2025全球技術報告之工業(yè)制造業(yè)洞察：互操作性、混合模式、AI 創(chuàng)新：工業(yè)制造數(shù)字化決勝點

KPMGglobaltechreport—industrialmanufacturing

insights

Interoperability,hybridmodelsandAI

innovationarethebattlegroundsfordigitalexcellenceinindustrialmanufacturing

KPMG.MaketheDifference.

KPMGInternational|

IMhasabove-averageTheroadaheadforMethodologyHowKPMGcanhelp

KeyfindingsManufacturing’sproactive

datamaturityindustrialmanufacturing

IMexcelsat

achievingAIROI

andprogressivespirit

Executivesummary

Executivesummary

Intherapidlyevolvinglandscapeofindustrialmanufacturing(IM),

organizationsareincreasinglyrecognizingtheimperativeofdigital

transformationtoenhanceoperationalefficiency,qualitycontrol,andsustainability.TheKPMGglobaltechreporthighlightsthatindustrialmanufacturingfirmsareattheforefrontofthistransformation,

showcasingthehighestlevelsofdigitalmaturityacrossvarious

technologycategoriescomparedtoothersectors.Thisreport

servesasacriticalresourceforunderstandingthecurrentstateofdigitaladoptioninindustrialmanufacturingandthestrategicstepsnecessaryforcontinuedadvancement.

TheresearchconductedbyKPMGsurveyed2,450executives

from26countries,including368leadersfromtheindustrial

manufacturingsector.Thefindingsrevealthat76percentof

industrialmanufacturingfirmsexpressastrongwillingnessto

embracecutting-edgetechnology,thehighestamongallsectors

surveyed.Notably,thesectorexcelsinAIadoption,with34percentoforganizationsachievingareturnoninvestment(ROI)frommultiple

AIusecases.However,thereportalsoidentifiessignificantmaturitygapsinareassuchassupplychain,procurement,andfinance

functions,whichhinderthefullrealizationofdigitalpotential.

Toaddressthesechallengesandcapitalizeontheopportunities

presentedbydigitaltransformation,severalrecommendationsemergefromthereport.First,organizationsshouldfocusonenhancingtheirdatastrategiestohelpensureseamlessintegrationandanalysis

acrossdisparatesystems.ThisiscrucialforunlockingthefullpotentialofAIandachievingdata-leddecision-making.

Second,upskillingtheworkforceisessentialtohelpbridgetheskillsgapexacerbatedbytheriseofAI.Trainingprogramsshouldtarget

analyticaldecision-makingandfosteradata-centricculture,enablingemployeestoleveragereal-timedataeffectively.Furthermore,

organizationsmustprioritizethedevelopmentofrobustcybersecuritymeasurestoprotectinternaldatanetworks,especiallyastheybegintosharedatainrealtimewithexternalpartners.

KPMGglobaltechreport—industrialmanufacturinginsights

?2025CopyrightownedbyoneormoreoftheKPMGInternationalentities.KPMGInternationalentitiesprovidenoservicestoclients.Allrightsreserved.

Lastly,fosteringacultureofinnovationandagilitywillbevitalformeeting

evolvingclientexpectationsregardingleadtimesandcustomization.By

investinginprocessesthatelevatethevoiceofthecustomer,manufacturers

cangaininsightsthatdrivenewproductlinesandrevenuestreams.

2

?2025CopyrightownedbyoneormoreoftheKPMGInternationalentities.KPMGInternationalentitiesprovidenoservicestoclients.Allrightsreserved.

IMhasabove-averageTheroadaheadforMethodologyHowKPMGcanhelp

KeyfindingsManufacturing’sproactive

datamaturityindustrialmanufacturing

IMexcelsat

achievingAIROI

andprogressivespirit

Executivesummary

Keyfindings

Manufacturing’sproactiveandprogressivespiritispropellingitsdigitalmaturity

Outoftheeightsectorspolled,manufacturingorganizationsaremostlikelytobeinthehigheststageofstrategicmaturityinthemajorityoftheninetechcategoriesmeasured.And

cutting-edgetechnology—thehighest proportionofallsectorssurveyed.

