The wide­spread adop­tion of tech­no­lo­gies focus­ing on inter­con­nectiv­ity and auto­ma­tion in indus­tri­al envir­on­ments, as well as recent advances in AI tech­no­logy, are at the root of the ever-increas­ing amount of data being gen­er­ated and pro­cessed on a daily basis. We will take an in-depth look at the stag­ger­ing num­bers used to describe these unima­gin­able amounts of data, as well as the soci­et­al chal­lenges and implic­a­tions of the ongo­ing tech­no­lo­gic­al paradigm shift.

Table of Contents

1.      Stan­isław Lem’s „Grow­ing Tor­rent of Information“

2.      2.5 Quin­til­lion Bytes Gen­er­ated Each Day

3.      Smart Factor­ies and Their Insa­ti­able Hun­ger for Data

4.      The Social Implic­a­tions of AI and Automation

5.      The Future of Data Processing

6.      Con­clu­sion

1. Stan­isław Lem’s „Grow­ing Tor­rent of Information“

„Recog­niz­ing the qual­ity of inform­a­tion, not accept­ing com­pletely irrel­ev­ant advert­ising, sec­ond­ary inform­a­tion that is simply unne­ces­sary for a human being, is a neces­sary con­di­tion for ‘stay­ing afloat’ amidst a grow­ing tor­rent of inform­a­tion.

This quote from renowned Pol­ish sci-fi writer and philo­soph­er Stan­isław Lem dates back to the essay Inform­acja o inform­acji from the early 1990s and is more applic­able than ever before. In today’s inform­a­tion age we are down­right flooded with data, the amount of inform­a­tion gen­er­ated every single day is stag­ger­ing. Over 90% of the inform­a­tion ever cre­ated by human­ity — books, music, data­base entries or whatever else you can think of — was cre­ated in the last dec­ade and exists almost entirely in a digit­al form. Fil­ter­ing and pro­cessing this data and dis­cern­ing valu­able data from junk data is a task requir­ing pro­cessing power and data stor­age far exceed­ing the cap­ab­il­it­ies of the human brain.

2. 2.5 Quin­til­lion Bytes Gen­er­ated Each Day

Human­it­ies data foot­print grows lar­ger and lar­ger every day, includ­ing pic­tures of cute kit­tens, You­Tube, Face­book, Ins­tagram and Twit­ter posts, videos and everything else. Skype alone has three bil­lion minutes of calls per day. Five bil­lion Snapchat videos and pho­tos are shared every 24 hours, 333.2 bil­lion emails are sent in the same peri­od. Every minute people spend one mil­lion US dol­lars online. Around five bil­lion people — or a good 65% of the world pop­u­la­tion — cur­rently inter­act with digit­al inform­a­tion every day. This pro­por­tion could increase to 75% by 2025.

The total quant­ity of data pro­cessed glob­ally glob­ally in the year 2018 already amoun­ted to 33 zetta­bytes — accord­ing to the Data Age 2025 study con­duc­ted by IDC Glob­al Data Sphere, sponsored Seag­ate, a big play­er in the stor­age industry. One zetta­byte writ­ten out as a num­ber is a 1 fol­lowed by 21 zer­oes, one quin­til­lion writ­ten out is a 1 fol­lowed by 18 zer­oes. To truly grasp the enorm­ous amount of data cre­ated daily, con­sider this fig­ure: 2.5 quin­til­lion bytes, writ­ten out in full looks like this:


Stor­ing this immense amount of data phys­ic­ally would require 33 bil­lion phys­ic­al drives with a capa­city of 1TB each. Accord­ing to fur­ther research con­duc­ted by IDC the amount of data pro­cessed glob­ally per year will reach an astound­ing 175 zetta­bytes by 2025.

