Enhancing Machine y Design by Using A i icial
In elligence
Juan Pablo Fiesco, Miguel Angel Ma eo-Casali, Bea iz And es(B), and Raul Pole
Resea ch Cen e on P oduc ion Managemen and Enginee ing (CIGIP), Uni e si a Poli ècnica
de València (UPV), Camino de Ve a S/N, 46022 Valencia, Spain
{j iesco,mma eo,band es, pole }@cigip.up .es
Abs ac . This pape examines he signi icance o Indus y 4.0 and a i icial in el-
ligence (AI) in he manu ac u ing sec o , pa icula ly by emphasising he ole o
design phase in he machine y li e cycle. The design phase o a machine is a
complex ask ha equi es an ad anced enginee ing and physics knowledge le el.
Ne e heless in he echnology e a, compu e -aided design ools acili a e he
design ask. The a ea o da a execu ion and simula ion o machine beha iou in
di e en scena ios is being esea ched and exploi ed by echnologies, such as he
In e ne o Things (IoT) o AI. Wi h his pape , h ee AI-based ools a e p oposed
and concep ualised o suppo AI-assis ed op imisa ion o gene a e design p opos-
als o manu ac u e indus ial equipmen , s uc u al componen s, mechanisms and
con ol componen s.
Keywo ds: a i icial in elligence ·design li e cycle ·equipmen indus y ·
Indus y 4.0
1 In oduc ion
In he mode n Indus y 4.0 en i onmen , echnological compe ences a e c ucial o com-
panies o succeed in global ma ke s. To op imise he added alue in he design p ocess
o complex p oduc s, such as machine y wi h physical and dynamic elemen s, p oduc
de elopmen p ocess managemen is being explo ed (Almoslehy & Alkah ani, 2021).
The design unc ion is esponsible o c ea ing a p oduc ha bes mee s cus ome s’
needs, and i p o ides in o ma ion on a ious aspec s o he p oduc be o e and a e
i s p oduc ion. P oduc design equi emen s encompass pe o mance, eliabili y, size,
cos , manu ac u ing, indus y s anda ds, go e nmen egula ions, in ellec ual p ope y
and sus ainabili y (P ei e , 2009). In he Indus y 4.0 con ex , he gene a ion and collec-
ion o la ge amoun s o da a in a ious o ma s h oughou he p oduc li e cycle a e
becoming mo e commonplace. To make sense o his in o ma ion, new da a p ocessing
and analysis me hods mus be de eloped o ans o m da a in o easily unde s andable
and explainable o ma s. This pape aims o iden i y wha kind o needs manu ac u ing
companies ha e du ing he p ocess o designing indus ial equipmen and how hese
needs can be add essed wi h AI ools. Thus he ollowing esea ch ques ions a ise:
RQ1 Wha needs o exis in he indus ial equipmen design p ocess?
© The Au ho (s), unde exclusi e license o Sp inge Na u e Swi ze land AG 2024
J. Bau is a-Valhondo e al. (Eds.): CIO 2023, LNDECT 206, pp. 342–347, 2024.
h ps://doi.o g/10.1007/978-3-031-57996-7_59
Enhancing Machine y Design by Using A i icial In elligence 343
RQ2 Wha is he po en ial o using AI in he indus ial equipmen design?
RQ3 Wha kind o AI ools can mee indus ial equipmen design needs?
Acco dingly, his pape is o ganised as ollows: Sec . 2desc ibes he ollowed
me hodology. Sec ion 3examines he s a e o he a on needs in he design p ocess
by examining he link be ween AI du ing he p oduc design p ocess. Sec ion 4p oposes
a se o AI ools o ul il he iden i ied needs. Finally, conclusions a e discussed in Sec . 5.
2 Me hodology
The me hodology used in his pape is based on e iewing he exis ing li e a u e o
iden i y he needs ha appea in he design p ocess by conside ing he cu en ola ile
cha ac e is ics o cus ome demand. The nex s ep in he e iew ocuses on how he
digi al echnologies ha suppo Indus y 4.0 ha e been implemen ed in o he design
p ocess o deal wi h sho li e cycle p oduc s. The pape is based on a concep ual case
ha desc ibes he ools de eloped in he AIDEAS (2022) Eu opean P ojec , whose
main aim is o de elop AI echnologies o suppo ing he en i e li e cycle o indus ial
equipmen . This pape ocuses on concep ualising AI ools o suppo he design phase
o he li e cycle in speci ic Eu opean machine y manu ac u ing companies.
