Elzant
Elzant
Elzant
THÈSE
présentée par : Chawki EL ZANT
soutenue le : 16 Juillet 2021
et co-encadrée par :
[Mr BENFRIHA Khaled]
Jury
M. Farid BAKIR, Pr, LIFSE, HESAM Président
M. Abdelaziz BOURAS, Pr, College of Engineering, Qatar University (QU) Rapporteur T
M. Abdelkhalak EL HAMI, Pr, département de mécanique, INSA Rouen Rapporteur H
Mme. Rima HLEISS, Maitre de Conférences, Faculté de Génie, U. Libanaise Examinatrice È
M. Améziane AOUSSAT, Pr, LCPI, HESAM Examinateur S
M. Khaled BENFRIHA, Maître de Conférences, LCPI, HESAM Examinateur E
M. Patrick LE MEN, Ingénieur, MBA, Directeur Ingénieurs2000 Invité
1
2
Résumé
L'industrie 4.0 est un système dynamique et intégré permettant de contrôler l'ensemble de la chaîne de valeur du
cycle de vie des produits. L'intégration verticale et horizontale et la fusion des composants physiques et virtuels
sont essentielles, de même que la mise en œuvre des tendances technologiques telles que les systèmes cyber
physiques et l'Internet des objets en tant que catalyseurs clés d'un haut niveau d'intégration à l'échelle mondiale.
L'introduction de l'Industrie 4.0, caractérisée par des processus de fabrication intelligents et connectés, intéresse
de plus en plus les entreprises qui s'efforcent de s'adapter à la demande croissante du marché et des clients.
L'objectif de ce projet de doctorat est d'aider les entreprises à évoluer de la fabrication automatisée traditionnelle
vers la numérisation du processus de fabrication en proposant un nouveau modèle d'intégration de l'Industrie 4.0.
La variété et la complexité des aspects de l'Industrie 4.0, ainsi que les défis techniques et de gestion doivent être
pris en compte pour fournir aux entreprises des approches d'intégration cohérentes sur la base d’un nouveau
Framework d'intégration. L'expérimentation a été réalisée à la plateforme technologique 4.0 des Arts et Métiers,
campus de Paris. Le modèle développé et le Framework proposé ont été vérifiés grâce à un travail intensif
d’expérimentation et de mise en œuvre à la plateforme 4.0.
Mots-clés : Industrie 4.0, Systèmes cyber-physiques (CPS), Internet des objets (IoT), Modèles d'intégration,
Transformation numérique, Framework d'intégration, Flexibilité, Modularité, Déploiement de données, Robotique
avancée, Système d'exécution de fabrication.
Abstract
Industry 4.0 is a dynamic and integrated system for employing control over the entire value chain of the lifecycle
of products. Vertical and horizontal integration and fusion of physical and the virtual components are critical for
Industry 4.0, as well as technology trends implementation such Cyber Physical Systems and Internet of Things as
key enablers of high level of integration at a global scale. Introducing Industry 4.0 as the smart and connected
manufacturing processes enabler is increasingly attracting companies trying hard to adapt to the growing market
and customer demands. Due to the emergence of Industry 4.0 concept, the aim of this PhD thesis is to help
companies to evolve from traditional automated manufacturing to the digital manufacturing level by proposing a
novel integration model of Industry 4.0. The variety and the complexity of Industry 4.0 aspects, along with
technical and managerial challenges are to be taken into consideration to provide companies with successful
integration approaches along with a novel integration framework. The experimentation following this research has
been realized at the platform 4.0 at Arts et Métiers Institute of Technology, Paris. The developed model and the
proposed framework have been verified through an intensive implementation and integration work among the
experimentations done at the industrial platform 4.0.
Keywords: Industry 4.0, Cyber-Physical Systems (CPS), Internet of Things (IoT), Integration Models, Digital
Transformation, Integration Framework, Flexibility, Modularity, Data Deployment, Advanced Robotics,
Manufacturing Execution System.
3
4
ÉCOLE DOCTORALE SCIENCES DES MÉTIERS DE L’INGÉNIEUR
[Laboratoire Conception de Produits et Innovation – Campus de Paris]
et co-encadrée par :
[Mr BENFRIHA Khaled]
5
6
Table des matières
1. Introduction ............................................................................................................................ 9
2. Contexte général de la 4ème révolution industrielle ..................................................................... 11
2.1 La numérisation des entreprises et des systèmes de production, un besoin industriel .... 11
2.1.1 Contexte de l’évolution industrielle 4.0 ....................................................................... 11
2.1.2 Changements et avantages de la numérisation des entreprises ................................. 13
2.2 Objectifs de la recherche dans le contexte de l’Industrie 4.0 et la transition numérique
des entreprises ......................................................................................................... 14
2.2.1 Continuité numérique des produits et les axes de recherche LCPI .............................. 14
2.2.2 Plateforme de production robotisée 4.0 ...................................................................... 15
2.3 Positionnement général de la présente thèse du doctorat ................................................ 16
3. Contexte scientifique et problématique de recherche ............................................................ 17
3.1 Analyse des aspects de l'intégration des principes de l'industrie 4.0 dans les entreprises
de production ........................................................................................................... 17
3.2 Périmètre de l’intégration des principes de l’industrie 4.0 dans les entreprises de
production................................................................................................................ 18
3.3 Projection des concepts de l'industrie 4.0 au cycle de vie du produit et positionnement de
la problématique de recherche ................................................................................. 20
3.4 Problématique, la question de la recherche ..................................................................... 21
4. La numérisation des systèmes de fabrication dans le contexte de l’industrie 4.0 – Etat de
l’Art ................................................................................................................................... 24
4.1 Intégration vertical dans le contexte de l’industrie 4.0................................................. 24
4.2 Revue de la flexibilité et de la modularité dans le contexte de l’Industrie 4.0............. 26
4.3 L’Evolution de Technologies de L’information et des Communications (ICT) et la
structure hiérarchique traditionnelle ........................................................................ 28
4.4 Intégration verticale et fonctionnalités de l’industrie 4.0 ............................................. 30
4.5 Analyse des architectures 4.0 existantes ......................................................................... 32
4.6 Synthèse de positionnement par rapport à l’état de l’art ........................................... 34
5. Hypothèse ............................................................................................................................. 36
5.1 Hypothèse proposée ........................................................................................................ 36
5.3 Démarche scientifique de la résolution de l’hypothèse proposée ................................ 38
6. Modélisation des systèmes de production 4.0 ............................................................................ 40
7
6.1 Modélisation générale des systèmes de production 4.0 .................................................. 40
6.2 Modélisation des systèmes de production 4.0, cas d’un système de production existant
................................................................................................................................. 45
6.3 Framework de la transition numérique des systèmes de fabrication 4.0 ......................... 47
7. Expérimentations, application à la plateforme industrielle 4.0 ................................................... 51
7.1 Objectifs et Protocole de l’expérimentation ................................................................. 52
7.2 Déploiement du Framework proposé .............................................................................. 53
7.2.1 Atelier flexible à la plateforme 4.0 : ............................................................................. 53
7.2.2 Système de commande (contrôle) flexible de la plateforme 4.0 : ............................... 55
7.2.3 Intégration des services modulaires :........................................................................... 56
7.2.4 Gestion intégrée des données : .................................................................................... 58
7.2.5 Ouverture pour une communication au niveau business de l'entreprise et les lignes de
production distribuées : ........................................................................................................ 59
7.3 Architecture innovante de la plateforme industrielle 4.0 ................................................ 59
7.3.1 Architecture système de la plateforme industrielle 4.0 ............................................... 60
7.3.2 Architecture informatique de la plateforme industrielle 4.0 ....................................... 61
7.3.3 Modèle basé sur le niveau de décision des machines de la plateforme industrielle 4.0
............................................................................................................................................... 62
7.3.4 Synthèse ....................................................................................................................... 63
7.4 Une méthodologie de conception pour la configuration des processus modulaires ........ 64
7.5 Application de l’approche modulaire mise en œuvre dans le programme de production
................................................................................................................................. 67
7.5.1 Opérations paramétriques standardisées .................................................................... 68
7.5.2 Modules ........................................................................................................................ 71
7.5.3 Orchestration du plan de production ........................................................................... 73
7.5.4 Planification des processus de production................................................................... 75
7.6 Application démonstrative .............................................................................................. 76
7.6.1 Etape n. 1 : .................................................................................................................... 77
7.6.2 Etape n. 2 : .................................................................................................................... 77
7.6.3 Référentiel de mesure proposé pour l’évaluation quantitative de la nouvelle
performance : ........................................................................................................................ 79
7.6.4 Analyse et discussion des résultats : ............................................................................ 82
8. Résultats et discussion générale ................................................................................................. 85
9. Conclusion et perspectives ......................................................................................................... 88
References ..................................................................................................................................... 91
8
1. Introduction
L'industrie européenne est confrontée à des défis économiques essentiels en raison du rythme croissant
des développements sociétaux et technologiques, tels que la diminution de la disponibilité des ressources
naturelles, l'augmentation des prix de l'énergie, et la globalisation des marchés. La production
industrielle est aujourd'hui tirée par la concurrence mondiale et la nécessité d'une adaptation rapide de
la production aux demandes en constante évolution. Les consommateurs sont de plus en plus exigeants
pour une innovation produit-service améliorée, une variété de produits, des normes de qualité plus
élevées, des services d'assistance et l'immédiateté ou la satisfaction des commandes. Ces défis
conduisent les entreprises industrielles à être capables de gérer l'ensemble de leur chaîne de valeur de
manière agile et réactive.