76%

ofindustrialmanufacturingfirmssaytheirworkforcehasanappetitetoembrace

ThesectorexcelsatachievingAIROI,butdisconnects

preventfurtherprogress

ManufacturingisoneofthethreesectorswhereorganizationsaremostlikelytobeatthemostmaturephaseofAIadoption,with

investment(ROI)inseveralAIusecases.

34%

alreadyachievingreturnon

80%

ofrespondentssayintheirleadershiproletheyempowertheirorganizationtostrategicallyinnovatesotheycancapitalizeonmarkettrendswithAI.

Whileithasabove-averagedatamaturity,thesectorcontinuestoholditselftohighstandards

Industrialmanufacturingperformsabovethecross-sectoraverageinthenumberofitsorganizationsthatareinour

toptwolevelsofdatamaturity.

KPMGglobaltechreport—industrialmanufacturinginsights3

IMhasabove-averageTheroadaheadforMethodologyHowKPMGcanhelp

Manufacturing’sproactive

andprogressivespirit

datamaturityindustrialmanufacturing

Executivesummary

Keyfindings

IMexcelsat

achievingAIROI

Manufacturing’sproactiveandprogressivespiritispropellingitsdigitalmaturity

Forthesecondyearrunning,industrial

manufacturingissettingthepacefordigital

transformation.Outofthesectorssurveyed,itis

theonewhoseorganizationsaremostlikelytobeinthehigheststageofstrategicmaturityinsixoutoftheninetechcategoriesmeasured.

Inthisstage,whichwecall‘proactive,’organizationshavesuccessfullydesignedandtestedastrategicvision,achievedleadershipfunding,andare

implementingthatstrategywhileadaptingittomarketdevelopments.

Inallninetechnologycategories,industrialmanufacturingisaheadofthecross-sectoraverageintermsoftheproportionoforganizationsintheproactivestage.

XaaStechnologies(includingpubliccloudormulti-cloud)

Cybersecurity

AIandautomation(includinggenerativeAI)

Dataandanalytics

Moderndelivery(includinglowcode/nocode)

Edgecomputing

Web3(includingblockchainandtokenization)

Quantumcomputing

VR/AR/XR(includingMetaverse)andspatialcomputing

35%

39%

30%

38%

31%

37%

28%

35%

26%

31%

25%

30%

25%

28%

22%

26%

24%

26%

AverageacrossallsectorsIndustrialmanufacturing

Q:Howwouldyoudescribeyourorganization’spositiontodayineachofthefollowingareas?[Thoseanswering‘Weareproactiveinprogressingagainstourstrategyandarecontinuallyevolving’]

Source:KPMGglobaltechreport2024

KPMGglobaltechreport—industrialmanufacturinginsights4

IMhasabove-averageTheroadaheadforMethodologyHowKPMGcanhelp

KeyfindingsManufacturing’sproactive

datamaturityindustrialmanufacturing

IMexcelsat

achievingAIROI

andprogressivespirit

Executivesummary

Strivingfordigitalmaturityandacultureof

innovationisacontinualgoalofandchallengeforourindustrialmanufacturingclients.Todaywe’rehelpingthemarchitectandimplement

strategiesthatembraceAI,machinelearninganddatatransformation,butwithaconcurrentfocusonhowtheirhumantalentcanthriveinsuchenvironments.Bothelementsareequallyimportanttotheirlong-termsuccess.

ClaudiaSaran

HeadofIndustrialManufacturingKPMGintheUS

Thismaturityisunderpinnedbyacultureoftechenthusiasm:

76percentofindustrialmanufacturingfirmssaytheirworkforce

hasanappetitetoembracecutting-edgetechnology—thehighestproportionofallsectorssurveyed.

Themanufacturingsectorappearstobetakingamorecomprehensiveapproachtoitstechnologyevaluation.Thisyear,executivesaredrawingonawiderrangeofsourcestoinformtheirinvestmentdecisions.

Thatsaid,thedrivershaveshiftedintermsofwhichhasthestrongestinfluenceontechchoices.While“followingcompetitors”isstillatopdecisiondriver(85percent)in2024,ithasfallentosecond,behind

“l(fā)ookingtothird-partyguidance”(89percent).

