Issuu

The Dark Factory Era: Impact of Automation on Employment and Future Consumer Demand

Tejas

Repale1 , Prem

Birajdar

¹ Final Year M.Sc. (Computer Applications) Student, Department of Computer Applications, Pratibha College of Commerce and Computer Studies, Chinchwad, Pune, Maharashtra, India

² Final Year M.Sc. (Computer Applications) Student, Department of Computer Applications, Pratibha College of Commerce and Computer Studies, Chinchwad, Pune, Maharashtra, India

Abstract - The dark factory era is one of the most significant turning points in industrial history.Factories that once depended entirely on human hands now run around the clock without a single worker on the floor, and what seemed like science fiction powered by advances in robotics, artificial intelligence, and the Internet of Things. This paper focuses on two consequences of that shift that do not get nearly enough attention together: the mass displacement of manufacturing workers and theseriousthreatthisposes to consumer demand The argument built on a wide review of recent research, is that automation's gains for individual businesses come at a collective cost. As wage income shrinks across the economy, so does the purchasingpowerthatkeepsconsumermarketsrunning.

The article talks about possible ways to fix the problem. It talks about how people and machines can work together,how people can acquire moneywithouthaving to work, and how individuals can learn new skills to adapt. The major point, though, is a little worrisome. If nothing is done,factories willkeep manufacturing more and more products, but fewer and fewer people will be able to afford them.

Key Words: dark factory, worker displacement, consumer demand, wage income, purchasing power, human machine collaboration, reskilling.

1. INTRODUCTION

A dark factory is exactly what it sounds like a productionfacilitywherethelightscanstayoffbecause there is no one there who needs them (Gisi, 2024; Mathur, 2024). Everytaskonthefloor is handledby machines:roboticarms,autonomousguidedvehicles, and AI driven quality control systems that work continuously without breaks, errors, or supervision (Gisi, 2024). The idea may have once sounded far fetched, but it became real faster than most people expected.FanucinJapanandLEGOinDenmarkwere amongthefirstcompaniestomakeithappen,andthey

did so in the early 2000s, long before artificial intelligencehadbecomeamainstreamconversation.

AllofthisishappeningbecauseofIndustry4.0.Thisis basically the fourth big industrial revolution, and it uses things like big data, machine learning, cloud computing,andcyberphysicalsystemstochangethe way products are made (Sima et al., 2020). From a businesspointofview,itjustmakessense.Youspend lessonworkers,youroutputismoreconsistent,and the factory never has to stop running. But what happens to people in all of this is a different story. When workers get removed from the production process, the damage does not stop with them. The money that used to flow into their households stops flowing. Neighbourhoods and towns that were built aroundfactoryjobsstartfallingapart.Andthepeople whogotdisplacedrealisethattheskillstheyworked hardtobuildovermanyyearsaresimplynotwanted anymore(Kurt,2019;Spencer,2018).

Factory work has historically been one of the most reliable routes into the middle class, in wealthy and developingcountriesalike(Hepaktan&Şimşek,2022). Whenautomationclosesthatroute,thedamagegoes beyondindividualhardship.Itbeginstoeatawayatthe consumerbasethatmoderncapitalismdependson.If theproductivitygainsfromautomationflowmostlyto those who own the machines, and if workers are displacedfasterthannewjobsemergetoreplacethem, the result could be a structural gap between what economiescanproduceandwhatpeoplecanactually afford to buy (Spencer, 2018; Dellot & WallaceStephens,2017).

This paperexamines thatrelationshipthrough three questions:first,whatimpactaredarkfactorieshaving onmanufacturingemploymentwithintheIndustry4.0 framework?Second,howmightautomationdrivenjob loss threaten future consumer demand? And third, what policy and strategic responses have been proposedtoaddressthesepressures?

