XR T aining in Indus y 5.0: Ad ancing
Human-Machine Collabo a ion wi h he XR5.0
T aining Pla o m
Jo ge Oli ei a1,2(B), Tomaz Sa ai a1, Ha sh Manoj Shah3,
Emmanouil Ma ogio gis4, A gy o Ma ogio gou5, and A hanasios Kiou is5
1 Imme si e Li es, Lispolis, n44, 1600-546 Lisbon, Po ugal
{jo ge.oli ei a, omaz.sa ai a}@imme si eli es.p
2 Uni e si y Lusó ona/HEI-Lab, Campo G ande, 376, 1749-024 Lisbon, Po ugal
3 HOLO-Indus ie 4.0 So wa e GmbH, Ca l-Zeiss-Ring 19, 85737 Ismaning, Ge many
[email p o ec ed]
4 Synelixis Solu ions S.A, 157 Pe issou and Chalkidos, 14343 A hens, G eece
[email p o ec ed]
5 Uni e si y o Pi aeus, 80 M. Ka aoli and A. Dimi iou, 18534 Pi aeus, G eece
{ma gy,kiou is}@unipi.g
Abs ac . Indus ial T aining, especially aining gea ed owa ds Indus y 5.0 –
e e ing o obo and sma machines wo king alongside people, is an e ol ing
ield, and ecen echnological ad ancemen s in Ex ended Reali y (XR) and A i-
icial In elligence (AI) ha e p opelled in e es owa d his goal. The combina ion
o hese echnologies allows he implemen a ion o imme si e, adap i e, and pe -
sonalized lea ning expe iences, which can be u ilized by he wo k o ce in on- and
o - he-job con ex s o add ess aining in inc easingly complex indus ial sys ems.
Howe e , he adop ion o XR-based aining aces se e al challenges, including
compu a ional demands, la ency, usabili y cons ain s, and pe sonaliza ion. To
add ess hese limi a ions, he XR5.0 T aining Pla o m p o ides a s a e-o - he-a
cloud in as uc u e and AI-enhanced aining solu ion designed o c ea e, man-
age, and display XR con en o use s wi h op imized pe o mance and accessibili y.
The pla o m is s uc u ed a ound h ee (3) co e componen s, namely: (i) he Holo-
ligh Hub, o managing and o ches a ing XR applica ions enabling low-la ency
s eaming ia a cloud-based in as uc u e; (ii) he XR T aining Asse Reposi o y
o ensu e secu e s o age o aining ma e ials; and (iii) he XR T aining Man-
agemen Sys em o he c ea ion, managemen , and isualiza ion o XR-na i e
aining p og ams. This pla o m add esses he limi a ions o exis ing aining
pla o ms while educing ha dwa e dependency by adop ing a de ice-agnos ic
app oach. T his ensu es a mo e e icien and scalable aining ecosys em, enhanc-
ing wo k o ce alignmen wi h Indus y 5.0 en i onmen s. This pape p esen s he
pla o m’s a chi ec u e, key unc ionali ies, and in eg a ion s a egies while dis-
cussing i s po en ial o ans o m indus ial aining h ough in elligen , imme si e,
and da a-d i en lea ning solu ions.
Keywo ds: Ex ended Reali y · XR S eaming · A i icial In elligence ·
T aining · Indus y 5.0
© The Au ho (s), unde exclusi e license o Sp inge Na u e Swi ze land AG 2026
L. T. De Paolis e al. (Eds.): XR Salen o 2025, LNCS 15738, pp. 93–100, 2026.
h ps://doi.o g/10.1007/978-3-031-97769-5_7
94 J. Oli ei a e al.
1 In oduc ion
Indus y 5.0 ma ks a signi ican shi om he au oma ion pa adigm o Indus y 4.0
owa ds a human-cen ic pa adigm, emphasizing on human-machine coope a ion [1].