Thesetactics,pairedwiththesector’sproactiveandprogressivespirit,appeartobepayingoff.

72%

ofindustrialmanufacturing

execssaythattheir

organizationissatisfiedwith

thevaluegeneratedbytheir

techinvestments,whichis

abovethecross-sectoraverage.

AccordingtoSaurabhBhatnagar,Partner,IndustrialAutomation,

IntelligenceandDigitalization,KPMGinIndia,meetingclient

expectationsaroundleadtimesandcustomizationisaprimeareaforthesectortogeneratevaluefromtechnologies.

“Clients’lead-timeexpectationsarebecomingshorterbythe

day,”saysBhatnagar.“Themarketrequiresmoreagile,responsiveproductioncapabilities—fromthesourcingofrawmaterialstothedownstreamsupplyofgoods.”

Asproductionspecificationsbecomemorecomplexandbespoke,manyorganizationsareturningtodigitalinterventions,systems,

processesandcontrolstobringcustomization,reliabilityandspeedintoproductioncycles.Theseupgradesshouldalsoextendto

operationalworkflowsinsupplychain,procurement,sales,andfinancefunctions,amongotherareas.

“Thevaluechainofindustrialmanufacturingisbeingconnectedandenhancedbydigitalfeatures,”saysBhatnagar.“Thishybridmodelismakingtheentirevaluechainfaster,informed,moredisciplinedandagileinrespondingtomarketneeds.”

Theseefficiencygainsarealsohelpingtoimprovetheenergy

efficiencyofproductionprocesses,accordingtoBhatnagar,for

instanceinreducingtheidletimeofmachinesorreducingbatch

cycletimesofcertainotherprocesses.Here,AIandmachinelearning(ML)areplayingcrucialrolesinfindingmoresustainableandgreen

operationalstrategiesformanufacturerstodeploy.

KPMGglobaltechreport—industrialmanufacturinginsights5

?2025CopyrightownedbyoneormoreoftheKPMGInternationalentities.KPMGInternationalentitiesprovidenoservicestoclients.Allrightsreserved.

Executivesummary

Keyfindings

Mfis’rtitive

Theroadaheadfor

industrialmanufacturing

IMhasabove-average

datamaturity

IMexcelsat

achievingAIROI

HowKPMGcanhelp

Methodology

ThesectorexcelsatachievingAIROI,butdisconnectspreventfurtherprogress

IndustrialmanufacturingistheleadingsectorinAIadoption.ItisoneofthethreesectorsmostlikelytobeatthemostmaturephaseofAIadoption,with34percentsecuringROIonseveraloftheirAIusecases.But,whiletherearepocketsofsuccess,thesectorneedstoaddresscriticalgapsthatpreventitfromaccessingthefullpotentialofAI,includingbutnotlimitedtoimprovingdata-basedpredictions,optimizingproducts,augmentinginnovation,enhancingproductivityandefficiency,andloweringcosts.

ProportionofsectorsatthehigheststageofAImaturity

28%29%

Tech

Retailandconsumerpackagedgoods

34%34%

34%27%

32%30%

LifesciencesIndustrialmanufacturing

Healthcare GovernmentFinancialServicesEnergy

Source:KPMGglobaltechreport2024

OneoftheprominentAIusecasesinthesectorisusingpredictivemaintenancetoenhanceequipmentreliability.Performance

diagnosticsAI,alongsidestrategicdatagenerationandstorage,allowsworkerstoanalyzereal-timedatafromvariousmachinecomponents.Thishelpsthemtomakeinformeddecisionsaboutequipment

functionality,performanceandreliability.Also,organizationsare

applyingAIandMLimagerecognitiontechnologiestoevaluate

thequalityofafinishedproducttoinforminterventionsthatwill

improvelaterbatches.FirmsarealsousingAIandMLtoupgradetheenvironmentalefficienciesoftheirequipmentasESGtargetsbecomemoreimportantacrossthesector.