International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056

Volume: 13 Issue: 05 | May 2026 www.irjet.net p-ISSN: 23950072

2. Literature Review

2.1 The Dark Factory as an Industrial Phenomenon

The darkfactorydid notsimplyappearone dayasa finishedidea.Itevolvedslowlyandquietlyoutofthe automationmovementthatbegangatheringpaceinthe late twentieth century, accelerating as artificial intelligence,robotics,andsensortechnologiesbecame not onlymore powerfulbut also farmore affordable Gisi(2024)lookedathowthisalldevelopedandmade an interesting point the dark factory was never really a sudden invention. It was more like the unavoidableendresultofdecisionsmanufacturershad been making for decades, always trying to produce more, spend less, and keep quality steady. In the picture he paints, the factory of the future basically runs on its own. Workers are not gone entirely, but they are no longer on the floor doing physical tasks. Instead they are somewhere behind the scenes, keeping systems running, fixing problems, and designing the things that machines will then go and build.

Hepaktan and Şimşek (2022) put all of this in the contextofIndustry4.0andarguethatfullautomation is really what defines this era of industrial change. Theirreasoningmakessensewhenyouthinkaboutit. Markets today move fast and do not give much warning. A business that can shift quickly without havingtorestructureitsentireworkforcehasaserious edge over one that cannot. Automated systems give companies exactly that kind of flexibility. Mathur (2024) adds to this by pointing out that demand for fullyautonomousdarkfactorysetupshasbeengrowing consistently,mostlybecauseAIkeepsgettingsmarter andtheequipmentkeepsgettingcheaper.

George (2026) takes a different angle. Rather than lookingatthetechnologyortheeconomics,hefocused onwhatthistransitionisactuallydoingtorealworkers andindustries,specificallyinIndia.

Whathefindsisaparadoxthatiseasytomissifyou areonlylookingatproductivityfiguresandinvestment trends.Yes,robotsaretakingoverthephysicalwork. Buttheneedforhumanskillhasnotgoneaway ithas simplychangedshape.Companiesarelookingfordata analysts,AIsupervisors,andpeoplewhocanintegrate complexsystems.Theproblemisthattheseskillsare rare, and for someone who has spent twenty years workingwiththeirhandsonafactoryfloor,theroadto

acquiring them is steep, uncertain, and often poorly signposted.

2.2 Industry 4.0 and the Labor Market

Economists andsociologistshave beentalkingabout whatautomationdoestojobsforalongtime,andthe answersarerarelyclear.Inawide-rangingsystematic reviewpublishedinSustainability,Simaetal.(2020) bringtogetheralotofresearchonhowIndustry4.0is changingboththewaypeopledeveloptheirskillsand the way they shop. They see a pattern that keeps happening: automation does make new kinds of technical jobs, but it also tends to get rid of more routineandsemi-skilledjobs.Anumberofthingsaffect whether overall employment levels go up or down. These include how quickly new technologies are adopted,howmanyworkerscanactuallygetreskilling opportunities, and how flexible or rigid the labor marketisaroundthem.

Kurt(2019)writesinProcediaComputerScienceabout Industry4.0fromadifferentangle:notjustasastory abouttechnology,butasastoryaboutpower.Hesays thatwhatisgoingoninworkplacestodayismorethan just machines taking over jobs. It changes the way employers and workers interact in a big way. When managementcanuseautomationasatoolorthreaten touseit,workerslosetheirpower.Theirabilitytofight back and ask for better pay or working conditions slowlygoesaway.Kurtwarnsthattheresultisthatthe productivitygainsfromautomationgoupratherthan beingsharedwidelyamongworkers.

In The Digital Factory, Altenried (2022) goes even further with this idea, saying that the logic of the factoryfloorhasspreadbeyondthewallsofthefactory to the gigeconomy, platform work, andremote jobs. These new types of work don't look anything like traditional manufacturing at first glance. Altenried's point, though, is that the basic dynamic is the same: centralized control, broken-up labor, and limited workerpower.Hesaysthatworkersintheseso-called "neweconomy"jobsareoftenworseoffthanfactory workers because they have less job security, fewer protections,andnotmuchofavoiceasagroup.

