SIKORA, A. VOLUME: 23 |NUMBER: 2 |2025 |JUNE
Resea ch A icle
HOW GENERATIVE AI AND THE INTELLIGENT
INDUSTRIAL INTERNET OF THINGS
COMPLEMENT EACH OTHER
Axel SIKORA 1,∗
1Ins i u e o Reliable Embedded Sys ems and Communica ion Elec onics, O enbu g Uni e si y,
Bads . 24, D77652 O enbu g, Ge many
axel.siko a@hs-o enbu g.de
*Co esponding au ho : Axel Siko a; axel.siko a@hs-o enbu g.de
DOI: 10.15598/aeee. 23i2.250310
A icle his o y: Recei ed Ma 17, 2025; Re ised Ap 22, 2025; Accep ed Ap 26, 2025; Published Jun 30, 2025.
This is an open access a icle unde he BY-CC license.
Abs ac . Gene a i e modeling is an a i icial in el-
ligence (AI) echnique o gene a e syn he ic a i ac s
om analyzing aining examples and om lea ning
hei pa e ns and dis ibu ion. Gene a i e AI (GenAI)
uses gene a i e modeling and ad ances in Deep Lea n-
ing o p oduce di e se con en a scale by u ilizing ex-
is ing da a. Whe eas adi ionally GenAI is mos ly us-
ing media con en s, such as ex , g aphics, audio, and
ideo, i can addi ionally be used also o da a om he
(Indus ial) In e ne o Things. This a icle p o ides
a sys ema ic o e iew on he mani old di e en p ac i-
cal oppo uni ies and challenges GenAI b ings o he
IIoT. I also p esen s selec ed examples om he au-
ho ’s esea ch wi h his eams. In doing so, i co -
e s he ele ance o GenAI o he comple e li ecycle o
IIoT, om design and de elopmen , o e es ing o de-
ploymen . This pape summa izes a keyno e p esen a-
ion om he 13 h In e na ional Con e ence on G een
and Human In o ma ion Technology (ICGHIT) in Jan-
ua y 2025 held in Nha T ang, Vie nam.
Keywo ds
Gene a i e A i icial In elligence, In e ne o
Things, Indus ial In e ne o Things, Gene -
a i e In e ne o Things.
1. In oduc ion
I goes back o Schumpe e ’s idea ha inno a ions can
be desc ibed as new combina ions o p e-exis ing ideas
and echnologies [1]. A i icial In elligence (AI) and
he In e ne o Things (IoT) can be seen as one suc-
cess ul example o such “Inno a ion by Combina ion”
o wo mega ends, which a e ueling each o he and
a e leading o an accele a ing pace o inno a ion. The
IoT connec s any hing, anywhe e, any ime [2]. Thus,
i p o ides a pla o m o a uly pe asi e and in el-
ligen en i onmen . Since a couple o yea s, AI and
mos no ably Edge AI simul aneously make use o and
enhance he Indus ial IoT (IIoT) [3].
Gene a i e modeling is an a i icial in elligence (AI)
echnique o gene a e syn he ic a i ac s om analyz-
ing aining examples and om lea ning hei pa -
e ns and dis ibu ion. Gene a i e AI (GenAI) uses
gene a i e modeling and ad ances in Deep Lea ning
(DL) o p oduce di e se con en a scale by u ilizing
exis ing da a. These models o en gene a e ou pu
in esponse o speci ic p omp s. Gene a i e AI sys-
ems lea n he unde lying pa e ns and s uc u es o
hei aining da a, enabling hem o c ea e new da a.
Whe eas GenAI is mos ly using media con en s, such
as ex , g aphics, audio, and ideo, i can addi ionally
be used also o da a om he (Indus ial) IoT. Thus,
he Gene a i e In e ne o Things (GIoT) is eme ging
and holds immense po en ial o e olu ionize a ious
aspec s o socie y, enabling mo e e icien and in elli-
gen IoT applica ions.