AnotherproductiveusecaseishowthesectorisusingAItoaddresstalentshortages.Fourinfivemanufacturingexecutivessaythat

AIisfillingskillsgapsamongknowledgeworkers—gapsthathadpreviouslypresentedamajorchallenge.

Bhatnagaradvisesthat,asindustrialmanufacturersincorporateAI

andMLintotheirbusinessmodels,upskillingprogramsshouldtargetanalyticaldecision-makingandsciencetechnologyskills,aswellascreatingadata-centricculture.

KPMGglobaltechreport—industrialmanufacturinginsights6

?2025CopyrightownedbyoneormoreoftheKPMGInternationalentities.KPMGInternationalentitiesprovidenoservicestoclients.Allrightsreserved.

Executivesummary

Keyfindings

Mfis’rtitive

Theroadaheadfor

industrialmanufacturing

IMhasabove-average

datamaturity

IMexcelsat

achievingAIROI

HowKPMGcanhelp

Methodology

ofindustrial

80%

manufacturerssayAIissavingthemtime

andallowingthemtobemoreproductive

andfocuson

higher-valueactivities.

“So,factoryworkerscanmakebetterreal-timedecisionsonthe

groundbasedondatathat’sbeingthrownatthem,”saysBhatnagar.“[Factory]floorworkersshouldtaketheleadonthedecisionsthatAIandMLarenotreliableenoughtomakerightnow.BeforeAI,floor

workersmade30to40decisionsaday.Now,theyjustneedtofocuson,say,10extremelycritical,high-valuedecisions.”

IntermsofotherAIusecasesatplayinthesector,therapid

developmentofindustrialandprocesscontrolapplications

underpinnedbyAIpackages,low-costcomputerhardwareand

graphical-user-interfacetechnologyhasledtotheemergence

ofvirtualinstrumentation.Alsoknownas‘softsensors,’virtualinstrumentationactsasasubstituteforphysicalsensorsand

combinesreal-timedata,digitaltechandAI-backedmathematicalmodelstoestimateproductquality.

Thesesoftsensorsprovidemeasurementsatpointsinproductionlineswhereitisimpossibletoinstallaphysicalsensorduetoprohibitive

costsorharshoperationalconditions.Theycombinemultiplereal-timedatasourcesofprocessvariables—suchaswaterflowrates,temperatures,pressureandspeedoftravelonaconveyor—andconvertthemintoanumbertoforecastthequalityofaproduct.

Forinstance,ratherthanwaitingtoevaluatebatchqualityattheendoftheproductionprocess,onelargeintegratedsteelmanufacturingplantisdeployingsoftsensorsatallstagesofitsmanufacturing

cycle.Thisend-to-endmonitoringprovidesopportunitiesfor

proactiveinterventionsthatcanpreservethestabilityofproductionandminimizethepresenceofcontaminantsthatcouldcause

batchestoberejected.“Thissavestheplantfromwasting

productioncapacityandenergyonproducingoff-specmaterials,”saysBhatnagar.

Whiletheseadvancedengineeringsystemsanddesign

methodologiesarecrucialtoinnovation,theinfluenceoftheseAI

usecasesisoftenrestrictedbyconnectivitygapsbetweensystemsanddatasets,saysMartinKaestner,TechnologyLeader,IndustrialManufacturing,KPMGintheUS.

“ThetruepotentialofAIisrealizedwhendatafromdisparate

systems,suchascustomerrelationshipmanagementand

procurementplatforms,isaggregatedandanalyzedholistically,”

saysKaestner.“Thisisparticularlycriticalinsectorssuchas

aerospaceanddefense,whereorganizationsoftenpossesssomeofthebest-engineeredproducts,yettheystruggletoidentifycriticalgapsintheiroperationalframeworksbecauseofsilos.”

“TheinabilitytosynthesizedataacrossvariousplatformsinmodernmanufacturingenvironmentswillskewthevisibilityAImodelshaveofcriticalareassuchasglobalsupplychaindynamics,”Kaestneradds.

Withoutacohesivestrategytofullyintegrateandanalyzedatafrommultiplesources,companiesriskmissingoutonvaluableAI-poweredinsightsthatcoulddriveefficiency,reducecosts,andacceleratetimetomarket.