3 S a e o he A
The design p ocess in ol es con inuous in e play be ween de ining he p oblem and
gene a ing solu ions, and mul iple i e a ions equi ed o achie e an op imal o obus
design solu ion. The ollowing subsec ions e iew design p ocess needs and how hese
needs a e ul illed by applying indus y 4.0 echnologies.
3.1 P oduc Design P ocess Needs
Enhancing he quali y o decision making du ing p oduc design and de elopmen is
c ucial, bu can be challenging gi en he as amoun o a ailable design in o ma ion.
Decision make s o en s uggle o access he igh in o ma ion hey need a he igh
ime, which makes he p ocess ime-sensi i e (Riesene e al., 2019). The design p ocess
o mechanical p oduc s can be complex and ime-consuming because o he nume ous
s eps and imp o emen i e a i e ac i i ies ha a e in ol ed. The e is also p essu e o
educe cos s in a compe i i e ma ke en i onmen (Ka ayel e al., 2013). T adi ional
design app oaches may need o be e ised o adequa ely mee hese needs. Today, p od-
uc and sys em design no only sa is y use equi emen s and speci ica ions, bu also
enable in e ac ions be ween di e en componen s o ini ia e co ec i e ac ions when-
e e necessa y by making ope a ions and he en i e p oduc li e cycle mo e in elligen
(Hou e al., 2008). To adap o he ola ili y o cus ome demands, cus omised p oduc ,
sho e p oduc li e cycles and inc eased small ba ch p oduc ion a e eme ging ends.
344 J. P. Fiesco e al.
3.2 Digi isa ion in P oduc Design P ocesses
To mee ola ile demands, new mode n p oduc design models eme ge o p omo ing
digi al design and compu e -aided echnologies a e e y link o he p oduc li e cycle (Lei
e al., 2022). To add ess he posed challenges, esea che s ha e ocused on es ablishing a
coope a i e and in eg a ed en i onmen using a ious compu e -aided sys ems, includ-
ing he Compu e -Aided Design, Manu ac u ing and Enginee ing (CAD/CAM/CAE)
applica ion so wa e, da abases and web-based se ices (Sa ic e al., 2018). Al hough
he e a e many suppo ools a ailable o la e design s ages, such as de ailed design,
he e a e ela i ely e y ew ools a ailable o ini ial concep ual s ages (Meni u e al.,
2003). Such lack o suppo can hinde a designe ’s abili y o manipula e, o ganise and
ep esen design da a by educing hei e ec i eness. The de elopmen o Indus y 4.0
has led o signi ican ad ances being made in digi al win (DT) echnologies by open-
ing he way o in eg a e AI, speci ically da a-d i en machine-lea ning (ML) models.
Howe e , a ecu ing p oblem wi h hese models is hei suscep ibili y o ain da a and
esul s lacking uniqueness (Fa biz e al., 2022). In he pas , DT echnology has been used
p ima ily o asks like aul de ec ion, p edic ing main enance needs and pe o mance
analysis. Howe e , mo e a en ion should be paid o he po en ial o DT echnology in
p oduc design. Speci ically, explo ing how he DT ( i ual p oduc ) can imp o e he
design p ocess, and o lead o mo e e icien , in o med and esou ced esul s (Tao e al.,
2019), has been limi ed. The DT echnology can play a c i ical ole in his p ocess by
p o iding a i ual ep esen a ion o he physical p oduc and acili a ing he in eg a ion
o Big Da a and ML o suppo p oduc design (Niu e al., 2021). As inc easingly mo e
design in o ma ion is s o ed in da abases, he in eg a ion o he DT wi h Big Da a and AI
can p o ide designe s wi h a new le el o unde s anding and apid design decisions by
means o AI algo i hms o p edic and adap any kind o design p oposals by p o iding
designe s wi h a compe i i e ad an age in ime- o-ma ke e ms.
4 AI o Suppo he P oduc Design P ocess
The aim o his pape is o iden i y he app op ia e ins umen s ha can be used in he
indus ial machine y design s age wi hin he scope o he Eu opean P ojec AI D i en
Indus ial Equipmen P oduc Li e Cycle Boos ing Agili y, Sus ainabili y and Resilience
(AIDEAS, 2022). The p ojec seeks o enhance he sus ainabili y, esilience and agili y o
Eu opean machine y manu ac u ing companies by le e aging AI-d i en echnologies o
suppo he comple e li e cycle o indus ial machine y, including i s design, p oduc ion,
usage and main enance.