©2025 ADVANCES IN ELECTRICAL AND ELECTRONIC ENGINEERING 83
SIKORA, A. VOLUME: 23 |NUMBER: 2 |2025 |JUNE
This a icle p o ides a sys ema ic o e iew on he
di e en p ac ical oppo uni ies and challenges GenAI
b ings o he IIoT. I is s uc u ed as ollows: sec ion 2
gi es a sho o e iew o he di e en GenAI echnolo-
gies, being ele an o he IIoT, whe e sec ion 3 lis s
possible use cases om he IIoT. A e ha , sec ion
4 explains a ew examples om he au ho ’s esea ch
a his ins i u ions, be o e concluding wi h a selec ion
o cu en esea ch di ec ions in sec ion 5 and a sho
summa y in sec ion 6.
2. Gene a i e AI Technologies
GenAI is de ined, and commonly dis inguished om
o he ypes o AI, by i s capabili y o “gene a e new
con en ” [4]. In he ypical case o Gene a o Ad e sa -
ial Ne wo k (GANs), GenAI uses wo neu al ne wo ks:
a gene a o and a disc imina o (c . Fig. 1). Thus, i
gene a es syn he ic a i ac s by analyzing aining ex-
amples (be i ca s, dogs, o IIoT da a), lea ning hei
pa e ns and dis ibu ion and hen c ea ing ealis ic
acsimiles. The disc imina o hen akes he eal ex-
amples om he da ase and he ake ones gene a ed
by he gene a o and ies o classi y hem as ei he
eal o ake. Based on his classi ica ion, i lea ns o
ge be e a disc imina ing images in he nex ound.
A he same ime, he gene a o lea ns how well he
gene a ed acsimiles ooled he disc imina o and im-
p o es he c ea ion o acsimiles in he nex ound.
GenAI is no new, i is only un il ecen ly ha la ge-
scale gene a i e models exempli ied by La ge Language
Models (LLMs) (e.g., GPT, LLaMA, and Gemini) and
Mul imodal Gene a i e Models (e.g., GPT-4V, DALL-
E, and S able Di usion) ha e made he b eak h ough
[6]. The e is a selec ion o se e al GenAI me hods be-
ing used, i.e. o GIoT applica ions [5,7]:
•Gene a i e Ad e sa ial Ne wo ks (GANs) a e
maybe he mos p e alen GenAI echnique be-
ing used oday in IoT da a syn hesis, consis ing o
gene a o and disc imina o ne wo ks [5, 7]. The
gene a o ne wo k aims o gene a e new da a by
lea ning eal da a dis ibu ion, while he disc im-
ina o ne wo k aims o dis inguish syn he ic da a
om eal da a. The wo ne wo ks a e ained
oge he in in e ac i e and compe i i e manne s,
esul ing in con inuous enhancemen o syn hesis
pe o mance.
•Va ia ional Au oencode s (VAEs) consis o he
encode and decode ne wo ks, whe e he encode
ne wo k comp esses he inpu da a o a la en ep-
esen a ion and he decode ne wo k lea ns o e-
cons uc syn he ic da a ha closely aligns wi h
he o iginal dis ibu ion [7].
Fig. 1: Gene alized P ocess Flow o Gene a i e AI.
•Gene a i e Di usion Models (GDMs) a e gene a-
i e models eme ging wi h he s a e-o - he-a pe -
o mance o image syn hesis. They consis o o -
wa d di usion and denoising p ocesses inspi ed by
non-equilib ium he modynamics heo y [7].
•Geome ic DL (GDL) ies o unde s and, in e -
p e , and desc ibe AI models in e ms o geome ic
p inciples [7].
•Flow-based Gene a i e Models (FGM) can ans-
o m inpu da a dis ibu ions om simple o com-
plex h ough a se ies o di e en iable and in e -
ible ans o ma ions ha a e implemen ed as neu-
al ne wo ks [5].
3. Use Cases o GenAI in he
IIoT
GenAI is a game-change o he IIoT, o e ing capa-
bili ies ha enhance e iciency, educe cos s, and d i e
inno a ion. By ex ending a ailable me hods o p e-
dic i e analy ics, eal- ime simula ions, and in elligen
au oma ion, GenAI is shaping he u u e o indus ial
ope a ions in p o ound ways. As indus ies inc eas-
ingly adop IIoT, GenAI will likely be pi o al in ensu -
ing sma e , sa e , and mo e sus ainable p ac ices.