This app oach in eg a es ad anced echnologies wi h human expe ise o c ea e mo e
p oduc i e and wo ke - iendly en i onmen s [2]. One o he key a eas whe e Indus y
5.0 is d i ing inno a ion is indus ial aining, mainly h ough he u iliza ion o ad anced
echnologies such as Ex ended Reali y (XR) and A i icial In elligence (AI) [3]. These
echnologies enable mo e imme si e, adap i e, and pe sonalized lea ning expe iences,
p o iding cus omized en i onmen s o aining indus ial ope a o s in complex indus-
ial p ocedu es [4]. Following a epo by Ma ke sandMa ke s [5], he XR ma ke is
p edic ed o g ow om $15.3 billion in 2020 o $77.0 billion by 2025, d i en by he
inc easing u iliza ion o AR in domains including indus ial aining, showcasing i s
o e all impo ance and alue.
Despi e hese oppo uni ies, se e al challenges con ibu e o he slow adop ion o
XR-based solu ions in he indus ial domain [6]. The cu en limi a ions include la ency
needed o indus ial se ings ha a e cons ained by echnical speci ica ions ela ed o
ende ing ame a e, he e esh a e o XR display, and he inpu lag om he in e -
ac ing con en due o highe compu a ional demands [7]. Addi ionally, XR’s di icul y
o beginne s and usabili y challenges (especially p onounced when XR sys ems do
no adequa ely add ess concep s like p esence and imme sion) a e limi ing XR-based
aining solu ion adop ion in eal-wo ld indus ial se ings [8]. Re iews o exis ing ain-
ing pla o ms o indus ial scena ios sugges a p edominan ocus on echnical and
echnological subjec s while o en neglec ing he human-cen ic aspec s o aining [9,
10].
To his con ex , in his manusc ip he XR5.0 T aining Pla o m is being in oduced,
as a componen o he XR5.0 p ojec . The la e is an EU- unded ini ia i e aiming o
o e come hese challenges by es ablishing speci ic guiding p inciples and s a egies o
implemen ing XR echnologies wi hin Indus y 5.0. XR 5.0 aims o ensu e seamless
human-machine collabo a ion, enabling ope a o s o lea n in highly in e ac i e en i on-
men s ha adjus in eal- ime acco ding o use pe o mance and needs [11]. As pa
o his p ojec , he XR5.0 T aining Pla o m will be de eloped ha in eg a es ad anced
XR echnologies wi h AI ools o enhance aining e iciency, adap abili y, and eal-
wo ld applicabili y [8]. This p ojec u ilizes XR s eaming solu ions ha s eam XR
con en di ec ly om he cloud o XR end-de ices o mi iga e la ency and compu a-
ional demand- ela ed issues, ensu ing a luid and mo e na u al in e ac ion in indus ial
se ings. I also con ains a module designed o c ea e XR-na i e aining p og ams, able
o c ea e a mo e dynamic, imme si e, and pe sonalized aining expe ience.
The es o his pape is o ganized as ollows. In Sec . 2 i is being desc ibed he
XR5.0 T aining Pla o m’s a chi ec u e and ision, including i s co e componen s and
key ea u es, while Sec . 3 concludes his wo k, discussing he u u e wo k o he pla -
o m. While his pape ocuses p ima ily on he a chi ec u al and unc ional design o
he XR5.0 T aining Pla o m, a u u e i e a ion will de ail he expe imen al design, da a
collec ion p o ocols, and analysis echniques planned o e alua ing aining ou comes,
usabili y, and sys em pe o mance in eal-wo ld indus ial scena ios.
XR T aining in Indus y 5.0 95
2 XR5.0 T aining Pla o m
The XR5.0 T aining Pla o m is a cloud-based XR aining and s eaming pla o m
ha will enable Indus y 5.0 ope a o s o on-demand ende XR aining applica ions
speci ic o selec ed indus ial scena ios wi hin he Indus y 5.0 ecosys em. The end-use s
o ope a o s will be able o le e age he aining pla o m o unde go imme si e and AI-
powe ed aining in Augmen ed Reali y/Vi ual Reali y (AR/VR) mode wi h eal- ime
in e ac ions. The pla o m suppo s adminis a i e ope a ions h ough web b owse s
and deli e s XR expe iences o ope a o s ia XR de ices, such as AR/VR headse s
(e.g., Me a Ques and Mic oso Hololens). The pla o m’s a chi ec u e (Fig. 1) ensu es
seamless managemen o applica ions and use s, obus o ches a ion o cloud esou ces,
and e icien XR con en deli e y.