3. Smart Factor­ies and Their Insa­ti­able Hun­ger for Data

Data is the found­a­tion of almost all integ­ral parts of our every­day life. Our most crit­ic­al infra­struc­ture, such as traffic, trans­port, energy, tele­com­mu­nic­a­tion and emer­gency ser­vices, simply can­not func­tion without data. Smart homes, the Inter­net of Things, autonom­ous driv­ing and all oth­er cur­rent tech­no­logy trends are highly data intens­ive and are driv­ing factors in the expec­ted expo­nen­tial increase in data usage in the near future.
The indus­tri­al sec­tor, a cru­cial top­ic for us here at Codelab, is argu­ably one of the major cata­lysts in this devel­op­ment. Indus­tri­al devices and sys­tems gen­er­ate massive amounts of inform­a­tion, even when con­figured in the less data hungry report by excep­tion mode, which has been an industry stand­ard for many years. The data col­lec­ted is typ­ic­ally related to large machines and includes sig­nals about the cur­rent work­ing con­di­tions of a machine, like tem­per­at­ure sig­na­tures, engine speed or oxy­gen con­tent for example. Any giv­en para­met­er is rep­res­en­ted as an elec­tric­al sig­nal, which can be stored, trans­ferred and then ana­lysed, in order to find reg­u­lar­it­ies, incon­sist­en­cies, sim­il­ar­it­ies and to dia­gnose prob­lems and anti­cip­ate fail­ures. The aggreg­ated data is then used to optim­ize the effi­ciency of pro­duc­tion­al pro­cesses and to determ­ine when to pro­ceed with pre­vent­ive main­ten­ance meas­ures in order to avoid pro­longed pro­duc­tions stops and thus increase profitability.

The rise of data-driv­en smart factories

The digit­iz­a­tion study by IDC and Seag­ate pre­dicts that there will be around 150 bil­lion con­nec­ted devices on our plan­et by 2025. We can assume, with a high degree of prob­ab­il­ity, that a sig­ni­fic­ant per­cent­age of those devices will be deployed in the indus­tri­al sec­tor. The fact that all those devices run soft­ware using excit­ing, state of the art tech­no­lo­gies and frame­works are the reas­ons why Codelab decided early on to focus on devel­op­ing soft­ware for the Indus­tri­al Inter­net of Things (IIoT).
Every year the trend of digit­iz­a­tion in the indus­tri­al sec­tor becomes more appar­ent and sig­ni­fic­ant. In our mod­ern, digit­al­ized world, more and more factor­ies are being gradu­ally upgraded and trans­formed into smart factor­ies which dif­fer from tra­di­tion­al non-smart factor­ies in fun­da­ment­al aspects. One key dif­fer­ence is that their oper­a­tion is com­pletely data-driv­en. Smart factor­ies are inten­ded to be semi-vir­tu­al spaces with inter­con­nec­ted machines con­tinu­ously col­lect data via count­less sensors.

The data being accu­mu­lated and incor­por­ated into the pro­duc­tion pro­cess can be split into two groups: extern­al and intern­al data. The intern­al data includes all intern­al com­pany data as well as data gathered from machines inside the fact­ory. As com­pan­ies have gradu­ally star­ted col­lect­ing data about how their products are used, espe­cially tech­nic­al products, the amount of extern­al data col­lec­ted is con­sist­ently increas­ing. These days the products them­selves usu­ally gath­er the data, some­thing that had been con­sidered the holy grail of the man­u­fac­tur­ing industry for some time. Com­bin­ing this extern­al data with the intern­al data res­ul­ted in sub­stan­tial leaps in max­im­iz­ing effi­ciency in the pro­duc­tion process.

With this new data driv­en approach, man­u­fac­tur­ers can sig­ni­fic­antly extend product lifespan and reduce product fail­ure by eval­u­at­ing its per­form­ance in the field based on beha­viour pro­file data from end users. Such an approach also has the poten­tial to spark a fun­da­ment­al change in the con­text of the fur­ther cus­tom­iz­a­tion of pro­duc­tion assortment.

The Role of Data Cen­ters in the Factor­ies of the Future

Data cen­ters are indis­pens­able when it comes to the pro­cessing of these enorm­ous amounts of data and as such data cen­ters are an integ­ral part of both pub­lic and private clouds. Data cen­ters are essen­tial for factor­ies of the future as they are used for provid­ing decent­ral­ized stor­age and archiv­ing, ser­vice deliv­ery, extens­ive ana­lyt­ics and com­mand and con­trol. As more and more pro­cesses are being digit­ized, both inside and out­side of factor­ies, none of them would be able to work as a single sep­ar­ated entity. This is due to a fun­da­ment­al shift in how glob­al pro­duc­tion and sup­ply net­works oper­ate. Wheth­er a fact­ory is respons­ible for core oper­a­tions or just exec­ut­ive ele­ments, it has to com­mu­nic­ate with many dif­fer­ent extern­al sys­tems. The top­ic of data pro­cessing tran­scends far bey­ond the auto­ma­tion of tra­di­tion­al man­u­fac­tur­ing and indus­tri­al prac­tices, like large-scale machine-to-machine com­mu­nic­a­tion (M2M or the Inter­net of Things (IoT). Extens­ive digit­al data ana­lys­is offers many great oppor­tun­it­ies to fur­ther increase its use­ful­ness and could be used to imple­ment a more organ­ic approach — man­aging energy resources and raw mater­i­als in a more cir­cu­lar way and on a lar­ger scale. Digit­iz­a­tion and auto­mat­iz­a­tion paired with big data ana­lys­is are part of the unstop­pable auto­ma­tion age, which comes with both pre­vi­ously unima­gin­able oppor­tun­it­ies and con­sequen­tial implications.

