)過數(shù)十年或者更長才逐漸被人們認(rèn)可接受。但是測試的法子都是假設(shè)某個東西很有前途值得一試，并來判斷它是否比正在使用的更好。我們沒有現(xiàn)成的方法判斷一個十分新奇的想法會獲得成功，這就需要出色的領(lǐng)導(dǎo)和鼓勵。歷史告訴我們很多在他們的想法被認(rèn)可以前面臨長期不斷的抨擊的人們獲得成功以后就是這樣，沒有它以前，人們不知道是怎么如何度