The e is a mul i ude o di e en uses cases o he
GenAI in he IIoT. In he ollowing, a sho o e iew
is gi en oge he wi h some abs ac examples b inging
pi o al bene i s ac oss he en i e IoT pipeline, encom-
passing da a gene a ion, da a p ocessing, in e acing
wi h IIoT de ices, and IIoT sys em de elopmen and
e alua ion. These bene i s a e ele an o all di e en
IIoT applica ion domains, e.g. au onomous ehicles,
obo ics, heal h ca e, and many mo e [6].
3.1. Enhanced Da a Analy ics and
Insigh s
IIoT de ices gene a e as amoun s o da a. GenAI
can
A.1 syn hesize complex da a and c ea e meaning ul
summa ies o pa e ns om aw senso da a, making
i easie o de i e ac ionable insigh s.
©2025 ADVANCES IN ELECTRICAL AND ELECTRONIC ENGINEERING 84
SIKORA, A. VOLUME: 23 |NUMBER: 2 |2025 |JUNE
A.2 analyze ends and anomalies by gene a ing sim-
ula ions o p edic ions, helping o an icipa e sys em
beha io s o de ec issues ea ly.
A.3 c ea e new da a ep esen a ions and ill gaps in
incomple e da ase s o imp o e da a quali y.
A.4 gene a e digi al wins and high- ideli y i ual
models o indus ial p ocesses, p o iding de ailed in-
sigh s and p edic ions.
3.2. Au onomous and Adap i e
Decision-Making
GenAI can enable IIoT de ices o:
B.1 gene a e dynamic a chi ec u es, whe e gene a-
i e models c ea e on- he- ly solu ions o unexpec ed
challenges, e.g. supply chain dis up ions o p oduc ion
bo lenecks.
B.2 simula e scena ios, so ha i ual en i onmen s
can be gene a ed o es di e en IoT esponses unde
a ious condi ions, enhancing eal-wo ld deploymen
eliabili y.
3.3. Enhanced P edic i e
Main enance
GenAI models excel a analysing complex, mul i a ia e
da a om IIoT senso s o:
C.1 p edic equipmen ailu es by gene a ing simu-
la ed ailu e pa e ns o p o ice ea ly wa nings abou
machine y needing main enance and deli e subs an-
ial cos sa ings by p e en ing unplanned down ime.
C.2 op imize main enance schedules and gene a e e -
icien main enance plans, minimizing down ime.
C.3 simula e wha -i scena ios by p edic i e simula-
ions o help o ganiza ions unde s and he long- e m
impac o a ious ope a ional s a egies.
In his sense, ca ego y “C. Enhanced P edic i e
Main enance” can be unde s ood as a special case o
ca ego y “A. Enhanced Da a Analy ics and Insigh s”
and as an ou come o ca ego y “B. Au onomous and
Adap i e Decision-Making”.
3.4. Enhanced Secu i y
Cybe secu i y is pa amoun o IIoT. GenAI can con-
ibu e by:
D.1 c ea ing syn he ic da a o ain AI models (e.g.
o anomaly de ec ion) wi hou exposing sensi i e eal-
wo ld da a, p ese ing p i acy.
D.2 simula ing cybe a ack scena ios and po en ial
ulne abili ies, as well as es ing IoT de enses agains
po en ial h ea s.
D.3 gene a ing adap i e secu i y p o ocols and
p oposing eal- ime, cus om solu ions o mi iga e
h ea s and o sa egua d agains ulne abili ies.
3.5. E icien Resou ce Managemen
Fo IIoT sys ems managing esou ces, GenAI can:
E.1 model he in luence o di e en esou ce alloca-
ion s a egies on sys em pe o mance.
E.2 op imize ope a ions and gene a e e icien sched-
ules o ou es o esou ce use.
E.3 simula e u u e demands and p edic and gene -
a e plans o balance supply and demand dynamically.