2.4 Consumer Demand and the Automation Paradox

Honestly, most of the research out there focuses on howautomationtakesawayjobs.Fairenough thatis anobviousandseriousconcern.Butwhatdoesnotget talkedaboutnearlyasmuchiswhatautomationdoes

International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056

Volume: 13 Issue: 05 | May 2026 www.irjet.net p-ISSN: 23950072

toconsumerdemand.Andsomeresearchersthinkthat is actually the bigger problem.Here is the basic idea behind what scholars call the "automation paradox." Businessesinvestinautomationtocutcostsandboost productivity.Makessensefromacompanyperspective. Butifenoughbusinessesdothisandenoughworkers losetheirjobsortakepaycutsasaresult,thosesame workers have less money in their pockets to spend. Lessspendingmeanslessdemand.Lessdemandmeans businesses sell less. And suddenly the productivity gains that looked so attractive start working against theveryeconomytheyweresupposedtohelp.Simaet al. (2020) make this point clearly. They show that whenjoblosseslinkedtoIndustry4.0 whichbecome widespreadenoughtoreduce incomeacrosslargepart of the population, consumer behavior shifts in ways thatactuallydamageseconomicgrowth.

Theparadoxisprettystraightforwardwhenyouthink about it a decisionthatmakes perfectsense for one company can become a serious problem when the wholeeconomystartsdoingthesamething.It'seasyfor any one company to decide to replace expensive human workers with automated systems: costs go down, output goes up, and profits follow. But when enoughcompaniesmakethissamesmartchoiceatthe sametime,badthingsstarttohappenonabiggerscale. Wageincomebeginstofallthroughouttheeconomy.

Peoplewhohavelosttheirjobsorhadtheirhourscut have less money to spend. And slowly, the money neededtobuythesamethingsthatautomatedfactories make so well starts to go down. This is a modern versionofaproblemthateconomistshavebeentrying tosolveforhundredsofyears.Theunderconsumption problemthatMalthusandKeyneswroteaboutintheir owntimesisbackinatwenty-first-centuryindustrial form.

2.5 Research Gap

Althoughtheimpactofautomationonemploymenthas been strictly examined and the automation paradox hasstartedtotakeacademicinterest,a gappersistsin thecurrentliterature.Themajorityofstudiesregard jobdisplacementanddemandcontractionasdistinct issuestobeexaminedindependently.Thereisnosingle frameworkthatshowsthewholechainofevents:from thedecisionatthefirmleveltoautomate,throughthe lossofjobsacrosstheeconomy,tothelossofconsumer purchasing power that keeps the industries that are driving automation going. This paper directly addressesthatgapbyconnectingjoblosstoadropin

consumer demand in a single conceptual model and lookingatwhatkindsofpolicychangescouldstopthe cyclebeforeitbecomesself-reinforcing.

3. Research Methodology

3.1

Research Design

Thisstudyusesamixed-methoddescriptive-analytical approach notbecauseitisthemostfashionablechoice, but because the problem genuinely demands it. Automationissimultaneouslyastorytoldinnumbers andastorytoldinlives,andanyresearchdesignthat captures only one of those dimensions will miss somethingimportant.

The qualitative side of the study works through thematic analysis reading across a wide body of published academic literature, industry reports, and real-world case studies to identify the patterns, tensions, and arguments that keep surfacing. The quantitative side draws on employment statistics, automation adoption rates, and consumer spending datafromsomeofthemostauthoritativeinternational sources available: the International Labour Organization,theOECD,theMcKinseyGlobalInstitute, andtheWorldEconomicForum.Thesedatasetstellus howmanyjobshavebeenlost,howfastautomationis spreading,andhowhouseholdspendingisshiftingthe kind of hard evidence that gives an argument its backbone.

But numbers, on their own, have a way of making humanproblemslooktidierthantheyactuallyare.A statisticaboutjobdisplacementdoesnottellyouwhat it feels like to be a fifty-year-old factory worker in a townwheretheplantjustswitchedtofullautomation, orhowlongitrealisticallytakestoretrainforarolein a completely different industry. That is where the qualitativeanalysisdoesitswork fillinginthetexture, the contradiction, and the human complexity that aggregate data smooths over. The two strands are strongertogetherthaneitherwouldbealone.