Intherealmofindustrialmanufacturing,theintegrationofadvancedengineeringsystemsanddesignmethodologiesisparamountforfosteringinnovationandenhancingoperationalefficiency.Whilethesectorhasbeenmeticulouslyworkingtoadoptthelatesttechnologies,includingartificialintelligence(AI)andextensivemachinelearningcapabilities,theseareoftensiloedwithinindividualsystemssuchascustomerrelationshipmanagementandprocurementplatforms.

MartinKaestner

TechnologyLeader,IndustrialManufacturingKPMGintheUS

KPMGglobaltechreport—industrialmanufacturinginsights7

Executivesummary

Keyfindings

Mfis’rtitive

inrilang

Methodology

HowKPMGcanhelp

IMhasabove-average

datamaturity

IMexcelsat

achievingAIROI

Whileithasabove-averagedatamaturity,theindustrycontinuestoholditselfto

highstandards

Eventhoughdatasiloesremainachallenge,strongdatafoundationssupportthesector’soverallAIprogress.Industrialmanufacturing

performsabovethecross-sectoraverageintheproportionofitsorganizationsthatareinourtoptwolevelsofdatamaturity.

“Theindustrialmanufacturingsectorhasbuiltarichdigital

architectureforitsdata,”saysBhatnagar.“Thesector’shigh-qualitydatamanagementprocessesandinfrastructuresuchassensors,

serversandcloudplatformshelpensurethattherightdataispulledfromtherightplacesattherightfrequency,formatandquality.”

Industrialmanufacturingisthesectormostlikelytociteimmaturedatamanagementasthetopfactorslowingtheirdigitaltransformationprogress.

KPMGglobaltechreport—industrialmanufacturinginsights8

Executivesummary

Keyfindings

Mfis’rtitive

inrilang

Methodology

HowKPMGcanhelp

IMhasabove-average

datamaturity

IMexcelsat

achievingAIROI

Proportionofexecutivesinthetoptwolevelsofmaturityindatamanagement

Datainvestments

Ensuringdatasysteminvestmentsalignwithprioritiesofallbusinessstakeholders

53%

56%

52%

56%

52%

57%

51%

58%

51%

58%

50%

54%

Datamonetization

Leveragingdataforcompetitiveadvantageinnewbusinessmodels

Datasecurity

Protectingdatathroughthesecurityofsystems,standardsandgovernanceprocesses

Datagovernance

Conductingauditstoaddressdataintegrityandcreatingframeworksthatprovideclearaccountability

Datainteroperability

Addressingdatasilosandpoordataintegration

Datascience

Usingreal-timeorpredictiveanalyticstoinformdecisions

AverageacrossallsectorsIndustrialmanufacturing

Howeffectiveareyourdataandanalyticsactivitiesinthefollowingareas?(Influential/embedded)

Source:KPMGglobaltechreport2024

Butratherthanbecomecomplacent,thesectorcontinuestohold

itselftohighstandardsandmaintainsambitiousdatagoals.Althoughitsaccesstodataishigh,industrialmanufacturingisalsothesectorwhoseorganizationsaremostlikelytobealerttothefactthat

furtherimprovementstothedatastrategywouldacceleratethepaceofdigitaltransformation.Forinstance,totakefulladvantageofgenerativeAI,industrialmanufacturerswillneedtoconstructareliable,trustworthydatainfrastructurethatiscustomizedtotheirbusinessneeds.

So,howcanthesector’sorganizationsmakesurethattheirdata

strategiesareenablersofprogress,ratherthanblockers?“Tomake

theirbusinessesmoreprofitableandcustomercentric,manufacturersneedtokeepincreasingtheirvisualizationcapabilitiesandworkflowstofacilitatedata-leddecision-makingandensuretheseinsightsreachtherightemployeesattherighttime,”saysBhatnagar.“Thereisstillworktobedonehere,buttheyareonthevergeofgettingthisright.”