The managed esou ces can be esou ces like ene gy,
a ic, o wa e , bu also he IIoT ne wo ks i sel .
3.6. Enabling C ea i i y in IIoT
Applica ions
GenAI can open doo s o inno a i e applica ions by:
F.1 c ea ing new de ice unc ionali ies and ea u es
based on speci ic en i onmen s.
F.2 enhancing sus ainabili y ini ia i es by gene a ing
solu ions o educing emissions and was e in indus ial
p ocesses.
F.3 gene a ing syn he ic en i onmen s o es ing, so
ha IIoT de elope s can simula e di e se scena ios o
enhance obus ness be o e deploymen .
3.7. Pe sonalized Use Expe iences
GenAI can use IIoT da a o:
G.1 gene a e cus omized in e ac ions, so ha de ices
can lea n use p e e ences and c ea e pe sonalized e-
sponses o se ings.
G.2 imp o e decision suppo and gene a e ac ion-
able insigh s and ecommenda ions o complex indus-
ial ope a ions.
G.3 design new se ices and sugges li es yle en-
hancemen s based on usage pa e ns.
3.8. Na u al Language In e aces
In eg a ing GenAI can allows IIoT sys ems o:
H.1 imp o e oice and ex in e ac ion. so ha de-
ices can p o ide sma , nuanced and con e sa ional
©2025 ADVANCES IN ELECTRICAL AND ELECTRONIC ENGINEERING 85
SIKORA, A. VOLUME: 23 |NUMBER: 2 |2025 |JUNE
Fig. 2: Sys ema ic O e iew o Use Cases and Applica ions o GenAI & IoT.
esponses.
H.2 c ea e mul ilingual and con ex -awa e in e ac-
ions and enhance accessibili y and use sa is ac ion.
4. Use Cases o GenAI in he
IIoT
This chap e shows some examples om selec ed
p ojec s om he au ho ’s eams a O enbu g Uni-
e si y∗and a Hahn-Schicka d Associa ion o Applied
Resea ch†.
4.1. Senso Design
As desc ibed in ca ego y A.4, deep lea ning and GenAI
can be used o gene a e digi al wins and high- ideli y
i ual models o indus ial p ocesses, p o iding de-
ailed insigh s and p edic ions.
In [8] and [9], a no el indi ec pho oacous ic senso
(PAS) has been de eloped ha uses deep lea ning ech-
niques (Fig. 2). S udies we e ca ied ou in con olled
se ings. As a esul , he senso ’s epea abili y and he
in luence o empe a u e and humidi y on he mic o-
phone ou pu ol age p o es ha deep lea ning mod-
els along he pipeline shown in Fig. 3 can e icien ly
be used o accu a ely desc ibe he senso ’s beha iou .
The indings demons a e he senso ’s consis en cha -
ac e is ics a e he pos -p ocessing s age.
∗h ps://i esk.hs-o enbu g.de/en
†h ps://www.hahn-schicka d.de/en
Fig. 3: Pho o o PAS senso [9].
4.2. Sys em Op imiza ion
In [10], an enhanced Angle o A i al (AoA) p edic-
ion me hod is p esen ed, which uses neu al ne wo ks
and he P ima y and Adjacen An ennas Rep esen a-
ion (PAAR) ans o ma ion. PAAR le e ages o a-
ional symme y in segmen ed an enna da a, o ans-
o m he da a in o a o a ion-in a ian o m. Thus,
PAAR imp o es p edic i e accu acy and s abili y, es-
pecially wi h limi ed aining da a. Expe imen s wi h
wo Digi al Video B oadcas ing - Te es ial (DVB-T)
da ase s we e conduc ed and e alua ed ou di e en
neu al ne wo k models, each a ying in pa ame e size.
The esul s demons a e ha he PAAR me hod
clea ly ou pe o ms he adi ional unp ocessed ap-
p oach, which we e e o as he plain app oach, in
©2025 ADVANCES IN ELECTRICAL AND ELECTRONIC ENGINEERING 86
SIKORA, A. VOLUME: 23 |NUMBER: 2 |2025 |JUNE
Fig. 4: Block diag am o he ins umen a ion and con ol wo k-
low o a no el indi ec PAS senso [8].