In he indus ial equipmen design phase, ca e ul planning and execu ion a e essen-
ial o ensu e op imal pe o mance and e iciency. By in eg a ing AI echnologies wi h
CAD/CAM/CAE sys ems, designe s can op imise he design o indus ial equipmen
s uc u al componen s, mechanisms and con ol componen s. Using hese sui able ools
can signi ican ly enhance he key ac i i ies in ol ed in he design phase, which can esul
in be e machine pe o mance and inc eased e iciency.
Nex a se o h ee ools is p oposed and concep ualised o in ol e AI ools in he
design phase o he machine y indus y li e cycle.
Enhancing Machine y Design by Using A i icial In elligence 345
The i s p oposed AI-based ool e e s o as he Machine Design Op imise (MDO).
The MDO consis s o an AI-powe ed ool ha assis s designe s o de ine he key
design pa ame e s in mul iphysical sys ems. AI is used o op imise design pa ame-
e s o machines by employing echniques like e olu iona y algo i hms o me aheu is ic
algo i hms o ind op imal con igu a ions ha mee speci ied objec i es, such as min-
imising ene gy use, maximising h oughpu o educing ma e ial usage. AI in he MDO
will enable algo i hms o be designed ha lea n om exis ing designs and use -de ined
cons ain s o gene a e new design al e na i es. AI-enabled simula ions and i ual p o o-
yping ools can be employed o simula e and e alua e di e en machine design a ian s.
Finally, AI algo i hms can analyse design speci ica ions, pe o m simula ions and com-
pa e designs o p ede ined c i e ia o de ec design laws o isks ea ly in he de elopmen
p ocess.
The second p oposed AI-based ool ocuses on da a syn hesis o aining he MDO
AI ool, and is called he Machine Syn he ic Da a Gene a o (MDG). The MDG syn-
hesises high-quali y da ase s by simula ions, which a e essen ial o analysing machine
design and aining op imisa ion algo i hms o p opose op imal design pa ame e s. As
a esul , AI solu ions will soon be accessible o small-scale and sho - e m p ojec s,
and will equi e ewe esou ces o ain he necessa y algo i hms. The MDG oolki
will enable o syn hesise la ge high-quali y da ase s by simula ions o he machine
design analysis and o aining he op imisa ion algo i hms ha will p opose op imal
design pa ame e s. AI MDG solu ions will be accessible o sho e ime se ies and
smalle olume p oduc ions, and will equi e ewe esou ces o aining he necessa y
AI algo i hms.
To in eg a e AI-assis ed op imisa ion modules wi h exis ing CAx sys ems, an in e -
ope abili y sys em is essen ial. CAx Addon aims o de elop AI-assis ed op imisa ion
modules (MDO and MDG) in o p oduc s ha can be easily in eg a ed wi h s anda d
CAD/CAM/CAE sys ems. This in ol es c ea ing use in e aces and APIs ha will
allow seamless in eg a ions wi h sys ems like SolidWo ks by making hem eady o
use in eal-wo ld scena ios. APIs and UIs will inco po a e he indi idual unc ionali ies
o he op imisa ion modules, bu will also conside he equi emen s o o he s anda d
CAx solu ions. These modules can signi ican ly enhance machine pe o mance and e i-
ciency by imp o ing indus ial equipmen designs. By employing AI-powe ed ools like
he MDO and he MDG, designe s can imp o e he indus ial equipmen design phase
and op imise pe o mance.
5 Conclusion
The in eg a ion o AI ools in o he equipmen design phase o e s nume ous bene i s,
including he educ ion o he design cycle ime and was e o mass, and enhancing
sus ainabili y h oughou he ex ended li e cycle o indus ial equipmen . The pape
concep ualises a se o AI-based oolki s, such as he MDO, he MDG and CAx Addon,
which can u he op imise design p ocess by allowing he c i ical aspec s o equipmen
o be op imally designed. Using AI ools in ends o s eamline he equipmen design
p ocess and o imp o e he design ea u es o a longe li e cycle by allowing equipmen
o be pu o good use. This ini ia i e explains a i s app oach ha de ines he essen ial
346 J. P. Fiesco e al.
cha ac e is ics o se e al AI ools ha a e employed o imp o e he machine li e cycle
wi h AI echniques. A u u e esea ch line p oposes iden i ying which ype o AI ech-
niques a e he mos sui able ones o de eloping hese ools. Ano he esea ch line is o
iden i y and de ine he equi emen s o each ool a he use le el o hem o be use ul.