Ces exigences ne peuvent pas être satisfaites que par des avancées radicales dans les processus de
fabrication actuels, ainsi que par l'intégration de tous les acteurs de la chaîne de valeur de l'entreprise
(fournisseurs et clients). Les aspects techniques de ces exigences sont abordés par l'application des
concepts génétiques des systèmes cyber-physiques (CPS), de l'Internet des objets industriels et de
l'intégration des nouvelles fonctionnalités et technologies aux systèmes de production industriels. La
représentation d'objets physiques tels que les machines, les outils, les pièces et leurs équipements dans
le monde numérique, avec la possibilité de communiquer avec d'autres objets, permet de construire un
système de contrôle de la planification, un pilotage de la production et enfin une logistique plus flexible,
plus efficace et plus transparente. Précisément, les entreprises ont besoin à la fois de structures virtuelles
et physiques qui permettent une coopération étroite et une adaptation rapide tout au long du cycle de vie
du produit, de l'innovation à la production et à la distribution (Erol and Schumacher, 2016). Cela
nécessite également des changements au niveau des différentes couches du système de production de
l’entreprise, des systèmes de contrôle au système d’information en termes de planification, d’exécution,
de vente et enfin de la gestion.
Ainsi, tous les fabricants doivent se préparer à entreprendre cette révolution industrielle potentielle pour
rester compétitifs sur un marché turbulent et hyperconcurrentiel. Les innovations technologiques et les
changements dans l'environnement des affaires affectent à la fois la performance à court terme et la
pérennité à long terme des entreprises. Lorsque les orientations et les options technologiques futures
sont incertaines, les entreprises doivent formuler une stratégie appropriée pour soutenir leur planification
avec les développements technologiques futurs tels que l'industrie 4.0. Du point de vue tant stratégique
que technologique, la transition vers l'industrie 4.0 nécessite un modèle stratégique complet qui visualise
chaque étape supplémentaire vers une entreprise de fabrication entièrement numérique.
Un problème majeur survient lorsqu'une entreprise décide de passer des fonctionnalités 3.0 à 4.0, et la
question principale est de savoir comment entreprendre cette transformation aux niveaux intégration,
9
outils et stratégie. L’intégration de nouvelles technologies et la mise au niveau des lignes de production
existantes doivent être étudié en profondeur. Ainsi, un processus d'intégration doit être développé avec
soin. De plus, le point initial dans cette démarche consiste à identifier le modèle de la transition
numérique et ses limitations.
La présente thèse de doctorat vise à offrir un modèle de transition qui peut être utilisé comme outil
conceptuel, par les académiciens et les industriels, dans la perspective d’une transformation réussie de
la fabrication traditionnelle vers la numérisation des systèmes de production dans le contexte de
l’Industrie 4.0. Il est évident qu'une compréhension approfondie des spécifications de l'Industrie 4.0 est
un préalable au développement du modèle de transition numérique et technologique. Notre démarche
scientifique consiste à développer un modèle conceptuel et l’appliquer sur un système de production
traditionnel afin d'identifier les évolutions requises et de définir en conséquence un Framework de la
transition numérique. Le Framework proposé sera appliqué à la plateforme industrielle 4.0 des Arts et
Métiers de Paris afin d’expérimenter et valider notre approche scientifique.
L’agilité offerte par l’industrie 4.0 permet aux usines de gérer une gamme de produits plus diversifiée
avec un délai de livraison plus court et de personnaliser rapidement les produits en fonction des
exigences spécifiques, garantissant ainsi une plus grande réactivité aux besoins des clients. Alors que
l'Industrie 4.0 offre d'énormes et nouvelles opportunités pour les entreprises ; son impact apporte
également de nouveaux défis. La question dynamique de la structure verticale traditionnelle de
production doit être analysée, puisque les systèmes de production doivent d'abord être capables de
s'adapter aux nouveaux produits et de répondre rapidement aux fluctuations de la demande.
Deuxièmement, les systèmes de production doivent permettre de déployer des données collectées à partir
des machines et des capteurs afin que le système optimise ses processus.
Ce document est organisé en neuf chapitres. Après l'introduction, le chapitre 2 présente le contexte
général de cette thèse de doctorat en termes de besoin industriel et besoin de recherche. Le chapitre 3
identifie le positionnement de la recherche ainsi que la problématique de la recherche. Le chapitre 4 est
dédiée à l’étude bibliographique et aux défis auxquels est confronté le processus de la numérisation des
systèmes de fabrication dans le contexte de l’industrie 4.0. Le chapitre 5 introduit l’hypothèse proposée.
Le chapitre 6 présente le modèle suivi du Framework de la transition numérique. Dans le chapitre 7, une
nouvelle architecture de transition de la fabrication numérique est développée et expliquée dans le cadre
des expérimentations, suivi d’une application démonstrative. Une discussion générale des résultats sera
menée dans le chapitre 8. Le chapitre 9 conclut la thèse de doctorat tout en discutant les résultats et les
apports scientifiques.
10
2. Contexte général de la 4ème révolution industrielle
Ce chapitre a pour objectif d’établir le lien entre les besoins industriels vis-à-vis la transition
numérique des entreprises et l’objectif de notre recherche scientifique afin de préciser le
positionnement général de la présente thèse de doctorat. Le contexte général de la 4 ème révolution
industrielle ainsi que les aspects économiques et industrielles qui accompagnent l’émergence de
l’Industrie 4.0 sont introduits dans la section 2.1. Le besoin industriel figurant dans la transition
numérique identifié dans la section 2.1.1 est justifié par les changements et les avantages de la
numérisation des entreprises identifiées dans la littérature qui seront présentés dans la section 2.1.2.
La section 2.2 présente par la suite le besoin de recherche vis-à-vis la nouvelle révolution industrielle
tout en positionnant parmi les axes de recherche du Laboratoire de Conception de Produits et
Innovation (LCPI), et en introduisant le projet de recherche de la plateforme 4.0. La section 2.3
positionne l’objectif de la présente thèse de doctorat parmi les axes de recherche définis.
11
Ford T-Model "Vous pouvez avoir n'importe quelle couleur tant qu'elle est noire", il s’agit bien d’une
production de masse, sans possibilité de personnalisation des produits. La troisième révolution
industrielle se caractérise par l'introduction de la microélectronique et de l'automatisation (Rojko,
2017). C'est l'ère des technologies de l'information. Des systèmes de gestion intégrés IMS ont été
introduits pour réduire considérablement les coûts de gestion, ce qui représente l'un des changements
majeurs dans le monde industriel. L'ensemble de la structure de l'entreprise a été affectée et réorganisée
en fonctions de leur cœur de métier et séparées tels que le département de recherche et développement,
les finances et le marketing (Santos et al., 2017).
Aujourd'hui, nous vivons la quatrième révolution industrielle, appelée « Industrie 4.0 » qui diffère de
la révolution industrielle précédente en termes de concept et d'outils. C'est la première révolution
industrielle à prendre en compte les ressources énergétiques renouvelables pour alimenter nos
processus de production. Demain, les usines 4.0 seront intégrées dans des villes intelligentes et
alimentées par l'énergie éolienne, solaire et géothermique selon (SenTryo, 2017). Du côté
technologique, l'industrie 4.0 ne consiste pas à découvrir un nouveau phénomène mécanique,
électrique ou même de communication, mais une véritable combinaison entre de nombreuses
technologies nouvelles. Grâce à l’utilisation élargie de l'Internet des objets (IoT), à l'introduction de
Cyber Physical Systems (CPS), Cloud Manufacturing (CM), Additive Manufacturing (AM) et de
nombreuses autres technologies et actionneurs, un nouveau niveau d'industrie et de fabrication pourra
être atteint. Cette révolution industrielle se caractérisent principalement par la connectivité et
l'interopérabilité, la flexibilité, l'intelligence, la modularité et la personnalisation de produits de masse.
En d’autres termes, l'industrie 4.0 aide les entreprises à produire des produits de plus en plus
individualisés avec un délai de mise sur le marché court dans le respect des exigences de la qualité.
Par ailleurs, la fabrication intelligente joue un rôle important dans l'industrie 4.0, les ressources sont
converties en objets intelligents afin qu'ils puissent agir, ressentir et se comporter dans un
environnement intelligent (Zhong et al., 2017).
L'industrie 4.0 est une révolution industrielle enracinée dans un nouveau phénomène technologique -
la numérisation. Cette numérisation / digitalisation nous permet de construire un nouveau monde
virtuel à partir duquel nous pouvons diriger le monde physique. Les industries d'aujourd'hui visent à
connecter tous les moyens de production pour permettre leur interaction en temps réel. Les usines 4.0
rendent possible la communication entre les différents acteurs et objets connectés d'une ligne de
production grâce à des technologies telles que le Cloud, le Big Data Analytics et l'Internet des objets
industriel (SenTryo, 2017).
Trois grands dimensions économiques et industrielles sont à l'origine de l'émergence de l'Industrie 4.0.
La diminution de la disponibilité des ressources naturelles, l'augmentation des prix de l'énergie,
12
l'augmentation de l'âge des employés et la mondialisation des marchés sont des défis majeurs dans le
contexte économique. Parallèlement, les demandes des clients deviennent de plus en plus difficiles,
allant de l'amélioration de l'innovation produit-service à la variété de produits avec des normes de
qualité élevées et des services d'assistance renforcés, ainsi que l'immédiateté et la satisfaction des
commandes. Outre les défis majeurs susmentionnés, l'avancement des technologies telles que la
disponibilité des services Internet, la large utilisation des objets connectés et de l'Internet des objets
(IoT) ainsi que le développement des technologies de l'information, l'infrastructure et les réseaux
informatiques sont des catalyseurs clés de la numérisation des systèmes de fabrication. Dans ce
contexte, la numérisation et l'intégration de l'Industrie 4.0 dans les systèmes de fabrication est une
solution réfléchie contribuant à améliorer la performance industrielle et répondant aux enjeux
économiques et industriels. L'évolution de l'Industrie 4.0 est progressive, cependant, son impact est
considérable sur la performance de l'entreprise, la productivité, les marchés et l’industrie en général
(Santos et al., 2017).