Fig. 5: Machine lea ning concep s used o p edic i e main e-
nance o bea ings [11].
da a-limi ed scena ios, educing he mean absolu e an-
gula e o (MAAE) by up o 40%.
Howe e , wi h ex ensi e aining da a, he plain ap-
p oach can su pass PAAR due o e o p opaga ion.
The s udy demons a es ha he PAAR me hod e ec-
i ely enhances AoA p edic ion, especially wi h spa se
aining da a.
4.3. P edic i e Main enance
GenAI models can analyse complex, mul i a ia e da a
om IIoT senso s o op imize main enance schedules
and gene a e he e icien main enance plans, minimiz-
ing unscheduled down imes o machines caused by ou -
ages o machine componen s in highly au oma ed p o-
duc ion lines, as lis ed in ca ego y C.2.
Conside ing machine ools such as g inding ma-
chines, he bea ing inside o spindles is one o he mos
c i ical componen s. Fig. 5 p o ides an o e iew o
Machine lea ning concep s used o p edic i e main e-
nance o bea ings.
The pape also p esen s he p edic ion o emain-
ing use ul li e, which is impo an o es ima ing he
p oduc i e use o a componen be o e a po en ial ail-
u e, op imizing he eplacemen cos s and minimizing
down ime. The a chi ec u e is depic ed in Fig. 4, e-
sul s a e shown in [12–14].
Fig. 6: The model and he ans e lea ning app oach o he
classi ica ion and RUL pa o he p oposed p edic i e
main enance solu ion [12].
4.4. IIoT Secu i y
Two use cases show he po en ial e iciency o GenAI
o he secu i y o IIoT sys ems, as an icipa ed in ca -
ego y D.
(1) Recen ly, he numbe o connec ed de ices
apidly g ows, hus ad e sa ies ha e mo e oppo u-
ni ies o gain access o IoT de ices and use hem o
launch wha is called la ge-scale a acks. Wi h he
apid p oli e a ion o In e ne o Things (IoT) de ices,
he need o e icien and e ec i e In usion De ec ion
Sys em (IDS) ailo ed o IoT en i onmen s has be-
come inc easingly pa amoun . Fo some yea s now,
comple e Secu i y In o ma ion and E en Managemen
(SIEM) sys ems ha e also been in use, which comp e-
hensi ely combine as many sui able echnologies (such
as in usion de ec ion and p e en ion, asse manage-
men , log analysis) as possible.
Secu i y In o ma ion and E en Managemen
(SIEM) sys ems a e a combina ion o di e en ca e-
go ies: Secu i y In o ma ion Managemen (SIM) and
Secu i y E en Managemen (SEM). SIEM echnology
enables he eal- ime analysis o secu i y ala ms gene -
a ed by ne wo k componen s o applica ions. By ana-
lyzing log in o ma ion, cohe en epo s can be gene -
a ed ha can also be used o compliance pu poses.
[15] explo es a ious echniques employed in con em-
po a y IoT IDS, including adi ional signa u e-based
app oaches like Sno and B o/Zeek, as well as eme g-
ing deep lea ning-based me hods.
[16] p o ides an o e iew on echniques and da ase s
used in he s udied wo ks, discuss he challenges o
using ML, DL and Fede a ed Lea ning (FL) o IoT
cybe secu i y.
©2025 ADVANCES IN ELECTRICAL AND ELECTRONIC ENGINEERING 87

SIKORA, A. VOLUME: 23 |NUMBER: 2 |2025 |JUNE
Fig. 7: A chi ec u e o he p oposed SIEM sys em om Kis e p ojec . I ea u es an au oma ed analys o simpli ied and cos -
e icien ope a ions also o small and medium size en e p ises [17].