Acrucialstepwillbetoprovidedashboardinterfacesthatpresentdatainsightsinaclearandintuitiveformat,sothatworkerscan

makedecisionsquickly.Thesuccessoftheseplatformswilldependheavilyondatagovernanceandinteroperabilitycapabilities,which,incomparisonwithothersectors,arebothkeyskillsforindustrial

manufacturing.Onbothcounts,industrialmanufacturingperformed

7percentagepointshigherthanthecross-sectoraverageof

51percent.

Interoperabilitywillbeanespeciallyimportantrequirementfor

trustworthydigitalinnovation,andparticularlywhenitcomestoAI.Tosustaintheirmomentumandmovedataathighspeedsandeveninrealtime,manufacturersmustfocuspartsoftheircybersecuritystrategiesonprotectinginternaldatatransfer.“Oneofthebiggestthingsmanufacturershavetoaddressisbuildingenoughsecurity

protectionaroundtheirinternalnetworkstoengineeringdesign

systems,”saysBhatnagar.“Especiallyforwhentheybegintosharedatainrealtimewiththeoutsideworld.”

KPMGglobaltechreport—industrialmanufacturinginsights9

Executivesummary

Keyfindings

Mfis’rtitive

Methodology

HowKPMGcanhelp

Theroadaheadfor

industrialmanufacturing

IMhasabove-average

datamaturity

IMexcelsat

achievingAIROI

Theroadaheadforindustrialmanufacturing

Asthemanufacturingsectorrespondsstrategicallytoshiftingsupplychaindynamicsandgrowingenvironmentaldemands,itsorganizationsshould:

Nurturetheproactiveandprogressivespiritthatispoweringtheirdigitaltransformationefforts.Theongoingevolutionofcybersecuritystrategiesshouldfeatureinitiativesthatenhancethesecurityprotectionofinternaldatanetworkspluggingintoengineeringdesignsystems,especiallyforreal-timedatasharing.

Improvevisualizationcapabilitiesandworkflowstohelpworkersmakebetterdecisionsbasedondatainsights.Andmakesurethattheseinsightsreachtherightemployeesattherighttimes.

InparallelwiththeriseofAIintheworkplace,preparetheworkforcebyupskillingfactoryworkerswithlearningprogramsthattargetanalyticaldecision-makingandsciencetechnologyskills.IMexecutivesaremoreinclinedthanmosttobelievethatGenAIwillboostproductivityandenhancecollaborationwithapositiveimpactinITandcreativejobs.1

Explorenewwaystoinnovatetomeetclients’expectationsofspeedandcustomization,andmakeoperationsmoreenergyefficient.By

capturingandcentralizingtheadhocdataprovidedbyconsumerswhentheysharetheirpreferencesandfeedbackoncertainproductfeatures,manufacturerscouldfindinsightsthatleadtonewproductlinesorrevenuestreams.Investinprocessesandsystemsthatelevatethevoiceofthecustomersothattheorganizationcangiveitstargetaudiencewhatitwants.

ToreallyoptimizetheuseofAIandnewtechnology,industrialmanufacturers

needtocombinewhattheydobest

(manufacturephysicalproducts)with

whatdigitaldoesbest(collectreal-time

dataandembedAI)todifferentiatetheirproductsandgainanewcompetitive

advantage.Itwillnotbeenoughtoadd

digitalfunctionalitytoanalogmachines—acompletere-imaginationisneeded.

CarmeloMariano

IndustrialManufacturingLeaderKPMGItaly

Manufacturingmightbeaheadofothersectorsindigitaltransformation,buttheneedtoinnovatedoesnotstophere.“Wewillcontinuetoseeaconstantpushforinnovation,”saysKPMGinIndia’sSaurabhBhatnagar.“Moretechnology,moreagility,moretailoringtosuitthecustomer,andfasterspeedto

delivery.Tomeettheseneeds,theindustrialmanufacturingsectorwillneedtorelyondigitalinterventionsevenmoreinfuture.”

1KPMGinUS

GettingaheadstartwithgenerativeAIinindustrialmanufacturing

,2023

KPMGglobaltechreport—industrialmanufacturinginsights10

Manufacturing’sproactiveIMexcelsat

andprogressivespiritachievingAIROI

Methodology