Fig. 1. XR5.0 T aining Pla o m A chi ec u e
2.1 T aining Pla o m A chi ec u e
As shown in Fig. 1, he co e o he XR5.0 T aining Pla o m is composed o he below
h ee (3) p ima y componen s:
•Hololigh Hub -I is he co e s eaming pla o m o managing, hos ing, and s eaming
XR applica ions.
•XR Asse Reposi o y - I e e s o a cloud-based ile s o age sys em o s o ing aining
ma e ials and p og ams.
•XR T aining Managemen Sys em - I allows adminis a o s o c ea e aining ma e i-
als and imme si e s ep-by-s ep aining modules, by linking aining ma e ials s o ed
in he eposi o y.
In addi ion o he co e componen s, he pla o m u ilizes wo (2) suppo ing laye s,
namely he G aphical Use In e ace (GUI) laye , which se es as he in e ac ion laye ,
as well as he Vi ualiza ion P o ide , which se es as he in as uc u e laye . Mo e
in o ma ion is p o ided below:
96 J. Oli ei a e al.
•G aphical Use In e ace (GUI) - I consis s o he Web B owse and he XR De ice
In e ace, whe e:
•The Web B owse can be used by adminis a o s/manage s h ough a cen al
dashboa d o manage XR applica ions, aining p og ams, and use ac i i ies.
•The XR De ice In e ace can be used by he ope a o s o access imme si e aining
h ough XR de ices (e.g., HoloLens 2, Me a Ques 3), enabling eal- ime in e ac ion
wi h XR con en .
•Vi ualiza ion P o ide - I consis s o High-pe o mance Vi ual Machines (VMs)
and he XR applica ions, whe e:
– The High-pe o mance VMs wi h G aphics P ocessing Uni s (GPUs) un XR
applica ions hos ed on he Hololigh Hub, ensu ing scalabili y and eliabili y.
– The XR applica ions buil in Uni y communica e wi h ex e nal se ices de el-
oped wi hin XR5.0 p ojec like he Human-Cen ic Digi al Twin and AI models
(AL, NSAI, GenAI) ia Applica ion P og amming In e aces (APIs), inco po a -
ing pe sonaliza ion and Gene al Da a P o ec ion Regula ion (GDPR)-complian
da a secu i y. I should be men ioned ha he XR applica ions a e buil wi h Uni y
based plugins, e e ing o: (i) he Hololigh S eam ha is used o enable XR
applica ion s eaming om he cloud o he XR end de ices; (ii) he XR Repos-
i o y plugin ha is used o download iles such as 3D models om he asse
eposi o y s o age in o he XR en i onmen ; and (iii) he XR T aining Plugin ha
is used o ende he aining p og ams de eloped using he web b owse in o he
XR en i onmen .
Wi hin he XR5.0 T aining Pla o m, he aining in e ac ion wo k low begins wi h
he adminis a o adding use s, XR applica ions, ela ed asse s on he pla o m, and
aining ma e ials, which can be u he combined in he aining p og ams ia he web
in e ace. Au ho ized ainees un he XR applica ion assigned o hem on he dedica ed
XR de ice, whe eas wi hin he XR applica ion, he ainees can access he aining
p og ams and ma e ials ia he XR T aining Plugin and un he p og ams in an imme si e
en i onmen .
2.2 Co e Componen s
The pla o m is de eloped a ound a i ualiza ion en i onmen using Amazon Web
Se ices (AWS)-based cloud se ices o managing mul iple XR applica ions, each
equipped wi h plugins ha allow he communica ion be ween Hololigh Hub, he Asse
Reposi o y Plugin, and he XR T aining Plugin p o iding eal- ime ende ing and con-
en deli e y o XR de ices. The pla o m will lean on i s co e componen s, u he
explained below.