The p ojec KISTE‡is di ing in o he di ec ion o
FL o anomaly de ec ion in he IIoT by in eg a ing
edge p ocessing wi h SIEM co ela ion o enable an
au oma ed analys o simpli ied ope a ions [17]. I
p oposes a no el amewo k o de ec ing anomalies in
IIoT ne wo ks by combining ede a ed lea ning (FL)
wi h a cus omized SIEM solu ion. The a chi ec u e
collec s eleme y and log da a om edge de ices, pe -
o ming local p ep ocessing and low-le el analysis be-
o e agg ega ing anomaly de ec ion models on an FL
se e hos ed wi hin a cen al secu i y moni o ing and
inciden de ec ion componen . De ec ed anomalies a e
co ela ed wi h ale s gene a ed by he SIEM moni-
o ing he co e IT ne wo k, p o iding a uni ied and
comp ehensi e iew o po en ial h ea s (c . Fig. 8).
(2) GANs ha e ea ned signi ican a en ion in a i-
ous domains due o hei gene a i e model’s compelling
abili y o gene a e ealis ic examples p obably d awn
om sample dis ibu ion. Image secu i y includes he
p o ec ion o digi al images om unau ho ized access,
modi ica ion, o dis ibu ion. This equi es a gua an-
ee o image p i acy, in eg i y, and au hen ici y o p o-
hibi hem om being exploi ed by malicious a acks.
GANs can also be u ilized o imp o ing image secu-
i y by exploi ing i s gene a ion abili y in enc yp ion,
s eganog aphy, and p i acy-p ese ing echniques.
‡h ps://kis e-p ojec .in o/ (p ojec websi e a ailable in Ge -
man only)
The su ey pape [18] e iews GANs-based image
secu i y echniques p o iding a sys ema ic o e iew o
cu en li e a u e and compa ing he ole o GANs in
image enc yp ion, image s eganog aphy, and p i acy
p ese ing om mul iple dimensions. Addi ionally, i
ou lines u u e esea ch di ec ions o u he explo e
he po en ial o GANs in add essing p i acy and image
secu i y conce ns.
4.5. Resou ce Op imiza ion
Resou ces can be modeled and op imized by GenAI, as
desc ibed in ca ego y E.1. The e iciency o blockchain
ne wo ks is one o such examples, as such ne wo ks
especially su e om scalabili y issues which hinde s
in eg a ion wi h IoT. Consequen ly, solu ions o im-
p o e blockchain scalabili y by minimizing he com-
pu a ional complexi y o consensus algo i hms o by
op imizing blockchain s o age equi emen s, ha e e-
cei ed a en ion. I.e, he ine iciencies o i s in e -pee
communica ion mus also be add essed. In his con-
ex , [19] p o ides a su ey on Ne wo k Op imiza ion
Techniques o Blockchain Sys ems.
One example [20] p oposes o le e age cloud e-
sou ces o s o ing blocks wi hin he chain using pa -
icle op imiza ion and gene ic algo i hms. An im-
p o ed hyb id a chi ec u e design uses con aine iza-
ion o c ea e a side chain on a og node o he de-
ices connec ed o i and an Ad anced Time- a ian
©2025 ADVANCES IN ELECTRICAL AND ELECTRONIC ENGINEERING 88
SIKORA, A. VOLUME: 23 |NUMBER: 2 |2025 |JUNE
Mul i-Objec i e Pa icle Swa m Op imiza ion Algo-
i hm (AT-MOPSO) o de e mine he op imal num-
be o blocks o be ans e ed o he cloud o s o -
age. This algo i hm uses ime- a ian weigh s o
he eloci y o he pa icle swa m op imiza ion and
he non-domina ed so ing and mu a ion schemes om
Non-domina ed So ing Gene ic Algo i hm (NSGA-
III). The p oposed AT-MOPSO showed signi ican ly
be e esul s han o he s a e o he a algo i hms
wi h ega ds o cloud s o age cos and que y p obabil-
i y. Impo an ly, he app oach also imp o ed ene gy
e iciency by 52%.
5. Fu u e Resea ch Di ec ions
F om he iewpoin o he au ho , amongs many o he
esea ch opics, h ee a e especially ele an and in e -
es ing:
•inc ease he e iciency o GenAI o IIoT especially
in combina ion wi h EdgeAI, so ha ope a ions
can be execu ed locally wi hou he o e head o
ull da a exchange and wi hou comp omising p i-
acy,
•iden i y no el applicabili y o GenAI o IIoT.