3.2 Research Approach

The study employs a deductive methodology consistently. The paperdoes notstartwith rawdata and build up to a theory. Instead it starts with establishedtheoreticalframeworksandusesthemto look at the evidence. Three frameworks are very important. The first is the Industry 4.0 framework, which gives us the technological and organizational backgroundweneedtounderstandhowandwhydark

International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056

Volume: 13 Issue: 05 | May 2026 www.irjet.net p-ISSN: 23950072

factories are popping up. The second is the Gartner HypeCycle,whichhelpsusunderstandwhereweare in the development of automation by showing the difference between unrealistic hopes and the more realistic productivity gains. The third is the underconsumption theory, which economists like Malthus and later Keynes came up with. It helps us understandhowdemandcanbelowerthanproductive capacitywhenwageincomeislow.

3.3 Data Sources

Thestudyusesfourdistincttypesofsources,eachof themtocontributinguniqueinsights.Theresearchis based on both theory and practice with academic journalslikeSustainability,ProcediaComputerScience, andNewTechnology,WorkandEmploymentserving asthefoundation.McKinsey,theILO,theOECD,andthe World Economic Forum all publish industry reports that give us the hard numbers on job trends, how quickly automation is being adopted, and how consumerspendingischanging.

The case studies of Fanuc in Japan and LEGO in Denmark shows us the argument by examining companiesthathaveundergonethetransitiontodark factories, focusing on its practical implications, especially for the workers involved. Gisi (2024), Altenried (2022), and Dellot and Wallace-Stephens (2017) wrote longer more detailed works that give theoreticalargumentsthespacetheydeserve.

3.4 Data Analysis Method

Thisstudyusesthreeanalyticalmethodstogether.First oneisthematicanalysisthatisusedtounderstandthe overall picture of the literature. It focus is on identifyingpatternsandtensionsthatappearsacross differentstudiesandgroupingthemintoclearthemes thatsupporttheargument.

Thesecondmethodiscomparativeanalysis.Theidea hereissimple tounderstandwherewearetoday,it helps to look at where we have been before. Automationisnotthefirsttechnologytoshakeupthe workforce. Mechanization, electrification, and computerization all did the same thing in their own time, andworkersandpolicymakers backthenwere justasworriedandjustashopefulaspeoplearetoday. Lookingatwhatactuallyhappenedduringthoseearlier shifts helps separate what is genuinely new about lights-out manufacturing from what is just history repeatingitself.

Thethirdmethodiscasestudyanalysis,whichisreally justaboutgettingbacktobasics.Ratherthanstayingat the theoretical level, this method looks at real companies,realworkers,andrealoutcomes.Whatdid specificbusinessesactuallydowhentheyautomated? Whathappenedtothepeoplewhoworkedthere?And whatcanwelearnfromthoseexperiencesthatmight beusefulgoingforward?Thesekindsofquestionsthat casestudyanalysisisdesignedtoanswer.

3.5 Ethical Considerations

Theinformationusedinthisstudycomesfrompublic sources such as peer reviewed journals, academic papers,andreportsfromwellknownorganizations.No researchwasconductedinvolvingrealpeople,andno private or sensitive data was accessed at any point. Every source has been properly cited and credited throughoutthepaperinlinewithacademicstandards.

4. Conceptual Framework: The Automation–Demand Paradox Model

Thecentralcontributionofthispaperisaconceptual framework that traces the chain of consequences runningfromfirm-levelautomationdecisionsthrough toeconomy-widedemandcontraction.Thelogicofthe modelrunsasfollows.

At the level of the individual firm, the decision to automate is straightforward and rational. Replacing human workers with robotic systems and AI driven processe reduces wage costs, increases output , eliminatesdowntimeassociatedwithhumanerror,and allows production to continue around the clock. Productivity rises. Unit costs fall. Profit margins improve. For a single company operating in a competitivemarket,thesearecompellingadvantages, anditisdifficulttoarguethatanyrationalfirmwould chooseotherwise.