2.1.2 Changements et avantages de la numérisation des entreprises
Selon (Bodrow, 2017), la survie économique de chaque entreprise dépend de sa volonté de changer.
L'industrie 4.0 applique des changements substantiels qui affectent les petites, moyennes et grandes
entreprises, en particulier dans les domaines de la production, la logistique, et l’optimisation des
processus de production basée sur le traitement des données.
Selon McKinsey and Company, l'Industrie 4.0 est l'intégration au secteur manufacturier d’une couche
IoT en utilisant les capteurs et actionneurs, en connectant les systèmes de production à la chaîne
d'approvisionnement et aux fonctions de demande du marché par des pouvoirs analytiques. Cela aura
pour conséquence de réduire les risques contractuels et de limiter les interventions humaines (Santos
et al., 2017). Ainsi, les structures de gestion sont censées être plus décentralisées, bien que plus
spécialisées.
Selon (Schlapfer et al., 2015), entre 20 et 50% d'économies pourraient être réalisées avec une agilité
accrue. Beaucoup pourrait être économisé dans la reconfiguration des équipements de production mais
encore dans les changements dynamiques des produits.
(Santos et al., 2017) ont répertorié les quatre principaux avantages de l'Industrie 4.0 : Développer des
processus industriels et des chaînes d'approvisionnement verts et durables, créer un nouvel avantage
stratégique sur le marché, réduire les coûts grâce à des machines autonomes et assurer un
développement plus rapide des produits.
(Rojko, 2017) a mentionné que l'industrie 4.0 pourrait entraîner une diminution de 10 à 30% des coûts
de production et de logistique et de 10 à 20% des coûts de gestion de la qualité. De plus, de nombreux
13
avantages peuvent être mis en avant :
- Un délai de mise sur le marché plus court pour les nouveaux produits.
- Permettre une production de masse personnalisée sans augmenter significativement les coûts de
production globaux.
- Environnement de travail plus flexible et plus convivial.
- Utilisation plus efficace des ressources naturelles et de l'énergie.
Partant du contexte industriel qui consiste à avoir un guide permettant à l’entreprise d’adopter une
transition numérique réussie vers la digitalisation des systèmes de fabrication en tenant compte du
système de production existant, la question majeure qui se pose n’est plus pourquoi, mais comment ?
Dans le contexte révolutionnaire de l’industrie 4.0, (Benfriha, 2020) a proposé une problématique de
recherche qui considère 2 axes principaux :
L’axe 1 porte le développement d’une démarche générique de transition vers les systèmes de production
digitalisée, dont l’objectif est d’aboutir à une approche progressive de la transition numérique des
entreprises (Benfriha, 2020). La transition numérique de l’entreprise nécessite une évolution du système
de production et l’intégration de nouvelles briques technologiques proposés par le 4.0. Cette approche
prend en considération les fonctionnalités recherchées, le niveau d’intégration et les performances visées
par l’entreprise, qui par conséquent définit le périmètre de cette transition numérique en termes
d’investissement au niveau technologique.
L’axe 2 aborde la phase de l’optimisation simultanée de la conception d’un produit et de son processus
de fabrication en se basant sur l’exploitation des données issues d’un processus de production digitalisé.
La fabrication digitalisée génère une masse de données issues de la couche IoT et d’autres processus
inhérents aux phases d’industrialisation et de production. L’exploitation de ces données collectées peut
générer des informations utiles à l’optimisation simultanée des processus de conception et de production
d’un produit en termes de couts, de consommation énergétique et de temps (lead-time). Cet axe renforce
la continuité numérique entre différentes phases du cycle de vie du produit.
14
Les activités de recherche du Laboratoire Conception de Produits et Innovation (LCPI, EA 3927) de
l’HESAM, Paris, s’inscrivent essentiellement dans le domaine du génie industriel telle que le montre
la Figure 1. Par conséquent, nos travaux de recherche portent sur l'intégration des concepts et
technologies 4.0 dans les systèmes de fabrication, et ont pour objectifs de développer une démarche
scientifique d’aide à la décision pour la transition vers la digitalisation des systèmes de fabrication.
Fig. 1 Axes de recherches et des activités du Laboratoire Conception de Produits et Innovation LCPI
Les universités, les centres de recherche et les entreprises s'intéressent de plus en plus à l'industrie 4.0
en termes de recherches méthodologiques qui contribuent à la mise en œuvre de ces nouvelles
technologies. L'objectif derrière ces efforts est d'exploiter le potentiel des nouvelles technologies et
concepts, en particulier l'IoT, les processus techniques et commerciaux, la virtualisation, la flexibilité,
l'interopérabilité, et l’intelligence artificielle (Rojko, 2017) afin de définir une approche
méthodologique pour accompagner les entreprises dans leur projet de transitions vers la digitalisation.
Dans ce cadre, la plateforme industrielle 4.0 a été développé pour faire partie des plateformes de
compétences du laboratoire CPI au sein des Arts et Métiers de Paris, en partenariat avec Le centre de
formation d'apprentis (CFA) ingénieurs 2000. Cette plateforme permet d’expérimenter des modèles et
architectures de systèmes de production complexe. Le 2ème objectif est à caractère pédagogique. En
15
effet, sur la base des travaux de recherche menés, il devient possible de proposer des parcours
innovants d’enseignement et de mise en situation pour l’apprentissage des principes de l’industrie 4.0.
Parmi les deux axes identifiés dans la section 2.2.1, la présente thèse de doctorat est la première thèse
menée au niveau de la plateforme 4.0 en considérant l’axe 1 : Transition vers les systèmes de production
4.0. L’objectif de cet axe consiste à développer une démarche scientifique de transition progressive vers
la numérisation de la production, offrant ainsi de nouvelles fonctionnalités en rupture avec les modèles
actuels. Il s’agit d’une approche générale qui s’impose et qui est adaptée à la trajectoire que l’entreprise
se fixe, compte tenu des performances souhaitées, des fonctionnalités visées, du gap technologique
attendu et de l’investissement budgétaire de l’intégration.
L’originalité de ces travaux de recherche réside dans la considération de toute les couches d’un système
de production y compris le niveau technologique mais aussi bien l’architecture informatique et la couche
de commande. Évidemment, l’aspects technologique, n’est pas le seul facteur de réussite, d’autres
facteurs sont tellement critiques pour l’aboutissement d’un tel projet spécifiquement l’évolution des
compétences. Cependant, seul l’aspect technologique sera abordé dans ce mémoire.
16
3. Contexte scientifique et problématique de recherche
3.1 Analyse des aspects de l'intégration des principes de l'industrie 4.0 dans
les entreprises de production
En se basant sur la littérature et aux axes de recherche identifiés, trois aspects principaux pourraient être
associés à l'intégration des principes de l'Industrie 4.0 dans les entreprises, illustrés dans le Tableau. 1.
Tableau. 1 Aspects de l'intégration de l'Industrie 4.0 dans les entreprises (Bodrow, 2017)
Selon (Bodrow, 2017), trois aspects principaux sont nécessaires pour réaliser l'intégration des principes
de l'Industrie 4.0 et pour appliquer potentiellement ses fonctionnalités promises dans les entreprises de
fabrication. Les changements dans la production sont dus à l'application des systèmes cyber-physiques
et de l'IoT industriel au travers des contrôleurs décentralisés via un réseau avancé. Les changements
dans la logistique sont pris en compte en adaptant les systèmes d'exécution de fabrication MES et les
systèmes de gestion des Ressources et Plannings ERP pour renforcer les processus de fabrication et
d'affaires en ajoutant des niveaux plus élevés de flexibilité en termes de management et de connectivité.
Enfin, l'influence du Big Data et du Cloud concerne le traitement et le déploiement de données collectées
17
à partir des différents processus de production, des machines et des produits par l’utilisation du Edge
Computing et du Cloud Computing.
La classification de ces aspects proposés par (Bodrow, 2017) est soutenue par (Dalenogare et al., 2018),
est basée sur les trois avantages principaux qui caractérisent l'industrie de futur d'un point de vue
intégration technologique: l'intégration verticale, l'intégration horizontale et la ‘end-to-end’ ingénierie.
L'intégration verticale fait référence à l'intégration des systèmes de technologies de l'information et de
la communication (ICT) à différents niveaux hiérarchiques d'une organisation, notamment entre les
niveaux de production et le niveau de gestion. D'autre part, l'intégration horizontale consiste en une
collaboration entre les entreprises au niveau d'une chaîne d'approvisionnement, avec un échange de
ressources et d'informations en temps réel. Enfin, l'ingénierie end-to-end est l'intégration de l'ingénierie
dans toute la chaîne de valeur d'un produit, de son développement jusqu'à l'après-vente. Cette dernière
dimension est assurée principalement par le traitement et le déploiement des données et de leur
traitement dans les systèmes de fabrication.
(Frank et al., 2019) ont affirmé que le concept de l'Industrie 4.0 représente une nouvelle étape
industrielle des systèmes de fabrication par l’intégration d’un ensemble de technologies émergentes et
convergentes qui ajoutent de la valeur à l'ensemble du cycle de vie du produit. En référence à (Frank et
al., 2019), les technologies de l'Industrie 4.0 peuvent être séparées en deux couches différentes en
fonction de leur objectif principal:
18
base permettent aux technologies frontales d'être connectées dans un système de fabrication intégré et
complet.
Fig. 2. Structure / Cadre théorique des technologies de l’industrie 4.0 (Frank et al., 2019)
Cette nouvelle initiative industrielle exige une évolution sociotechnique du rôle humain dans les
systèmes de production, dans lequel toutes les activités de travail de la chaîne de valeur seront réalisées
avec des approches intelligentes (Smart Working) et ancrées dans les technologies de l'information et
de la communication (ICT).