This could be done by „gene a ing“ and imple-
men ing no el use cases h ough he GenAI-based
Schumpe e -combina ion o exis ing echnologies,
and
•op imize he GenAI app oaches and go e en u -
he in he sys em modelling.
6. Summa y
This sho keyno e pape p o ides a sys ema ic
o e iew o he po en ial use cases and applica ions
o GenAI o he Indus ial In e ne o Things. I also
showcases some selec ed esea ch p ojec s om he au-
ho ’s eams wi h p omising and o wa d-looking e-
sul s.
I will be in e es ing o con inue his esea ch jou -
ney, o iden i y u he use cases and o imp o e he
exis ing app oaches.
Acknowledgmen
The au ho is ex emely g a e ul o collabo a e wi h
such g ea co-au ho s and pe sonali ies wi h excellen
b ains in his eams. He is hank ul o all he in ense
discussions, ui ul ideas, and objec i e-o ien ed e -
o s.
The esul s would no ha e been possible wi hou he
gene ous unding o he espec i e minis ies, agencies,
and dono s.
We acknowledge suppo by he Open Access Publica-
ion Fund o he O enbu g Uni e si y o Applied Sci-
ences.
Re e ences
[1] HANAPPI, H., E. HANAPPI-EGGER. New
Combina ions :Taking Schumpe e ’s concep se-
ious. Munich Pe sonal RePEc A chi e. 2004.
h ps://mp a.ub.uni-muenchen.de/28396/.
[2] SIKORA, A. Wi eless p o ocols o massi e IoT:
S anda d o scalable ne wo ks. Elek onik In e -
na ional. 2020, pp. 14-17.
[3] SHARMA, P., e al. Deep Lea ning in Resou ce
and Da a Cons ained Edge Compu ing Sys ems.
Machine Lea ning o Cybe Physical Sys ems.
2020, pp. 43-51. DOI: 10.1007/978 3 662-62746-
4_5.
[4] DE SILVA, D., e al. Oppo uni ies and Chal-
lenges o Gene a i e A i icial In elligence: Re-
sea ch, Educa ion, Indus y Engagemen , and So-
cial Impac . IEEE Indus ial Elec onics Mag-
azine. 2025, ol. 19, no. 1, pp. 30-45.
DOI: 10.1109/MIE.2024.3382962.
[5] JOVANOVIĆ, M., M. CAMPBELL. Gene a i e
A i icial In elligence: T ends and P ospec s .
Compu e . 2022, ol. 55, no. 10, pp. 107-112.
DOI: 10.1109/MC.2022.3192720.
[6] WANG, X., e al. The In e ne o Things in
he E a o Gene a i e AI: Vision and Challenges.
IEEE In e ne Compu ing. 2024, ol. 28, no. 5, pp.
57-64. DOI: 10.1109/MIC.2024.3443169.
[7] WEN, J., e al. F om Gene a i e AI o Gene -
a i e In e ne o Things: Fundamen als, F ame-
wo k, and Ou looks. IEEE In e ne o Things
Magazine. 2024, ol. 7, no. 3, pp. 30-37.
DOI: 10.1109/IOTM.001.2300255.
[8] SRIVASTAVA, A., e al. Tempo al Beha io
Analysis o he Impac o Combined Tempe -
a u e and Humidi y Va ia ions on a Pho oa-
cous ic CO2Senso . IEEE Applied Sensing Con-
e ence (APSCON), Goa, India. 2024, pp. 1-4.
DOI: 10.1109/APSCON60364.2024.10465885.
[9] SRIVASTAVA, A., e al. Da a-d i en Modelling o
an Indi ec Pho oacous ic Ca bon dioxide Senso .
IEEE Applied Sensing Con e ence (APSCON),
Goa, India. 2024, pp. 1-4. DOI: 10.1109/AP-
SCON60364.2024.10465802.