Theproblememergeswhenthisindividuallyrational decision is made simultaneously by enough firms acrossenoughsectors.Atthatpoint,whatwasgoodfor eachcompanyindividuallybeginstoproduceoutcomes thatarebadfortheeconomycollectively.Wageincome theprimarysourceofspendingpowerforthemajority of households begins to decline across the economy. Workerswhohavebeendisplaceddonotimmediately findequivalentemploymentelsewhere.Newtechnical rolesarecreated,butnotinsufficientnumbers,notin the same locations, and not accessible to workers whose skills were built around the routines that machines have now taken over. The result is a

International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056

Volume: 13 Issue: 05 | May 2026 www.irjet.net p-ISSN: 23950072

widening gap between productive capacity which automationpushesupward andconsumerpurchasing power whichdisplacementpushesdownward.

This is the automation demand paradox. High productioncapacitycoexistswithweakeningconsumer demand. Factories can make more than ever before, butthepopulationofpeopleabletoaffordwhatthose factoriesproduceisshrinking.Leftunaddressed,this dynamic becomes self-reinforcing: falling demand reduces revenue, which reduces investment, which slowsgrowth,whichfurtherweakensemployment completingacyclethatnoindividualfirm,actingalone, haseithertheincentiveortheabilitytobreak.

The framework positions deliberate policy interventionasthenecessarycircuitbreaker.Without structural responses including reskilling programs, redistributivemechanisms,andnewformsofeconomic governancetheproductivitygainsofautomationwill continue to concentrate at the top of the income distribution while the consumer base that sustains industrialcapitalismgraduallyerodes.

5. Analysis and Discussion

5.1 How Dark Factories Are Reshaping Manufacturing Employment

5.1.1TheScaleandPatternofDisplacement

Automation is not some future threat that manufacturingworkersneedtoworryaboutsomeday. Itisalreadyhereanditseffectsarealreadybeingfelt. TheMcKinseyGlobalInstituteestimatesthatasmany as375millionworkersworldwidecouldbeforcedto changejobsentirelyby2030.

Manufacturing workers are particularly vulnerable because the work they have always done is exactly what machines are now good at. Tasks like putting partstogether,runningmachines,checkingfordefects, and moving materials around the factory floor are straightforward and repetitive enough that robotic systems can handle them quickly, consistently, and cheaply.

Whatreallysetsthisroundofautomationapartfrom earlieronesishowfaritreaches.Inthepast,machines mostly took over the simplest and most boring jobs, whichmeantthatworkerswithadecentlevelofskill could still find their place. That has changed. The AI systems being used today can recognize patterns, figure things out when the situation is unclear, and

adjust on the fly. Not long ago people assumed only humanscoulddothosethings.HepaktanandŞimşek (2022)makeanimportantobservationhere.Workers inthemiddleoftheskillrange,whoactuallymakeup the biggest part of the manufacturing workforce in most countries, are now facing the same kind of job lossesthatlowerskilledworkersexperiencedinearlier roundsofautomation.Nobodyinthefactoryisassafe astheyusedtobe.

5.1.2 Job Creation Versus Job Destruction

Itisimportanttoacknowledgethatautomationdoes not only destroy jobs it creates them too. The operation and maintenance of complex automated systemsrequiresdataanalysts,AIsupervisors,robotics engineers, and systems integrators. Companies that automate successfully tend to grow, and growth generatesemploymentofitsownkind.George(2026) notes that in India, the dark factory transition is generatingrealdemandforanewclassoftechnically skilled workers, even as it eliminates the roles that previouslyemployedfarlargernumbers.

Thecriticalquestion,however,isnotwhethernewjobs arebeingcreated,butwhethertheyarebeingcreated fastenough,intherightplaces,andaccessibletothe rightpeople.Theevidenceonthisisconsiderablyless encouraging.Simaetal(2020)findthatwhileIndustry 4.0doesgeneratenewcategoriesofemploymentthe netshorttermeffectontotalemploymentisnegative in most manufacturing. The jobs created tend to requireskillsthatdisplacedworkersdonotcurrently possess.Theytendtobeconcentratedinurbancentres andtechnologyhubs,whiledisplacementfallshardest on workers in regions whose economies are built aroundtraditionalmanufacturing.Andtheytendtobe createdonatimescalethatisfarslowerthanthepace ofdisplacementitself.