Hololigh Hub. The Hololigh Hub (Fig. 2) is a cloud-based sys em designed o hos ,
manage, and s eam XR applica ions in eal- ime o ensu e a low-la ency and smoo h XR
expe ience ac oss di e en de ices, such as he Me a Ques and Mic oso Hololens. This
componen cen alizes XR applica ion managemen , allowing o ganiza ions o manage
use s and XR applica ions and se ings e ec i ely. The Hololigh Hub uses AWS cloud
XR T aining in Indus y 5.0 97
in as uc u e o o ches a e and un XR applica ions and le e ages Hololigh S eam o
ensu e imme si e, lag- ee, and secu e XR s eaming.
Fig. 2. Hololigh Hub A chi ec u e
The s ys em ollows a mic ose ices-based amewo k, wi h an API Ga eway sending
eques s o di e en se ices ela ed o use and applica ion managemen , analy ics, and
esou ce o ches a ion. I also in eg a es ex e nal se ices such as Single Sign-On (SSO)
o secu e login, a da abase o managing use and applica ion da a, and AWS i ual
machines o high-pe o mance XR ende ing.
XR Asse Reposi o y. The XR T aining Asse Reposi o y (Fig. 3) is designed o p o ide
lexible and secu e s o age o indus ial aining ma e ials. The co e elemen o he XR
T aining Asse Reposi o y is a Web App designed o manage he s o age o iles and
me ada a linked o aining ma e ials. I s p ima y pu pose is o ensu e lexibili y, allowing
iles o ei he be s o ed close o he Vi ualiza ion P o ide o enhanced pe o mance
o owa ds he edge o add ess secu i y conce ns o echnical cons ain s (Fo example,
speci ic 3D models would need o s ay on-p emises due o secu i y conce ns).
The app inco po a es Owncloud, an open-sou ce ile-sha ing solu ion, o s o e
aining- ela ed asse s and se es as a connec o be ween Owncloud and T aining Pla -
o m componen s like Hololigh Hub and he XR T aining Managemen Sys em. I
enables adminis a o s o c ea e aining ma e ials, assign pe missions, sha e con en
wi h o he use s, and alloca e s o age space o speci ic use s based on hei equi emen s.
Addi ionally, he Web App allows XR De ices o access ( iew/download) aining
ma e ials ia he XR aining plugin o enables use s o in e ac wi h he con en h ough
a s anda d Web b owse .
XR T aining Managemen Sys em. The XR T aining Managemen Sys em (Fig. 4)
enables he c ea ion, managemen , and isualiza ion o he XR T aining Ma e ials and
P og ams. I consis s o wo main modules: (i) he T aining P og ams Au ho ing Tool
(TPAT), a web applica ion o c ea ing and managing aining con en , and (ii) he XR
98 J. Oli ei a e al.
Fig. 3. XR Asse Reposi o y A chi ec u e
T aining Plugin (XRTP), a Uni y plugin ha allows he isualiza ion o XR aining
p og ams and ma e ials on XR de ices.
Fig. 4. XR T aining Managemen Sys em A chi ec u e
The TPAT p o ides a s uc u ed en i onmen o de eloping XR aining con en .
I includes op ions o uploading and managing aining esou ces, adap ing adi ional
ma e ials o XR, and gene a ing new aining con en . The suppo ed o ma s o ma e i-
als o be uploaded include PDF iles, ideo iles (i.e., MPEG4), images (i.e., PNG) and 3D
Models cu en ly suppo ing di e en o ma s (i.e., FBX) ha can be uploaded and used
as XR aining ma e ials. The sys em allows manage s o o ganize aining p og ams
in o s uc u ed wo k lows, using ea u es such as in e ac i e checklis s, wo k low-based
ins uc ions, and mul imedia con en (e.g., images, ideos). The TPAT also in eg a es
AI-powe ed ools o au oma e he c ea ion o con en , o ins ance, by ex ac ing and
c ea ing ex om PDF ma e ials. I is also planned o include AI assis an s o suppo
ope a o s wi hin he XR en i onmen .
The XRTP is esponsible o isualizing XR aining con en in imme si e en i-
onmen s. I p o ides an in ui i e in e ace op imized o XR in e ac ions, suppo ing
hand acking, and allowing a na u al and imme si e expe ience. T aining ma e ials such
as checklis s, images, ideos, and wo k lows a e p esen ed h ough dedica ed iewe s,
allowing ope a o s o in e ac dynamically wi h con en . A T aining P og am Manage
o ganizes a ailable p og ams, guiding use s h ough s uc u ed lea ning expe iences.