©2025 ADVANCES IN ELECTRICAL AND ELECTRONIC ENGINEERING 89
SIKORA, A. VOLUME: 23 |NUMBER: 2 |2025 |JUNE
[10] SELLE, P., A. SIKORA, O. AMFT. En-
hanced Angle o A i al (AoA) P edic ion
h ough Neu al Ne wo ks and P ima y and Ad-
jacen An ennas Rep esen a ion T ans o ma ion
(PAAR). IEEE In e na ional Con e ence on Ad-
anced Ne wo ks and Telecommunica ions Sys-
ems (ANTS), Guwaha i, India. 2024, pp. 1-6.
DOI: 10.1109/ANTS63515.2024.10898694.
[11] SCHWENDEMANN, S., Z. AMJAD, A.
SIKORA. A su ey o machine-lea ning
echniques o condi ion moni o ing and p e-
dic i e main enance o bea ings in g inding
machines. Compu e s in Indus y. 2021, ol. 125.
DOI: 10.1016/j.compind.2020.103380.
[12] SCHWENDEMANN, S., A. RAUSCH, A.
SIKORA. A Hyb id P edic i e Main enance
Solu ion o Faul Classi ica ion and Remain-
ing Use ul Li e Es ima ion o Bea ings Using
Low-Cos Senso Ha dwa e. P ocedia Com-
pu e Science. 2024, ol. 232, pp. 128-138.
DOI: 10.1016/j.p ocs.2024.01.013.
[13] SCHWENDEMANN, S., A. RAUSCH, A.
SIKORA. De ailed S udy o Di e en Deg ada-
ion S ages o Bea ings in a P ac ical Re e ence
Da ase . IEEE 28 h In e na ional Con e ence
on Eme ging Technologies and Fac o y Au oma-
ion (ETFA), Sinaia, Romania. 2023, pp. 1-8.
DOI: 10.1109/ETFA54631.2023.10275478.
[14] SCHWENDEMANN, S., A. SIKORA. T ans e -
Lea ning-Based Es ima ion o he Remaining Use-
ul Li e o He e ogeneous Bea ing Types Using
Low-F equency Accele ome e s. J. Imaging. 2023,
ol. 9, no. 2. DOI: 10.3390/jimaging9020034.
[15] ZAHARY, A. T., N. A. AL-SHAIBANY,
A. SIKORA. A e iew o In usion De ec-
ion Sys ems o he In e ne o Things. 1s
In e na ional Con e ence on Eme ging Tech-
nologies o Dependable In e ne o Things
(ICETI), Sana’a, Yemen. 2024, pp. 1-6.
DOI: 10.1109/ICETI63946.2024.10777253.
[16] MESSAAD, M. A., C. JERAD, A. SIKORA. AI
App oaches o IoT Secu i y Analysis. Ad ances
in In elligen Sys ems and Compu ing. 2021, ol.
1353, pp. 47-70. DOI: 10.1007/978-981-16-0730-
1_4.
[17] DANESHGADEH, S., e al. A Fede a ed
Lea ning-Based F amewo k o Anomaly De ec-
ion in IIoT Ne wo ks: In eg a ing Edge P ocess-
ing wi h SIEM Co ela ion. unde e iew a : IEEE
In e na ional Con e ence on Edge Compu ing and
Communica ions (IEEE EDGE 2025). 2025.
[18] MHAWI, M. Y., H. N. ABDULLAH, A. SIKORA.
GANs o Image Secu i y Applica ions: A Li e -
a u e Re iew. I aqi Jou nal o In o ma ion and
Communica ion Technology. 2024, ol. 7, no. 2.
DOI: 10.31987/ijic .7.2.296.
[19] ANTWI, R., e al. A Su ey on Ne wo k
Op imiza ion Techniques o Blockchain
Sys ems. Algo i hms. 2022, ol. 15, no. 6.
DOI: 10.3390/a15060193.
[20] NARTEY, C., e al. Blockchain-IoT pee de-
ice s o age op imiza ion using an ad anced
ime- a ian mul i-objec i e pa icle swa m
op imiza ion algo i hm. Algo i hms. 2022.
DOI: 10.1186/s13638-021-02074-3.
©2025 ADVANCES IN ELECTRICAL AND ELECTRONIC ENGINEERING 90