5.1.3 The Skills Gap

Running through the employment debate is the problem of the skills gap. The workers most at risk fromautomationarepreciselythoseleastlikelytohave accesstothereskillingopportunitiesthatmighthelp themtransitionintonewroles.AsKurt(2019)argues, the power dynamics of the modern workplace make thisworse:whenemployerscancrediblythreatento automateanyfunctionthatworkerspushbackagainst, theleverageneededtodemandbettertraining,more securecontracts,orlongertransitionsupportsimplyis notthere.DellotandWallace-Stephens(2017)makea

International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056

Volume: 13 Issue: 05 | May 2026 www.irjet.net p-ISSN: 23950072

similarpoint,notingthattheworkersmostexposedto automation are also those with the least access to continuingeducation,thefewestfinancialresourcesto sustainthemselvesthroughperiodsofretraining,and themostfragilesafetynetstofallbackonifretraining fails.

5.2 Automation, Purchasing Power, and the Demand Crisis

5.2.1

The Erosion of Wage Income

The connection betweenjob losses from automation andfallingconsumerdemandisprettystraightforward. It runs through wages for most families, especially thoseintheworkingandmiddleclasses,wagesarenot justonesourceofincomeamongmany.Theyarethe main source. Everything else, including how much peoplespend,howmuchtheyborrow,andhowthey planforthefuture,isbuiltaroundthatpaycheck.

Whenwagesfallorjobsstartfeelinglesssecure,people pullbackonspending.Andthisdoesnotjusthappento thepeoplewhoactuallylosetheirjobs.Italsohappens tothepeoplewhoarestillemployedbutareworried theycouldbenext.Spencer(2018)lookscloselyatthis pattern and makes an interesting point. The underconsumption problem that economists noticed during the early days of industrial capitalism is showingupagainintheageofAI.Thebasiclogichas not changed at all. When the gains from increased productivitygomostlytobusinessownersratherthan beingsharedwithworkers,peoplesimplydonothave enoughmoneytobuywhatisbeingproduced.

The factories that worked so hard to become more efficient end up creating a problem for themselves. Thereisplentybeingproducedbutnotenoughpeople withenoughmoneytobuyit.Inthatsensethesuccess ofautomationstartsworkingagainsttheveryeconomy itwassupposedtostrengthen.

5.2.2 The Paradox at the Macroeconomic

Level

This is a difficult problem to solve through normal market forces and here is why. Every individual company that chooses to automate is making a perfectly reasonable decision. If one business holds back while its competitors are automating, it will simply become too expensive to keep up and eventually go under. The problem is that when all businessesmakethissamereasonabledecisionatthe sametime,thewagesthatordinarypeopledependon

tocovertheirbasicneedsstartdisappearingacrossthe board.

Sima et al. (2020) argue that this is not just a temporarydipthatmarketswillnaturallycorrectover time.Itissomethingdeeper.Itisafundamentalshiftin therelationshipbetweenhowthingsareproducedand how they are consumed. Normal market ups and downscanusuallysortthemselvesoutgivenenough time. But this is a structural change in how people behave as consumers and it needs a deliberate and plannedresponseratherthanjustwaitingforthingsto balanceoutontheirown.

5.3 Policy and Strategic Responses

5.3.1

Reskilling and Workforce Transition Programs

Ifyouaskmostpeoplewhatshouldbedoneaboutjob lossesfromautomationtheansweryouwillhearmost oftenisreskilling.Theideaissimpleenough.Takethe workers who have lost their jobs and give them the trainingtheyneedtomoveintothenewkindsofwork thatautomationisopeningup.Whenyoufirsthearit thatactuallysoundslikeaprettysolidplan.Ifworkers are losing jobs because their skills no longer match what employers need, then helping them build new skillsseemsliketheobviousanswer.Simaetal.(2020) support this view, arguing that developing human capital is an essential part of any serious long term responsetoIndustry4.0.