L'industrie 4.0 est enracinée dans la fabrication avancée ou également appelée concept de fabrication
intelligente, ce qui signifie un système adaptable où les lignes flexibles ajustent automatiquement les
processus de production à plusieurs types de produits et à des conditions changeantes. Cela permet
d'augmenter la qualité, la productivité et la flexibilité et peut aider à réaliser des produits personnalisés
à grande échelle.
Les technologies intégrées dans les produits finaux (produits intelligents) font également partie du
concept plus large de l'industrie 4.0. Les produits intelligents peuvent fournir des informations en retour
pour le développement de nouveaux produits et de fournir de nouveaux services et solutions au client.
Ainsi, certains chercheurs considèrent les produits intelligents comme le deuxième objectif principal de
19
l'Industrie 4.0, car ils permettent de nouveaux modèles commerciaux tels que les systèmes produit-
service, qui créent de nouvelles opportunités pour les fabricants et les prestataires de services.
Au cœur du concept de l'Industrie 4.0, les technologies de fabrication intelligente constituent le pilier
central des activités opérationnelles internes, tandis que les produits intelligents prennent en compte
la valeur ajoutée externe des produits, lorsque les informations et les données des clients sont intégrées
au système de production. La fabrication intelligente considère les technologies pour la fabrication des
produits (système de production), tandis que les produits intelligents considèrent les technologies liées
à l'offre de produits.
Par conséquent, nous supposons que la fabrication intelligente est le point de départ et le premier
objectif de l'industrie 4.0, tandis que le produit intelligent est son extension. Cette vision suit
l'évolution chronologique récente du concept Industrie 4.0, qui trouve ses racines d'abord dans
les systèmes de fabrication avancés et ses connexions avec d'autres processus de l'entreprise. En
ce qui concerne la dimension Smart Manufacturing, les principales technologies associées
comprennent six objectifs principaux : l'intégration verticale, la virtualisation, l'automatisation,
la traçabilité, la flexibilité et la gestion de l'énergie, et seront abordées dans la section état de l’art.
Sur la base de la catégorisation proposée par (Frank et al., 2019), chaque dimension pourrait être associée
à une phase du cycle de vie du produit d'un point de vue technique. Comme le montre la Figure 3,
l'intégration des aspects de l'Industrie 4.0 dans les entreprises de fabrication est associée aux différentes
phases du cycle de vie du produit. L'objectif de cette figure est de mettre en évidence notre intérêt pour
la dimension Fabrication et donc de clarifier notre positionnement de recherche axé sur la
numérisation des systèmes de fabrication dans le contexte de l’industrie 4.0 et notamment en
termes de transformation digitale des entreprises. L'aspect Smart Product concerne principalement
la phase de conception en adaptant le Product-Service-Systems (PSS) comme modèle de conception. Ce
20
modèle est basé sur l’intégration des capteurs et des logiciels au produit dans la perspective d'intégrer le
concept d'orientation service aux produits connectés. Le Smart Working Environment concerne
essentiellement la phase d'industrialisation du cycle de vie du produit, en intégrant des outils innovants
de support aux opérateurs et en assistant les ingénieurs de production dans les tâches de la production et
de la planification. L'aspect Systèmes de Production Intelligents concerne directement la phase de
fabrication, et représente l'objectif principal de ces travaux de recherche qui seront discutés dans la
section suivante. Le système de production intelligent ou le système de fabrication flexible concerne
principalement la flexibilité et l'intégration verticale. Enfin, la Smart Supply Chain est associée aux
services liés au produit sur l'ensemble de la chaîne de valeur concernée principalement par l'intégration
horizontale.
Fig. 3 Projection de l’intégration des principes de l’Industrie 4.0 dans les entreprises de production
En synthèse, nos travaux de recherche portent sur l'intégration des concepts et technologies 4.0 dans les
systèmes de fabrication. A cet effet, nous sommes amenés à développer un modèle d'aide à la décision
pour aider les entreprises dans la numérisation des systèmes de fabrication dans le but d’améliorer la
flexibilité et la performance des systèmes et processus de production, ainsi que la qualité du produit
grâce aux données générées et à leur exploitation.
Les entreprises de production peuvent se concentrer sur les différents besoins qu'elles peuvent avoir
lorsqu'elles accordent la priorité à la mise en œuvre des technologies de fabrication intelligente
susmentionnées afin d’atteindre un niveau de maturité compatibles avec l'Industrie 4.0.
Cela suggère que les technologies de fabrication intelligentes peuvent être interdépendantes et créer une
synergie active pour l'industrie 4.0. Cette intégration synergique « Smart Manufacturing » soutenues par
l’intégration technologiques comme outils de la ‘numérisation’ s’insère potentiellement dans la
21
configuration et l’évolution fonctionnelle des ‘processus de production’, ce qui produit un effet direct à
l’évolution de l’architecture du système de fabrication. Les nouveaux systèmes de fabrication sont en
mesure d’ajuster leurs capacités de production et d’utiliser des fonctionnalités avancées. La flexibilité
des systèmes de fabrication s'applique non seulement aux composants matériels d'une usine tels que les
robots et les machines, mais également à son infrastructure des technologies de l'information et de la
communication (ICT) composée de ressources de calcul, de stockage et de réseaux.
L’IoT aboutit aux systèmes cyber-physiques CPS qui consiste à intégrer des objets physiques de l'usine
avec la dimension virtuelle de l'usine, y compris les données intégrées, l'intelligence artificielle et la
simulation. Par conséquent, les entreprises manufacturières qui visent un niveau de maturité plus élevé
de l'Industrie 4.0 devraient adopter systématiquement la plupart des technologies de fabrication
intelligente, car ces technologies sont interdépendantes. L’adoption de cette technologie avancée aboutit
à son tour à la génération de données, qui une fois traitées, contribuent à améliorer et optimiser les
processus de production.
Cette analyse nous conduit à définir le positionnement de notre problématique de recherche Quel
modèle pour une transition réussie vers la numérisation des systèmes de fabrication 4.0 ? qui se
situe au croisement de trois domaines définis par la Figure 4, la ‘numérisation’, le ‘système de
fabrication’, et les ‘processus de production’. L’évolution du système de fabrication dans le contexte de
l’Industrie 4.0 est considérée comme le point de départ pour répondre à la question de recherche
identifiée. Cette évolution devra tout d’abord prendre en considération la situation actuelle du système
de fabrication pour établir par la suite sa transition numérique en lui assurant des fonctionnalités
avancées. La numérisation se figure dans les outils technologiques proposés par l’Industrie 4.0 pour
appliquer les fonctionnalités visées au système de production. Les fonctionnalités et les technologies
avancées intégrées dans le système de production s’imbriquent apparemment dans l’évolution et la
numérisation des différents processus de production dans l’objectif d’améliorer leur performance.
22
Fig. 4 Positionnement de la problématique de recherche
La définition de ces trois dimensions nous aide à orienter notre direction de recherche afin de répondre
à la problématique identifiée. A ce stade, notre démarche consiste à dresser un état de l’art élargi afin
d’identifier les concepts fondamentaux nécessaires à la construction d’un modèle de système de
production 4.0 dans l’objectif de guider les entreprises dans la transition numérique des systèmes de
production. Ces concepts de natures différentes mais complémentaires devraient s’imbriquer,
s’interfacer et s’interconnecter afin de créer un système cohérent productif et assez détaillé pour un
déploiement pragmatique. Le modèle proposé sera accompagné par un Framework de transition
numérique des systèmes de production comme outil de support visant à guider l'intégration numérique
4.0 dans les systèmes de production existants.
23
4. La numérisation des systèmes de fabrication dans le contexte de
l’industrie 4.0 – Etat de l’Art
Les défis auxquels est confrontée l'intégration des technologies 4.0 dans les systèmes de fabrication
peuvent être classés en deux dimensions principales, la dimension stratégique et la dimension technique.
La première dimension concerne la stratégie et la perspective d’évolution de l’entreprise en ce qui
concerne ses activités principales. Les technologies 4.0 et la numérisation poussent considérablement
les entreprises à s’adapter au marché saturé et mondialisé, où la production en quantité « production de
masse » n’est plus adaptée à la demande de plus en plus individualisée « personnalisation de masse ».
Face à ce problème, la réactivité et l'adaptabilité des systèmes de production sont des atouts majeurs
face à ce type de demande (Erol and Schumacher, 2016). Ainsi, l'évolution des systèmes de production
devient nécessaire et impérative afin de s'adapter aux défis commerciaux. Par conséquent, des défis
techniques se posent, plusieurs travaux ont indiqué que les investissements élevés, la complexité, le
savoir-faire requis ainsi que l'inadéquation de l'infrastructure informatique existante et des technologies
perturbent l'intégration verticale et horizontale. Ces défis nous conduit à vérifier la capacité des systèmes
de fabrication existants à s'adapter à la numérisation en tenant compte de son architecture et de son
infrastructure informatique.
La section suivante est consacrée au classement des enjeux liés à l'intégration de la technologie 4.0 dans
les systèmes de fabrication qui s’imbriquent dans la dimension de ‘l’intégration verticale’. Ce
classement prend en considération les trois domaines identifiés dans le chapitre précèdent ‘Système de
Production’, ‘Numérisation’ et ‘Processus de Production’ afin de répondre à notre problématique de
recherche. Les concepts principaux identifiés dans la section 4.1 précisément la flexibilité, la modularité,
les technologies de l’information et des communications (ICT), et l’intégration verticale sont présentés
respectivement dans les sections qui suivent. Ensuite, une analyse bibliographique des architectures et
modèles récentes proposées dans la littérature sera menée dans la section 4.5. Finalement, la section 4.6
apporte une synthèse de positionnement par rapport à l’état de l’art.