Planned ea u es include a i ual dashboa d o managing iews wi hin he XR space
and he adap a ion ea u e ha will consis o an adap i e in e ace ha pe sonalizes he
use expe ience based on in e ac ion da a.
XR T aining in Indus y 5.0 99
2.3 In eg a ion o Componen s
As i has been al eady explained, he XR5.0 T aining Pla o m in eg a es he h ee (3)
co e se ices (i.e., Hololigh Hub, XR T aining Asse Reposi o y, XR T aining Manage-
men Sys em) o ensu ing seamless deli e y o XR aining con en , use managemen ,
and sys em scalabili y. These componen s a e designed o wo k in an in eg a ed manne ,
allowing en e p ises o deli e scalable and e icien aining expe iences. The pla o m
employs a mic ose ices-based a chi ec u e wi h API-d i en communica ion, cloud-
cen ic da a exchange, and e en -d i en o ches a ion, whe eas SSO p o ides secu e,
uni ied use au hen ica ion. The Hololigh Hub manages au hen ica ion and session
con ol, while he XR Asse Reposi o y s o es aining ma e ials accessed ia he XR
T aining Plugin. Real- ime s eaming and dynamic esou ce alloca ion op imize pe o -
mance, whe eas he o ches a ion o hese componen s enables e icien managemen
o applica ions, aining modules, esou ces, asse s, use pe missions, and secu e ile
sha ing. This in eg a ion ensu es secu e access con ol, e icien con en managemen ,
and dynamic esou ce alloca ion, enabling a obus and scalable XR aining expe ience.
To his con ex , u u e e inemen s aim o enhance o ches a ion, au hen ica ion lows,
and s eaming pe o mance.
3 Conclusion
In his pape , he XR5.0 T aining Pla o m has been p esen ed, as a obus solu ion o
add ess he challenges o indus ial aining in he Indus y 5.0 pa adigm. This pla o m
in eg a es XR echnologies wi h AI and cloud-based sys ems, p o iding imme si e,
adap i e, and human-cen ic aining en i onmen s ha go beyond adi ional me hods.
By elying on componen s such as he Hololigh Hub, XR Asse Reposi o y, and XR
T aining Managemen Sys em, he pla o m is lexible enough o allow c ea ing, man-
aging and deli e ing XR con en o aining in di e en indus ial scena ios aiming o
enhance expe ien ial lea ning. I s a chi ec u e ensu es scalabili y, low-la ency s eaming,
and secu e access, con ibu ing o e ec i e deploymen in eal-wo ld applica ions.
This app oach ad ances beyond he s a e-o -a by o e ing an in eg a ed s olu ion ha
combines XR, AI, and a cloud-based in as uc u e o deli e imme si e, adap i e and
pe sonalized lea ning expe iences be e aligned wi h human ac o s. Being an XR-based
solu ion, he pla o m can suppo bo h on- he-job and o - he-job aining, p o iding
a hyb id lea ning model whe e ounda ional knowledge can be acqui ed in con olled
en i onmen s and la e ein o ced h ough eal-wo ld applica ion being pa icula ly el-
e an o sec o s s uch as manu ac u ing and main enance. Howe e , one limi a ion o he
pla o m is i s eliance on high-bandwid h in e ne connec ions o eal- ime s eaming,
which can be a limi a ion in emo e o low-connec i i y indus ial se ings.
As he XR5.0 p ojec e ol es, u u e wo k will ocus on enhancing o ches a ion
capabili ies, explo ing he me ics o he adap a ion and pe sonaliza ion, and expanding
in e ope abili y ac oss de ices and indus ial con ex s. Ul ima ely, he XR5.0 T aining
Pla o m lays he ounda ion o scalable, in ui i e, and e icien XR aining solu ions
ha ake ad an age o human-machine collabo a ion in indus ial en i onmen s. As
pa o u u e wo k, expe imen al alida ion will include pilo s udies in indus ial se -
ings o measu e imp o emen s in aining e iciency, use engagemen , and lea ning
100 J. Oli ei a e al.
e en ion, using bo h quan i a i e me ics (e.g., ask comple ion ime, e o a es) and
quali a i e use eedback. This will ensu e a comp ehensi e assessmen o he pla o m’s
e ec i eness.