In practice however things are more complicated. Reskillingprogramsdoworkbutonlyundertheright conditions.Theytendtobemosteffectivewhenthey are properly funded, run over a long period of time, andcloselytiedtowhatemployersareactuallyhiring for. The problem is that many government run programsfallshortonallthreeofthesefronts.Timing matterstoo.Theseprogramsworkbestwhenworkers canaccessthemearly,beforefinancialpressurestarts buildingup andlife circumstances make itharder to committoretraining.TheOECD'sAIPrinciplesmakea similarpoint,callingonmembergovernmentstoinvest intransitionframeworksthatanticipatedisplacement andplanaheadforitratherthanscramblingtorespond afterthedamageisalreadydone.

5.3.2 Universal Basic Income and Redistributive Models

A more radical response that is being talked about moreandmoreseriouslyistheuniversalbasicincome

International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056

Volume: 13 Issue: 05 | May 2026 www.irjet.net p-ISSN: 23950072

(UBI).Thisisaregular,unconditionalcashtransferto allcitizensthatisenoughtomeettheirbasicneedsand is paid for by taxes on the productivity gains of automation.Inthiscase,theappealofUBIisclearand strong. Automation puts productive capacity in the hands of capital owners and takes away the wage income that workers rely on. A way to redistribute some of that productivity back to the general public would help fight poverty, keep consumer demand strong,andgiveworkerswhohavelosttheirjobsthe financial breathing room they need to retrain or transition at a sustainable pace. Dellot and WallaceStephens (2017) describe this as part of a broader concept they call demand recycling. The idea is straightforwardthebenefitsthatautomationgenerates needtoflowbackthroughtheeconomyinawaythat keepspeopleabletoaffordthings,ratherthanpilingup inthehandsofafewwhileeveryoneelsestrugglesto getby.

5.3.3 Human–Robot Collaboration and Inclusive Governance

Governancedoesnotgettalkedaboutenoughinthis debatebutitreallyshould.Becauseattheendofthe daysomeonehastodecidehowautomationgetsused and who it benefits. The OECD AI Principles tried to pushthingsintherightdirection.Countriesthatsigned up agreed that AI and automation should be transparent, accountable and actually focused on peopleslivesratherthanjustonmakingbusinessesrun cheaperandfaster.

Theproblemisthatgoodprinciples onpaperdo not automaticallytranslateintorealchangeontheground. Ifworkers,communitiesandgovernmentsdonothave a genuine seat at the table when these decisions are beingmade,businesseswillkeepcallingtheshots.And businessesnaturallyprioritizetheirowninterests.The automationparadoxalreadyshoweduswhathappens when everyone does what is good for themselves without anyone thinking about the bigger picture. Withoutrealgovernancestructuresinplacethecosts ofautomationwillkeepfallingonthepeoplewhocan leastaffordthem.

6. Future Scope

This study has covered a lot of ground but honestly thereisstillsomuchmoretoexplore.Theresearchon lights-out manufacturing and inequality is still relativelyyoungandtherearesomeprettyimportant questionsthatnobodyhasfullyansweredyet.

Developing Countries

Almosteverythingwrittenonthistopicfocusesonthe UnitedStatesandWesternEurope.Butwhataboutthe restoftheworld?CountriesinAsia,Africa,andLatin Americahavehistoricallyusedmanufacturingasaway toliftpeopleoutofpoverty.Ifautomationmakescheap labor less attractive to global companies, those countriescould beinserioustrouble.Thisisanarea thatreallyneedsmoreattentionfromresearchers.

Gender and Race

Some studies have touched on how automation hits womenandminorityworkersharderthanothersbut this has not been explored nearly enough. Future research needs to look more carefully at how race, gender,age,andeducationlevelallcombinetoshape whogetsleftbehindwhenfactoriesautomate.Looking atjustoneofthesefactorsatatimemissesalotofthe realpicture.

Mental Health and Social Impact

Most research counts job losses and measures wage drops. But losing a job does more than reduce your income. It affects how people feel about themselves, howtheyrelatetotheirfamilies,andhowconnected they feel to their communities. The human cost of lights-outmanufacturinggoeswaybeyondwhatshows up in economic data and future studies should start takingthatseriously.