Les résultats détaillés de l'état de l'art concernant les défis stratégiques et techniques nous ont conduit à
identifier les concepts principaux de l'évolution des systèmes de fabrication dans le contexte de
l'Industrie 4.0. Ces concepts s’agissent d’une superposition de principes et de briques technologiques
qui concernent les différents domaines affectés par la numérisation des systèmes de production comme
le montre la Figure 5.
24
Fig. 5 Concepts principaux de la numérisation des systèmes de production
En fait, la diversité des défis et problématiques d'intégration identifiées, nous ont incité à entreprendre
un travail de synthèse et de classification. La Figure 6 classe les défis identifiés ci-dessus et les attribue
à un sujet central, en considérant la dimension de l'intégration verticale comme l’enjeu principal de la
transition numérique des systèmes de production. Tout d'abord, les ICT et le système informatique de
production sont une dimension essentielle identifiée dans la littérature, considérés comme un principal
frein aux évolutions nécessaires pour s'adapter à la nouvelle configuration du système de fabrication
numérique (Xu et al., 2018a). Parallèlement, la conception du système informatique porte un autre défi,
en particulier lorsqu'il s'agit de faire face aux technologies de pointe, de mettre en œuvre des services
spécifiques dans le système de fabrication (Bodrow, 2017) ou encore d'adapter le cloud computing pour
améliorer l'analyse des données (Kamble, 2018). Deuxièmement, la modularité des systèmes de
fabrication est un autre défi majeur identifié par (Kaushal et al., 2016). Le concept de modularité
comprend plusieurs aspects tels que le développement des CPS et leur intégration dans le système de
fabrication (Goerzig and Bauernhansl, 2018) ainsi que le déploiement de la modularité dans le système
25
de contrôle de l'atelier SFCS (Mes and Gerrits, 2019). Troisièmement, le système de fabrication flexible
dans le contexte de l'industrie 4.0 est un défi majeur (Theorin et al., 2017), qui comprend la question de
la mise en œuvre des technologies avancées (Zhou et al., 2015), le déploiement des opérations flexibles
en atelier (Erol and Schumacher, 2016) et la configuration des machines connectées. Enfin, ces divers
défis clés pourraient tous être classés dans le périmètre de l’intégration verticale (Dalenogare et al.,
2018), qui concerne en outre la dimension infrastructure informatique (Rojko, 2017), la mise en œuvre
de l'IoT (Kamble et al., 2018), ainsi que les problématiques d'intégration de données (Rüßmann et al.,
2015).
Fig. 6 Cadre des défis d'intégration de l'industrie 4.0 dans les systèmes de fabrication
Suite à cette analyse des concepts et des défis de la numérisation des systèmes de production, nous
proposons dans la section suivante un état de l'art dédié aux concepts de la flexibilité et de la modularité.
Dans la section 3, nous passons en revue le système de contrôle de l'atelier et la structure ICT
traditionnelle. L’intégration verticale en faveur des fonctionnalités 4.0 est présentée dans la section 4.
26
Plusieurs recherches conventionnelles ont été menées dans le cadre de la personnalisation du produit
notamment sur le principe de la reconfiguration des plans de production avant l’émergence de l’Industrie
4.0. (Schierholt, 2001) a identifié le concept de reconfiguration de processus qui vise à simplifier la
génération de plan de production pour une nouvelle variante de produit. A cet effet, Schierholt
(Schierholt, 2001) a identifié deux principes pour les systèmes de reconfiguration des plans de
production : la reconfiguration interactive et la configuration basée sur l'automatisation. (Zheng et al.,
2008) ont proposé un « Systematic Knowledge Model » pour plusieurs variantes de produits et ont
appliqué une méthode de reconfiguration rapide par l’application de règles standards pour la
reconfiguration. Une approche basée sur « Tree Unification » est proposée par (Zhang and Rodrigues,
2009) pour développer des plans génériques à partir des données de production en lien avec des familles
de produits, ce qui permet aux entreprises de répondre à une diversité de produits personnalisés au
moindre coûts et délais.
Alors que, dans le contexte de l'industrie 4.0, une reconfiguration de processus peut être envisagée grâce
à la flexibilité des systèmes de production. Cette nouvelle révolution industrielle se caractérise
principalement d'une part par des machines de production connectées et interopérables qui sont pilotées
intelligemment. Et d'autre part, par de nouvelles fonctionnalités telles que la flexibilité, la prédiction, la
modularité et la personnalisation de masse des produits. En d'autres termes, l'industrie 4.0 facilite la
reconfiguration des processus afin de fabriquer des produits de plus en plus individualisés avec un délai
de mise sur le marché court et conformes à la qualité requise. Beaucoup pourrait être économisé dans la
reconfiguration et la réutilisation des équipements de production ou encore des changements
dynamiques du produit selon (Schlapfer et al., 2015).
Cependant, plusieurs travaux de recherche sont en cours afin d'introduire des solutions rapides, efficaces
et rentables en vue d'améliorer la fabrication flexible dans les lignes de production existantes. Basé sur
des technologies de l'industrie 4.0 telles que ; le Big Data et l'analyse, les systèmes robotiques
autonomes, le cloud computing, l'Internet des objets industriel, la simulation et le prototypage, la
fabrication additive, la réalité augmentée, l’Intégration horizontale et verticale et la cybersécurité
(Rüßmann et al., 2015), le système de fabrication peut accroître la flexibilité des processus de fabrication
(Kamble et al., 2018).
D'un autre côté, (Mes and Gerrits, 2019) ont abordé le fait que les systèmes de contrôle hiérarchiques
ne sont pas en adéquation avec le besoin de produire une gamme diversifiée. Par conséquent, des
concepts de fabrication flexibles et reconfigurables ont été introduits sur le principe de modules de
commande autonomes et intelligents. Ces modules interagissent dynamiquement les uns avec les autres
pour atteindre des objectifs de performance. Certains industriels comme ABB (Bloch et al., 2018) ont
commencé à développer ce que l'on appelle le Modular Type Package (MTP), une initiative visant à
27
modifier l'infrastructure du processus de contrôle dans l'atelier, ouvrant la voie au concept « Small
Automation » tel que défini par (Strategy&, 2015). Selon (Mourtzis et al., 2016) la planification des
processus assistée par ordinateur (CAPP) doit prendre en compte des variantes de produit personnalisées
dans un environnement de fabrication dynamique, plutôt que de fonctionner indépendamment des
données de production de l'atelier et de souffrir du manque d'interfaçage avec les systèmes
informatiques. En revanche, ils ont proposé un système orienté services basé sur le cloud computing
pour faciliter la collaboration en temps réel entre la planification adaptative et l'exécution des processus
en tenant compte de la disponibilité des machines.
La flexibilité des machines de production et la modularité des opérations de fabrication sont deux
concepts indissociables pour réaliser de nouvelles fonctionnalités en termes de production flexible. Le
FMS est une exigence préalable à la mise en œuvre d'un processus modulaire d'opérations de fabrication.
Cela signifie que les machines fonctionneront indépendamment ou en coordination avec les humains
pour fabriquer des produits orientés client (Sung, 2018). Afin d'atteindre ces objectifs dans l'industrie,
une attention particulière doit être accordée à plusieurs niveaux de fabrication, en particulier les
machines (outils et équipements d'atelier) et les systèmes de contrôle et d'exécution de la fabrication.
Les technologies de l'information et des communications (ICT) mettent l'accent sur le rôle des
communications, des ordinateurs, des logiciels, du stockage et les plateformes numériques d’échange
entre les différents acteurs afin d’exploiter et agir sur les informations. La structure traditionnelle des
ICT dans les systèmes de production conventionnels est présentée par la pyramide CIM de fabrication
intégrée par ordinateur Fig. 7.
28
Au niveau de la gestion de l'entreprise, les commandes sont traitées dans le système ERP en termes de
planification des ressources d'entreprise, de planification des activités, de gestion de la chaîne
d’approvisionnement, des ventes et distribution et enfin en termes de comptabilité et de gestion des
ressources humaines. Dans les outils ERP traditionnels, le processus de décision est centralisé au plus
haut niveau de la pyramide d'automatisation. La plupart des solutions ERP disponibles ne prennent pas
en charge une adaptation rapide dans la planification de la production en raison d'événements imprévus
(Rojko, 2017).
Dans le cas d'une production entièrement automatisée, la flexibilité est un des défis considérables au
niveau du contrôleur. En effet, une fois le processus figé dans la couche API de l'interface de
programmation d'application, seule la reprogrammation des contrôleurs permet d'utiliser l'atelier pour
une autre tâche que celle déjà active. Pour chaque nouveau produit, il sera nécessaire de créer et de
redévelopper un nouveau programme, permettant de fabriquer un autre produit plutôt que l'ancien. Cette
situation représente l'architecture industrielle conventionnelle. (Karnouskos et al., 2010) ont affirmé que
la communication et la mise en réseau pour l'échange de données entre les appareils et les services, y
compris les capteurs, les appareils, les contrôleurs PLC, le MES et l'ERP, doivent être adaptés pour une
production dynamique au sein de l'Industrie 4.0. Selon (Theorin, 2012), les entreprises de fabrication
doivent reconfigurer les usines de production dans un court laps de temps, et par conséquent, des
dispositifs d'automatisation flexibles et adaptables - capables d'être faciles à intégrer, configurer, étendre
et réutiliser - sont nécessaires. Lorsque les tâches de contrôle deviennent de plus en plus complexes, les
systèmes de contrôle souffrent d’un manque de flux d'informations bien structurés entre la planification
des processus et leur mise en œuvre en raison du faible niveau d’implémentation des méthodes de
programmation spécifiques au fournisseur (Ollinger et al., 2014) qui peut générer un effort de
réingénierie des programmes de contrôle (Theorin, 2012).