In line wi h his, his pape aims o be a basis o how AI can be applied in he machine y
design p ocess o, hus, be able o make a me hodology o concep ual amewo k o
di e en ools.
Acknowledgemen s. The esea ch ha led o hese indings ecei ed unding om wo sou ces.
The i s sou ce o unding was om he Ho izon Eu ope F amewo k P og amme (HORIZON) wi h
G an Ag eemen No. 101057294 “AI D i en Indus ial Equipmen P oduc Li e Cycle Boos ing
Agili y, Sus ainabili y, and Resilience (AIDEAS)”. The second sou ce o unding was om he
Regional Depa men o Inno a ion, Uni e si ies, Science, and Digi al Socie y o he Gene ali a
Valenciana “P og ama In es igo” ( e . INVEST/2022/330), which he Eu opean Union suppo ed
- Nex Gene a ionEU unde he Plan de Recupe ación, T ans o mación y Resiliencia.
Re e ences
AIDEAS (2022) AI D i en Indus ial Equipmen P oduc Li e Cycle Boos ing Agili y, Sus ain-
abili y and Resilience. Eu opean Union’s Ho izon Eu ope Resea ch and Inno a ion P og amme
unde g an ag eemen No. 101057294
Almoslehy, S.A.M., Alkah ani, M.S.: Key app oaches, isks, and p oduc pe o mance in managing
he de elopmen p ocess o complex p oduc s sus ainably. Sus ain. (Swi z.) 13(9), 4727 (2021).
h ps://doi.o g/10.3390/su13094727
Fa biz, F., Habibullah, M.S., Hamadicha e , B., Maszczyk, T., Agga wal, S.: Knowledge-
embedded machine lea ning and i s applica ions in sma manu ac u ing. J. In ell. Manu .
(2022). h ps://doi.o g/10.1007/s10845-022-01973-6
Hou, J., Su, C., Wang, W.: Knowledge managemen in collabo a i e design. In: P oceedings o
2008 IEEE In e na ional Con e ence on Se ice Ope a ions and Logis ics, and In o ma ics,
IEEE/SOLI 2008, ol. 1, pp. 848–852 (2008). h ps://doi.o g/10.1109/SOLI.2008.4686517
Ka ayel, D., Ozkan, S.S., Va anse e , F.: In eg a ed knowledge-based sys em o machine design.
Ad . Mech. Eng. 2013, 1–10 (2013). h ps://doi.o g/10.1155/2013/702590
Lei, Y., Vyas, S., Gup a, S., Shabaz, M.: AI based s udy on p oduc de elopmen and p ocess
design. In . J. Sys . Assu . Eng. Manage. 13, 305–311 (2022). h ps://doi.o g/10.1007/s13198-
021-01404-4
Meni u, K., Ri a d, H., Béda d, C.: Speci ica ions o compu e -aided concep ual building design.
Des. S ud. 24(1), 51–71 (2003). h ps://doi.o g/10.1016/S0142-694X(02)00009-1
Niu, X., Wang, M., Qin, S.: P oduc design li ecycle in o ma ion model (PDLIM). In . J. Ad .
Manu . Technol. 118, 2311–2337 (2021). h ps://doi.o g/10.1007/s00170-021-07945-z/Pub
lished
P ei e , M.: Design equi emen s. In: Ma e ials Enabled Designs, pp. 23–50. Else ie (2009).
h ps://doi.o g/10.1016/b978-0-7506-8287-9.00002-1
Riesene , M., Schuh, G., Dölle, C., Tönnes, C.: The digi al shadow as enable o da a analy ics in
p oduc li e cycle managemen . P ocedia CIRP 80, 729–734 (2019). h ps://doi.o g/10.1016/j.
p oci .2019.01.083
Enhancing Machine y Design by Using A i icial In elligence 347
Sa ic, I., Pe an, N., Mumino ic, A., Colic, M.: De elopmen o in eg a ed in elligen cad sys em
o design o sha s. Tehnicki Vjesnik 25, 99–104 (2018). h ps://doi.o g/10.17559/TV-201705
21194820
Tao, F., e al.: Digi al win-d i en p oduc design amewo k. In . J. P od. Res. 57(12), 3935–3953
(2019). h ps://doi.o g/10.1080/00207543.2018.1443229