Long Term Economic Consequences

Theautomationparadoxraisesaquestionthatnobody has fully answered yet. If enough workers lose their jobsorseetheirwagesfall,willtherestillbeenough consumer demand to keep the economy growing? Futureresearchneedstotakealongerandharderlook atwhatwidespreadlights-outmanufacturingactually does to economic growth, consumer spending, and wealthdistributionovertime.

Different Technologies Different Effects

Mostresearch treats automationasone bigcategory butinrealitydifferenttechnologiesdoverydifferent thingstodifferentworkers.Industrialrobots,AIquality controlsystems,andautonomouslogisticsnetworksdo notallhavethesameeffects.

International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056

Volume: 13 Issue: 05 | May 2026 www.irjet.net p-ISSN: 23950072

7. Conclusion

The dark factory represents one of the most consequentialdevelopmentsinthehistoryofindustrial production. By removing human workers from the productionfloorentirely,itcompletesatrajectorythat began with the first industrial revolution and accelerates sharply under the conditions of Industry 4.0 where artificial intelligence, robotics, and networkeddigitalsystemshavemadefullautomation not only technically feasible but economically irresistible for firms operating in competitive global markets.

This paper has argued that the consequences of this transition extend well beyond the workers directly displaced.Whenautomationeliminateswageincome onasufficientscale,itbeginstoerodetheconsumer demandthatsustainsindustrialcapitalismitself.Thisis the automation–demand paradox: the same technological progress that makes factories more productivethreatens,at themacroeconomiclevel,to hollow out the market of buyers that gives that productivity its economic meaning. It is a modern restatement of a problem that Malthus and Keynes identifiedinearliereras butitstwenty-first-century form is faster, broader, and less likelyto self-correct throughnormalmarketadjustment.

Thepaper'sconceptualframeworktracesthischainof consequencesfromtherationalfirm-leveldecisionto automate, through economy-wide wage income decline, to structural demand contraction and positions deliberate policy intervention as the necessary circuit breaker. Reskilling programs, universal basic income models, human–robot collaboration frameworks, and inclusive governance structureseachofferpartialbutmeaningfulresponses. None of them alone is sufficient. Taken together, as part of a coordinated, anticipatory, and equitable approachtomanagingthedarkfactorytransition,they representthedifferencebetweenanautomationfuture that is broadly shared and one that is dangerously concentrated.

Thestakesofgettingthisrightarehigh.Aneconomic system in which machines produce everything and people can afford to buy nothing is not a system in equilibrium itisasystemincrisis.Theresearchand policy agenda must therefore move faster than the technology itself, building the frameworks for distribution, retraining, and governance that will determinewhetherthedarkfactoryeraisremembered

as a moment of shared prosperity or of structural rupture.

8. References

Altenried, M. (2022). The digital factory:The human laborofautomation.UniversityofChicagoPress.

Dellot, B., & Wallace-Stephens, F. (2017). The age of automation: Artificial intelligence, robotics and the future of low-skilled work. RSA Actionand Research Centre.

George, A. (2026). Dark factories and the future of workinIndia:Navigatingthehumancostoflights-out manufacturing.JournalofIndustrialTransitions,4(1), 22–39.

Gisi, P. J. (2024). Dark factory and the future of manufacturing: Building an adaptive, sustainable, smartproductionsystem.ProductivityPress.

Hepaktan,C.E.,&Şimşek,G.(2022).Industry4.0and its effects on the labour market. Ekonomi Bilimleri Dergisi,14(1),1–19.

Kurt, R. (2019). Industry 4.0 in terms of industrial relations and its impacts on labour life. Procedia ComputerScience,158,590–601.

Mathur, A. (2024). Autonomous manufacturing systemsandtheriseoffullydarkproductionfacilities. International Journal of Advanced Manufacturing Technology,131(3),1145–1160.

Sima,V.,Gheorghe,I.G.,Subic,J.,&Nancu,D.(2020). InfluencesoftheIndustry4.0revolutiononthehuman capital development and consumer behaviour: A systematicreview.Sustainability,12(10),4035.

Spencer,D.A.(2018).Fearandhopeinanageofmass automation: Debating the future of work. New Technology,WorkandEmployment,33(1),1–12.