29
Le dernier niveau de la pyramide d'automatisation est le niveau machines. Contrairement aux deux
couches supérieures, ce niveau dispose d’un contrôle naturellement indépendant (Rojko, 2017). D'un
autre côté et répondant aux initiatives de production flexibles, les machines doivent fonctionner
indépendamment ou en coordination avec les humains pour produire une fabrication orientée client. De
ce fait, la configuration de la machine doit être adaptée afin d’évoluer vers une entité indépendante
capable de collecter des données, de les analyser et de les interpréter (Sung, 2018).
Dans la continuité, l'industrie 4.0 est principalement représentée par le CPS, l'IoT et le cloud
computing (Henning et al., 2013; Hermann et al., 2016; Jasperneite, 2012; Lasi et al., 2014; Moeuf,
2018). L'IoT est le principal catalyseur des systèmes de fabrication avancés tels que le cloud
computing, de même, il est un support principal pour le développement d'usines intelligentes (Thames
and Schaefer, 2016). Le CPS permet un niveau plus élevé d'intégration et de coordination entre les
éléments physiques et informatiques (Gürdür et al., 2016). Les recherches indiquent qu'avec
l'introduction du CPS, les machines pourront communiquer entre elles en s’appuyant sur des systèmes
de contrôle décentralisés et donc contribuer à l’optimisation de la production. L'industrie 4.0 comprend
généralement de nombreux composants complexes et a de larges applications dans de nombreux
secteurs industriels (Xu et al., 2018b).
(Frank et al., 2019) ont subdivisé les technologies liées à la dimension Smart Manufacturing en six
objectifs principaux : intégration verticale, virtualisation, automatisation, traçabilité, flexibilité et
gestion de l'énergie.
L’intégration verticale de l’usine comprend des systèmes (ICT) avancés qui intègrent tous les niveaux
hiérarchiques de l’entreprise - de l’atelier aux niveaux intermédiaire et supérieur - aidant les actions de
prise de décision à être moins dépendantes de l’intervention humaine. Pour atteindre l'intégration
verticale, la première étape niveau atelier est la numérisation de tous les objets et paramètres physiques
avec des capteurs, des actionneurs et des automates programmables. Les données sont ensuite collectées
par le contrôle de supervision et l'acquisition de données (SCADA), afin de piloter la production et le
diagnostic en atelier. Au niveau de la gestion des informations, Le Manufacturing Execution System
MES obtient des données de SCADA, fournissant le statut de la production au système ERP. Lorsque
tous les systèmes sont correctement intégrés, les informations sur les ordres de fabrication circulent
également en sens inverse (en aval), de l'ERP au MES puis au SCADA, aidant à établir un lien
transparent entre les ressources de l'entreprise et le management des ordres de fabrication. Par
conséquent, l'intégration verticale offre plus de transparence et de contrôle du processus de production
et contribue à améliorer la prise de décision en atelier. Le Smart Manufacturing intègre des machines en
réseau dans l'atelier via une communication de machine à machine (M2M), ce qui contribue à améliorer
30
l'adaptabilité à différents types de produits. Le M2M consiste en un système de communication avec
interopérabilité, qui rend les machines capables de se comprendre, facilitant leur adaptation dans les
lignes de fabrication.
Cette capacité est prise en charge par la mise en service d’un émulateur virtuel, qui émule les différents
codes PLC des machines et valide les procédures de configuration, évitant ainsi les temps d'arrêt
prolongés dus à la longue configuration de l'équipement. Cette simulation est plus performante avec la
fabrication numérique, qui, outre les codes PLC, considère également les données de tous les objets
virtualisés de l'atelier, puis simule les processus des opérations, en tenant compte de plusieurs
paramètres pouvant affecter la production. La fabrication intelligente favorise également une
robotisation améliorée. En effet, les robots peuvent effectuer des tâches avec plus de précision que par
le passé, augmentant la productivité tout en étant beaucoup moins sujets à la fatigue.
De plus, l'intelligence artificielle prend en charge la production 4.0 de plusieurs manières. Dans les
machines, des outils analytiques avancés peuvent analyser les données collectées à partir de capteurs
pour surveiller et prévoir les pannes des machines, les surcharges ou tout autre problème. Cela ouvre la
voie à une maintenance prédictive qui permet d'éviter les temps d'arrêt dus à des pannes inattendues
pendant le processus de production. Les machines dotées d'une intelligence artificielle peuvent
également identifier automatiquement les non-conformités des produits dans les premières étapes du
processus de production, augmentant ainsi le contrôle qualité et réduisant les coûts de production. En
outre, l'intelligence artificielle complète également des systèmes comme l'ERP, prévoyant les demandes
de production à long terme et les transformant en commandes de production quotidiennes, en tenant
compte des commandes de dernière minute et des restrictions d'exploitation.
Concernant la traçabilité interne, des capteurs sont intégrés aux matières premières et aux produits finis
dans l’entrepôt de l’usine. La traçabilité interne peut soutenir des systèmes adaptables avec des lignes
flexibles, dans lesquelles les machines lisent les exigences des produits dans les capteurs embarqués et
effectuent les actions nécessaires à leur fabrication. Les lignes flexibles peuvent également comprendre
des machines modulaires qui sont facilement connectées à une ligne de fabrication avec une
configuration minimale. Cela permet la production de différents types de produits en petits lots, avec
une perte de productivité minimale. Enfin, pour améliorer l’efficacité de l’usine, la fabrication
intelligente comprend également la gestion de l’énergie (surveillance et amélioration de l’efficacité
énergétique). La surveillance de l'efficacité repose sur la collecte de données de la consommation
d'énergie dans les réseaux électriques, tandis que son amélioration est obtenue grâce à des systèmes
intelligents de gestion de l'énergie qui planifient des étapes intensives de production à des moments où
les tarifs d'électricité sont favorables, par exemple.
31
Pour résumer, (Lin et al., 2019) ont évoqué l'importance d'améliorer les systèmes de fabrication pour
faire face aux événements externes imprévisibles et à la complexité interne, en particulier lorsque le
système de contrôle hiérarchique traditionnel est centralisé, rigide et incapable d'agir rapidement pour
répondre aux changements et les dysfonctionnements. Quatre éléments clés sont identifiés par (Lin et
al., 2019) pour l'évolution des systèmes de fabrication comme représenté par la Fig. 8: Agilité pour la
gestion des événements imprévus, Adaptabilité pour la diversité des produits, Efficacité pour
l'optimisation des processus et Réorganisation pour créer de nouvelles fonctionnalités.
Fig. 8 Éléments clés pour l'évolution des systèmes de fabrication (Lin et al., 2019)
Dans l’analyse bibliographique de ce chapitre, nous évoquons quelques architectures innovantes dans la
littérature qui ont été présentées dans le document de thèse principal afin de répondre aux attentes de
l'industrie 4.0. L’objectif de l’analyse des architectures 4.0 existants est de définir un positionnement
original par rapport à l’état de l’art qui nous mène par la suite à identifier notre hypothèse de recherche.
(Lin et al., 2019) ont proposé une architecture de contrôle de fabrication cyber-physique, prenant en
considération les faiblesses de la pyramide CIM traditionnelle en termes de limitations et de complexité
en ce qui concerne la reconfiguration de la ligne de production via le MES, et le manque de mise à jour
en temps réel entre le système ERP et le MES s’agissant de l'état de la production. Par conséquent, une
autre architecture a été proposée basée sur la transformation de machines en entités CPS avec des
capteurs embarqués, connectées entre elles et communiquant directement avec une plateforme cloud /
32
fog computing jouant le rôle de gestion des opérations. Les microcontrôleurs industriels MCU ont été
utilisés pour contrôler les machines et interagir entre elles avec la plateforme cloud. Ce système a la
capacité d'auto-prédire et d'auto-diagnostiquer les données traitées à l'aide de techniques d'apprentissage
via le cloud.
De même pour la plateforme cloud du modèle mentionné, (Bonnard et al., 2019) a proposé une
plateforme de big data / analytics pour la mise en œuvre de l'Industrie 4.0 dans les PME. Cette
architecture consiste à utiliser une interface de programmation (API ou API web) qui respecte les
contraintes d'architecture REST (Representational State Transfer) collectant des données à partir des
objets connectés, d'automates, de SCADA, de MES et d'ERP conçus dans une architecture orientée
services, publiés via une interface Web ou une application mobile et connectée à une base de données
MySQL. La plateforme prend en charge plusieurs protocoles permettant la communication avec
différentes entités de l'atelier telles que OPC-UA, Modbus ou Profibus, la communication sans fil ou
Ethernet. Les données collectées seront stockées sur une machine locale afin d'être ensuite envoyées via
API REST vers le serveur cloud. La plateforme offre la possibilité de traitement de données par l'IA
analytique et les algorithmes génétiques, ce qui permet un soutien aux décisions.
(Block et al., 2018) ont introduit une approche PPC de planification et de contrôle de la production,
basée sur le concept d'utilisation de CPS connectés, collectant des données lorsqu'aucune base de
données centrale n'est requise, partageant des informations avec d'autres CPS et contenant un modèle de
l'objet physique considéré. Les Données de simulation de fabrication de base CMSD sont utilisées par
(Block et al., 2018) pour l'échange de données et lié à AutomationML afin d'assurer la communication
avec les processus de niveau supérieur et le MES. Les CPS décentralisés sont considérés comme un
niveau de calcul de périphérie puisque toutes les données sont traitées et simulées sur la machine CPS.
Enfin, un PPC événementiel est déployé dans une station centrale pour encapsuler des unités dans une
architecture SOA orientée services.