4. The Social Implic­a­tions of AI and Automation

Anoth­er import­ant aspect of the devel­op­ment of the factor­ies of the future is the pro­cessing of  these large amounts of data, which would not only require sim­ul­tan­eous effect­ive com­mu­nic­a­tion and data ana­lys­is but also act­ive, stra­tegic plan­ning pre­dict­ing trends. Tra­di­tion­ally these func­tions were occu­pied by human work­ers, some­thing that may be about to change soon, as the amounts of data are cur­rently way too large to be ana­lysed effect­ively by human entit­ies. Infer­ence speed is anoth­er area in which com­puters and AIs are far super­i­or to humans. If we extra­pol­ate, it is safe to assume that the wide-spread use of AI sys­tems in the factor­ies of the future is inev­it­able due to the sheer amount of data being amassed. Stud­ies con­duc­ted in this field show many jobs tra­di­tion­ally per­formed by humans are already per­formed by autonom­ous sys­tems. Num­bers are rising every year, lead­ing to fun­da­ment­al changes in the labour mar­ket. For fur­ther read­ing on this top­ic please refer to Mind Chil­dren: The Future of Robot and Human Intel­li­gence by Hans Moravec and spe­cific­ally the ‘The Land­scape of human com­pet­ence’ graphic.

5. The Future of Data Processing

It is still a mat­ter of debate wheth­er it is pos­sible to con­struct an Arti­fi­cial Gen­er­al Intel­li­gence (AGI) that would be cap­able of repro­du­cing gen­er­al­ized human cog­nit­ive abil­it­ies in soft­ware. This the­or­et­ic­al AGI would be able to find solu­tions for com­plex unfa­mil­i­ar prob­lems, much like a human would. The devel­op­ment in this field is still in its infancy and it could take dec­ades or even cen­tur­ies until AGIs are widely adop­ted. There are many great resources widely avail­able for more in-depth inform­a­tion on this top­ic, for example Life 3.0: Being Human in the Age of Arti­fi­cial Intel­li­gence by Max Teg­mark. Cur­rently AGIs are not cru­cial for the fur­ther auto­ma­tion of factor­ies, which cur­rently rely on Arti­fi­cial Intel­li­gence (AI). Depend­ing on future tech­nic­al advance­ment AGIs might nev­er be suc­cess­fully deployed, or AI sys­tems being trained on a much nar­row­er field of expert­ise, might turn out to be more than enough to meet the require­ments of the factor­ies of the future. Please con­sult this art­icle about the dis­tinc­tion between AGI and AI. Wheth­er future indus­tri­al atom­iz­a­tion will be driv­en by AGI or AI tech­no­logy is cur­rently impossible to pre­dict, but one thing we know for sure is that we as humans will need more extens­ive sup­port from our semi-autonom­ous or fully autonom­ous sys­tems and their arti­fi­cial counterparts.

6. Con­clu­sion

Will the factor­ies of the future be com­pletely auto­mated and able to func­tion without any human work­ers? While pre­dic­tions with a 100% accur­acy regard­ing this ques­tion are clearly impossible, prom­in­ent trends are evid­ent and there is no turn­ing back to out­dated pro­duc­tion mod­els. This leads us to the more gen­er­al con­clu­sion that data is one of the corner­stones of The Fourth Indus­tri­al Revolu­tion (4IR) or Industry 4.0, some­thing we explored pre­vi­ously in a sep­ar­ate art­icle which cov­ers the import­ance of Industry 4.0 for the man­u­fac­tur­ing industry.