Acknowledgmen s. The esea ch p esen ed in his pape has ecei ed unding om he EU unded
P ojec XR5.0 (GA no 101135209).
Disclosu e o In e es s.. The au ho s ha e no compe ing in e es s o decla e ha a e ele an o
he con en o his a icle.
Re e ences
1. Zalozhne , A.Y., Ginz, V.N.: Indus y 4.0: unde lying echnologies. Indus y 5.0: human-
compu e in e ac ion as a Tech B idge om Indus y 4.0 o Indus y 5.0. In: 9 h In e na ional
Con e ence on Web Resea ch (ICWR) (2023)
2. Mose , T., Wol a sbe ge , J., So ko, S., Abu Naim, B.: Ex ended eali y based educa ion
and aining o human-cen ic Indus y 5.0 skill enhancemen . In: NOMS 2024–2024 IEEE
Ne wo k Ope a ions and Managemen S ymposium (2024)
3. Rauschnabel, P.A., Felix, R., Hinsch, C., Shahab, H., Al , F.: Wha is XR? Towa ds a
amewo k o augmen ed and i ual eali y. Compu . Hum. Beha . 133, 107289 (2023)
4. Mose , T., Schlage , A.: T aining and assis ance in indus ial en i onmen s using ex ended
eali y. In: IEEE Con e ence on Vi ual Reali y and 3D Use In e aces Abs ac s and
Wo kshops (VRW), pp. 97–103 (2024)
5. Ma ke s and Ma ke s: Augmen ed and i ual eali y ma ke size, sha e and
g ow h (2024). h ps://www.ma ke sandma ke s.com/Ma ke -Repo s/augmen ed- eali y-
ma ke -82758548.h ml
6. Wang, B., Zheng, L., Wang, Y., Fang, W., Wang, L.: Towa ds he indus y 5.0 on ie : e iew
and p ospec o XR in p oduc assembly. J. Manu . Sys ., 777–811 (2024)
7. Simões, B., C eus, C., Ca e e o, M.P., Ochaí a, A.G.: S eamlining XR echnology in o indus-
ial aining and main enance p ocesses. In: Web3D ‘20: P oceedings o he 25 h In e na ional
Con e ence on 3D Web Technology, ol. 5, pp. 1–7 (2020)
8. Daling, L.M., Schli meie , S.J.: E ec s o augmen ed eali y, i ual eali y-, and mixed eal-
i y–based aining on objec i e pe o mance measu es and subjec i e e alua ions in manual
assembly asks: a scoping e iew. Hum. Fac o s 66(2), 589–626 (2024)
9. Caze i, G.T., de San a-Eulalia, L.A., Se a im, M.P., Anholon, R.: T aining o Indus y 4.0: a
sys ema ic li e a u e e iew and di ec ions o u u e esea ch. B azilian J. Ope . P od. Manag.
19, 1–21 (2022)
10. C nob nja, J., S e ano ic, D., Rome o, D., So ic, S., Ma jano ic, U.: Digi al ans o ma ion
owa ds Indus y 5.0: a sys ema ic li e a u e e iew. In: Al nes, E., Romsdal, A., S andhagen,
J.O., on Cieminski, G., Rome o, D. (eds) Ad ances in P oduc ion Managemen Sys ems.
P oduc ion Managemen Sys ems o Responsible Manu ac u ing, Se ice, and Logis ics
Fu u es. APMS 2023. IFIP Ad ances in In o ma ion and Communica ion Technology, ol
689. Sp inge , Cham (2023)
11. Kiou is, A. e al.: XR5.0: Human-cen ic AI-enabled ex ended eali y applica ions o Indus-
y 5.0. In: 36 h Con e ence o Open Inno a ions Associa ion (FRUCT), Lappeen an a,
Finland, 2024, pp. 314–323 (2024)