En plus, plusieurs études de recherche considèrent la dimension management des feuilles de route
génériques de transition numérique dans les entreprises. Nous citons (Kamble et al., 2018) qui ont
proposé un Framework de l'industrie 4.0 construit avec trois composants, les technologies de l'industrie
33
4.0, l'intégration des processus et les résultats de cette intégration. En se basant sur une revue
approfondie, ce Framework a été proposé compte tenue des installations promises des technologies de
l'industrie 4.0 et de sa capacité pour intégrer les processus et les activités de l'entreprise, dans l’objectif
de rendre le système de fabrication plus flexible, économique et respectueux de l'environnement.
Un modèle en trois étapes pour la transformation de l'Industrie 4.0 « Envision, Enable, Enact » est
proposé par (Erol and Schumacher, 2016). L'objectif principal du modèle est de guider les entreprises
dans le développement de leurs objectifs spécifiques ainsi que d'un ensemble de mesures pour atteindre
leur vision et leur feuille de route spécifiques. (Ghobakhloo, 2018) propose une feuille de route
stratégique pour la transition Industrie 4.0, prenant en compte l'ensemble de la chaîne
d'approvisionnement et de valeur de l'entreprise, sous forme de recommandations visant à faciliter la
transition vers l'Industrie 4.0
Cependant, notre démarche qui sera contextualisée à travers l’hypothèse proposée prend en
considération l’état actuel du système de production proche d’un fonctionnement 3.0 et essaye d’aboutir
une démarche innovante qui figure dans la proposition d’un modèle suivi d’un Framework de la
transition numérique des systèmes de fabrication tout en s’appuyant sur l’architecture existante du
système de production.
La recherche menée ci-dessus porte principalement sur le développement de modèles de transition des
systèmes de production conventionnelle vers des systèmes de production numérisée où l’analyse des
données collectées et la reconfigurabilité des lignes de production peuvent accroître sensiblement les
performances opérationnelles. L’évolution des systèmes de production repose sur l'intégration verticale
des technologie 4.0 et produisent de nouvelles fonctionnalités telles que la flexibilité, la modularité, la
reconfiguration, l’interopérabilité et l’intelligence. L'architecture des systèmes de fabrication avancés
est complexe, hétérogène, dynamique et indéfinie. La modélisation et la conception de ces systèmes
doivent se caractériser par des attributs tels que l'interopérabilité, la flexibilité, la décentralisation et
l'abstraction. Un système de fabrication avancé peut être basé sur une combinaison de plusieurs
architectures. Certains sont considérés comme des approches de modélisation telles que SOA et «
Manufacturing as a Service (MaaS) » et d'autres sont des architectures telles que « Industrial Internet of
Things (IIoT) » et CPS. Les approches de modélisation et l'architecture peuvent être combinées puisque
le paradigme de modélisation est choisi pour le système de contrôle en fonction du type et des objectifs
de l'application de production, puis le système de fabrication doit être conçu à travers une architecture
dédiée (Borangiu et al., 2019).
34
La problématique dynamique de la structure verticale traditionnelle et de la pyramide d'automatisation
est notre principale préoccupation. Sur la base de ce qui a été discuté, l'évolution des systèmes de
production vers une approche 4.0 est fortement dépendante de l'intégration verticale. Deux facteurs clés
doivent être pris en compte pour atteindre l'objectif d’une transition numérique. Le premier est un
facteur de flexibilité identifié par (Lin et al., 2019). Un système de production doit être capable de
s'adapter aux nouvelles commandes (sous la capacité de l'usine) et de répondre rapidement aux
fluctuations de la demande et du workflow tout en assurant la personnalisation de masse. Le deuxième
facteur identifié par (Sung, 2018) concerne la gestion des données. Un système de production doit
être capable d'exploiter les données collectées à partir des machines, des capteurs, des pannes et des
données historiques, afin d'apprendre et d'auto-optimiser ses processus de production.
Grâce à la dimension « données », dans le contexte de l'Industrie 4.0, les systèmes de fabrication gagnent
en niveau d’intelligence. La fabrication intelligente vise à convertir les données acquises tout au long
du cycle de vie du produit en intelligence de fabrication afin d’optimiser la fabrication de différents
points de vue (Tao et al., 2018). Nous citons, l'amélioration des performances de production,
l’amélioration de la sécurité des travailleurs et des machines, l'amélioration la qualité des produits et la
facilitation de développement de produits intelligents. De ce fait, les industriels doivent d'abord traiter
une grande quantité de données générées tout au long du cycle de vie du produit, y compris les images
d'inspection visuelle et le traitement du contrôle qualité (Bajic et al., 2018). Deuxièmement, les systèmes
de fabrication évolués doivent être capables d'intégrer ces modules flexibles et de les activer par
l'intelligence artificielle (IA) de manière aussi peu intrusive que possible.
35
5. Hypothèse
Après avoir décliné les principes de l'Industrie 4.0 appuyée par un état de l’art détaillée, nous abordons
dans ce chapitre une hypothèse qui consiste à dire que la transition vers la digitalisation des systèmes de
production obéit à de grandes recommandations qui doivent être contextualisés en fonction de l’état des
lieux et des objectifs visés par les entreprises. La section 5.1 introduit notre hypothèse. La section 5.2
présente la démarche suivie pour valider l’hypothèse proposée.
L’analyse menée dans le chapitre précédent nous conduit à reprendre et à justifier le positionnement de
notre problématique de recherche présentée dans le chapitre 3, Quel modèle pour une transition
réussie vers la numérisation des systèmes de fabrication 4.0 ? qui se situe au croisement des trois
axes suivants ; la ‘numérisation’, le ‘système de fabrication’, et les ‘processus de production’.
L’agilité offerte par l’industrie 4.0 permet aux usines de gérer une gamme de produits plus diversifiée
avec un délai de livraison plus court et de personnaliser rapidement les produits en fonction des
exigences spécifiques, garantissant ainsi une plus grande réactivité aux besoins des clients. Alors que
l'Industrie 4.0 offre de nouvelles opportunités pour les entreprises, elle soulève dans le même temps de
nouveaux défis. Un état de l’art détaillé a été fourni pour définir les principes et les concepts de la
numérisation des systèmes de fabrication, et identifier les défis principaux respectives à cette intégration
numérique.
Nous constatons dans la littérature que le sujet abordé par cette thèse de doctorat représente une
originalité en termes d’actualité et d’importance, ce qui explique la rareté des approches qui s’intéressent
à l’évolution des systèmes de production existants vers la transition numérique en considérant le système
de production dans son intégralité. Les architectures présentées dans le contexte de l’intégration de
l’industrie 4.0 dans les systèmes de production sont innovantes mais en revanche se basent en majorité
sur le développement de nouvelles initiatives de transfert de data qui ne sont pas reconnu jusqu’au
aujourd’hui dans les entreprises et qui exigent des changements radicaux dans les architectures
existantes ce qui peut causer des ruptures potentielles dans les sites de production. En outre, un deuxième
aspect critique à mentionner au cours de cette analyse concerne les limitations d’une telle proposition
d’architecture, comme celle développée par (Block et al., 2018), à des applications dans le « process
manufacturing » qui marquent une grande différence avec le domaine de fabrication industrielle.
Le sujet de la numérisation des systèmes de production exige une approche spécifique qui considère
l’intégralité du système de production avec ses différentes couches ainsi que son architecture actuelle.
Cette approche vise à évoluer le système de production tout en assurant les fonctionnalités attendues par
les entreprises.
36
La question dynamique de la structure verticale traditionnelle de production doit d'abord être capables
de s'adapter aux fluctuations de la demande. Deuxièmement, les systèmes de production doivent
permettre de collecter les données à partir de machines et de capteurs afin d'apprendre et d'auto-optimiser
ses processus. Ainsi, les principaux défis et orientations définis dans l’état de l’art en fonction de notre
direction de recherche et de notre intérêt sont la mise en œuvre d'un système de fabrication flexible et
modulaire dans le contexte de l'intégration verticale. Par conséquent nous proposons l'hypothèse
suivante qui vient répondre à notre problématique de recherche identifié comme suivant :
L'évolution des systèmes de production vers la numérisation dans le contexte de l'Industrie 4.0
nécessite d'agir sur trois axes principaux : le système de production, y compris le système de
contrôle de l'atelier SFCS, les machines et technologies avancées, et enfin le traitement et le
déploiement des données.
A cet effet, la Figure 9 présente les différents axes d'intérêt identifiés par l'hypothèse. Le premier axe
« Production System & SFCS » concerne principalement le système informatique de la production, sa
capacité et ses exigences à s'adapter à la flexibilité et la modularité attendues. Le système d'information
de production comprend l'agilité du réseau informatique ainsi que les plateformes de production
numériques telles que MES, ERP et autres logiciels. Cet axe est également concerné par le système de
contrôle de l'atelier et sa reconfiguration qui doit répondre aux fluctuations de la demande et permettre
un contrôle flexible de la production.
Le deuxième axe « Traitement et déploiement des données » consiste à activer les flux de données dans
les sens descendant et ascendant. Après la mise en œuvre des capteurs et de l'IoT, les données doivent
être collectées à partir des machines et d’autres ressources pour y être analysées en vue d’une prise de
décision et de l’optimisation du la production.
Notre démarche présentée par la Figure 10 consiste à proposer un modèle conceptuel se basant sur les
principes qui accompagnent l’intégration verticale définis dans la littérature qui répond aux différentes
fonctionnalités des systèmes de production numériques. Ensuite, ce modèle conceptuel sera étendu pour
être adapté à la situation d'un système de production traditionnel.
Par la suite, et dans le but d’appliquer le modèle conceptuel développé, nous proposons un Framework
innovant de la transition numérique des systèmes de fabrication qui prend en considération les
différentes dimensions conceptuelles de notre modèle et les transforme en une structure de support
détaillée pour déployer l’intégration de l’industrie 4.0 dans les systèmes de production existants.