Data is chan­ging the world. The flood of data is becom­ing increas­ingly more ubi­quit­ous and influ­ences more and more aspects of our lives. Whatever the future holds in the con­text of tech­no­lo­gic­al devel­op­ment, the chal­lenge lies not only in hand­ling the aston­ish­ing amount of data but also the sub­sequent drastic social changes. We can only be cer­tain of one thing: we cer­tainly live in the age of information.

Tomasz Brzo­zowski — Author

Tomasz is a team man­ager at Codelab with an extens­ive in soft­ware devel­op­ment and pro­ject man­age­ment. He has extens­ive exper­i­ence in man­aging teams for indus­tri­al devel­op­ment pro­jects, lever­aging his expert­ise in cut­ting-edge tech­no­logy to deliv­er suc­cess­ful out­comes for cli­ents. With a curi­ous and ana­lyt­ic­al mind, Tomasz is always on the lookout for emer­ging trends and enjoys shar­ing his insights through writ­ing. Out­side of work, he val­ues time with his fam­ily and stays cur­rent on the latest tech­no­logy through gam­ing and oth­er interests.

Mads Carstensen - Con­sult­ing, text revi­sion, addi­tion­al research

Mads is the Mar­ket­ing Man­ager at Codelab respons­ible for all sales related mar­ket­ing activ­it­ies. Since he was a child he has been fas­cin­ated by IT and the Eng­lish lan­guage. After hav­ing spent a year abroad in Northamp­ton­shire, UK, Mads began his career in mar­ket­ing in Ber­lin in 2009 at a prom­in­ent start-up in Ber­lin. He has since been hon­ing his mar­ket­ing skills in the IT and indus­tri­al fields. When he’s not work­ing, Mads enjoys going to con­certs and spend­ing time fid­dling with com­puter parts, DAWs, DJ con­trol­lers or guitars.

About Codelab

Codelab is a val­ued part­ner for soft­ware devel­op­ment and digit­al­iz­a­tion for com­pan­ies in the auto­mot­ive, indus­tri­al, and enter­prise sec­tors. The core busi­ness of Codelab focuses on ser­vices and solu­tions in the realm of embed­ded sys­tems, IoT, full-stack devel­op­ment, and con­sult­ing. The 250 employ­ees are loc­ated in Ber­lin and two tech­no­logy cen­ters in Szcze­cin and Wro­claw. Codelab is a sub­si­di­ary of the pub­licly-traded Beta Sys­tems Soft­ware AG and has been oper­at­ing in over 40 coun­tries for 35 years.

Find out more about how Codelab is help­ing busi­nesses tackle digit­al­iz­a­tion and devel­op­ment bot­tle­necks.

Any remarks, sug­ges­tions, thoughts, feed­back or cor­rec­tions can be sent to mads.carstensen(at)


Marr, Bern­ard. “How Much Data Do We Cre­ate Every Day? The Mind-Blow­ing Stats Every­one Should Read.” For­bes, For­bes Magazine, 21 May 2018,

“How Much Data Is Cre­ated Every Day? —”,

“How Much Data Is Cre­ated Every Day? —”,

“The Flood of Data — The Future Trans­form­a­tion — Mer­ck Group.” Mer­ck Group,


“How Much Data Is Cre­ated Every Day? —”,

“How Much Data Is Cre­ated Every Day? —”,

IDC sponsored by Seag­ate. “Data Age 2025: The Evol­u­tion of Data to Life-Crit­ic­al.” Seag­ate,

Tech Coun­cil, For­bes. “AI vs AGI: What’s the Dif­fer­ence?” For­bes, For­bes Magazine, 17 Sept. 2018,

McKin­sey & Com­pany. “Where Machines Could Replace Humans—and Where They Can­’t (Yet).”,

Teg­mark, Max. Life 3.0: Being Human in the Age of Arti­fi­cial Intel­li­gence. Knopf Doubleday Pub­lish­ing Group, 2017.

Moravec, Hans. Mind Chil­dren: The Future of Robot and Human Intel­li­gence. Har­vard Uni­ver­sity Press, 1988.

Statista. “Num­ber of daily act­ive Twit­ter users world­wide from 1st quarter 2010 to 2nd quarter 2021.” 3 Aug. 2021,

Hoot­suite. “Ins­tagram by the num­bers: Stats, demo­graph­ics & fun facts.” 22 Apr. 2021,

You­Tube. “Press.” 2021,

Face­book. “Face­book Reports First Quarter 2021 Res­ults.” 28 Apr. 2021,

Snapchat. “Fast Facts.” 2021,