38
Enfin, le Framework de la transition numérique sera appliqué à la plateforme industrielle 4.0 introduite
dans la section 2.2.2. Les travaux expérimentaux d’intégration seront bien détaillés dans le chapitre 7.
A partir du déploiement de la nouvelle architecture système nous proposons par la suite une
méthodologie de conception pour la configuration des processus modulaires suivie de l’application de
cette approche modulaire dans le système de production. Une application quantitative sera menée
finalement sur la flexibilité du système déployé afin d’expérimenter notre démarche scientifique.
En synthèse, l'objectif du Framework proposé est donc de développer un nouveau processus modulaire
pour attribuer au système de production de nouvelles fonctionnalités :
- La flexibilité de fabrication
- La modularité du système informatique
- La décentralisation du niveau de décision
- La personnalisation
- La portabilité
39
6. Modélisation des systèmes de production 4.0
A partir des concepts définis et cités précédemment, (Benfriha, 2020) a proposé une initiative de
modélisation du concept de systèmes de production 4.0. Tout d'abord, le modèle est présenté dans son
originalité adaptée à une modélisation générale d’un système de production 4.0, en associant les trois
axes de l’hypothèse proposée aux différentes concepts évoqués par le modèle. Deuxièmement, le modèle
sera étendu pour être adapté à la situation d'un système de production traditionnel qui doit être mis à
niveau vers un système de production 4.0. La dernière section introduit le Framework de transition
numérique qui engendre une structure de support visant à guider l'intégration du 4.0 dans le système de
production existant.
Le but ici est d'exploiter les concepts développés dans la littérature et de construire un nouveau modèle
de système de production 4.0. Nous avons identifié trois concepts représentatifs des systèmes de
production qui pourraient être rencontrés dans l'industrie, abordés par (Bonnard et al., 2019), par (Lina
et al., 2019) et (Gorecky et al., 2016). En général, tous les concepts sont concourants, cependant leur
spécificité trouve son origine dans le contexte d'utilisation. Avant de se lancer dans un processus de mise
en œuvre des concepts de l'Industrie 4.0 dans une entreprise, plusieurs experts recommandaient une
organisation optimisée de la production comme préalable, notamment par l'utilisation de méthodes
comme le lean manufacturing. Il convient de mentionner à ce stade que toute implémentation d'un
système de production 4.0 doit considérer le modèle d'architecture de référence pour l'industrie 4.0
RAMI4.0 comme un cadre de la normalisation de référence pour la numérisation industrielle.
A partir d'une analyse des concepts définis dans la littérature dans le périmètre des trois axes bien définis
par l’hypothèse proposé, nous présentons ici une initiative de convergence qui consiste à se baser sur un
nouveau concept général simplifié qui réorganise les différentes technologies et architectures impliquées
dans une vision complète et cohérente d'un futur modèle de système de production 4.0. Ce modèle
conceptuel général considère différentes briques technologiques et concepts de base essentiels,
représentées par les 5 axes identifiés par (Benfriha, 2020) Figure 11, qui constituent les principes de
base de la modélisation d’un système de production numérique 4.0. Les trois axes identifiés par
l’hypothèse « Production System & SFCS », « Traitement et déploiement des données », « Machines et
Technologies Avancées » dans l’objectif d’effectuer la transition numérique des systèmes de production
se reposent comme le montre la Fig. 11 sur les 5 axes conceptuels essentiels du processus de
modélisation proposé par (Benfriha, 2020) : les applications, la data, les systèmes CPMS, les systèmes
IIoT, et les systèmes IT.
Le 1er axe considère que le système de fabrication doit être configurable dans les systèmes de
fabrication physique cybernétique (CPMS). Cet axe suppose que les machines disponibles intègrent
40
nativement la capacité de travailler dans un environnement connecté ou de les faire évoluer pour être
compatibles avec un système de production 4.0. L'objectif de CPS est de construire un système
d'ingénierie contrôlable, fiable, évolutif et capable d'interagir en temps réel. Les CPS sont basés sur les
derniers développements dans les domaines de l'informatique, des technologies de l'information et de la
communication, ainsi que des sciences et technologies de fabrication dans la perspective d'augmenter
les performances de production grâce à des systèmes de contrôle intelligents basés sur les données
collectées et transformées (Joost et al., 2012). Pour répondre aux nouvelles demandes du marché, de
nouveaux paradigmes technologiques doivent être appliqués aux systèmes de fabrication dans le but
d'améliorer l'agilité, la réactivité et d'assurer la qualité des produits. CPS peut relever les défis des futurs
systèmes de fabrication en créant le pont entre les éléments physiques et les systèmes de télécommande.
La combinaison de plusieurs CPS interconnectés et améliorés par les technologies cloud a conduit à
augmenter les processus de nouvelles fonctionnalités 4.0. Une ligne de production intelligente peut
inclure de nombreux systèmes cyber-physiques reconfigurables capables d'interagir et d'échanger des
données en temps réel avec d'autres systèmes distants. Ces aspects qui contribuent à la flexibilité
opérationnelle posent, à la fois, l'hétérogénéité et la complexité des applications CPS qui représentent
un réel défi dans leur développement où les aspects de fiabilité, de sécurité et de certification sont
prioritaires.
41
Le 2ème axe concerne les dispositifs IIoT qui superposent les systèmes CPMS par une couche de
capteurs et de caméras. La superposition d’une couche IIoT ouvre la voie à des applications de
modélisation virtuelle des processus de fabrication (DTS et DDT) et permet le traitement des données
via le Cloud computing ou Edge computing. L'Internet des objets IoT est l'une des technologies de
rupture avancées et considéré par le McKinsey Global Institute comme une avancée technologique qui
« transformera la vie, les affaires et l'économie mondiale ». Le système IIoT va bien au-delà de la
collecte de données à partir de capteurs, il est considéré comme un système holistique qui gère les
données de la collecte à l'analyse d'une grande quantité de données pouvant être utilisées pour améliorer
les performances globales des systèmes industriels. (Khan et al., 2020) donnent la définition la plus
adéquate du système IIoT comme « le réseau de composants industriels intelligents et hautement
connectés qui sont déployés pour atteindre un taux de production élevé avec des coûts d'exploitation
réduits grâce à une surveillance en temps réel, une gestion et un contrôle efficaces des procédés
industriels ».
La pertinence des systèmes IIoT est également prise en compte pour la planification et l'ordonnancement
dynamiques de la production dans l'atelier. Le système IoT peut intégrer le système de l'atelier, les
données des capteurs de localisation sont utilisées pour prendre les décisions logistiques internes les
plus optimales ; des accéléromètres peuvent être déployés à des fins de maintenance prédictive
(surveillance des vibrations), etc.
Le 3ème axe aborde les applications logicielles nécessaires à l'exécution et au contrôle d'un système de
production avec un soin particulier sur la problématique d'interfaçage logiciel. En effet, diverses
applications interviennent dans différentes phases de fabrication de manière ponctuelle ou continue. La
problématique de l'architecture logicielle et de l'interfaçage est un point crucial dans la réussite d'un
processus de fabrication intégré et autonome.
L'approche traditionnelle de la planification des processus dans diverses petites et moyennes entreprises
est basée sur les connaissances et l'expérience du personnel impliqué dans les activités de fabrication.
La phase de conception et de développement d'un produit intelligent doit prendre en compte l'inclusion
des données des trois étapes suivantes : planification des processus, planification des opérations et
séquençage des opérations. L'objectif principal doit être de collecter autant de données que possible afin
que le réseau digitalisé et virtuel des machines de fabrication et du produit puisse être établi directement.
La configuration des outils de planification des processus et de séquençage des opérations doit fournir
une efficacité accrue pour l'optimisation des processus et une capacité assurée pour la personnalisation
masse des produits.
Selon (Kumar et al., 2019), il existe trois critères importants associés au logiciel de planification de
produit : la planification des processus, le séquencement des opérations et enfin la sélection de la
42
machine. Le logiciel de planification des produits recueille des informations provenant de différentes
parties de la chaîne d'approvisionnement liées aux activités de fabrication importantes. La masse du
produit, son matériau, sa forme, son état de surface, ses caractéristiques géométriques et sa qualité
constituent les sous-critères des critères de planification du processus. Le deuxième critère du logiciel
de planification de produit est le séquencement des opérations. Le logiciel de planification du produit
doit prendre en compte les aspects qui influencent les processus de fabrication, tels que l'opération de
finition, le nombre d'opérations requises, la méthode de fixation et la priorité de l'opération. Le troisième
critère majeur pour les logiciels de planification de produits est la sélection des machines et les critères
de planification. Les sous-critères importants associés incluent le temps de fonctionnement, les outils
nécessaires, la qualité requise, le coût de la machine, la disponibilité de la machine et le nombre de
produits commandés. La sélection de la machine constitue une partie importante du logiciel de
planification des produits, qui se fait grâce aux informations sur les machines disponibles dans le
système de fabrication. La prise en compte de ces critères permet au système de fabrication de réaliser
l'opération avec l'objectif global de minimisation des coûts et du temps de fabrication.
Le 4ème axe pose la question des données générées ou importées à plusieurs points de vue, stockage,
sécurité, traitement, filtrage, partage et visualisation. Il est certain que l'Industrie 4.0 générera plus de
données, et plus qu'assez pour savoir collecter, analyser et utiliser afin d'en tirer une valeur ajoutée.
L'analyse des données permet de générer des informations susceptibles d'augmenter les performances
du système de production. Cet axe est sans aucun doute au cœur de cette révolution industrielle et
représente un enjeu majeur en termes de gestion, d'optimisation et de cybersécurité. Les entreprises
doivent être en mesure d'identifier et de localiser les données dans leurs formats d'origine dans un lac
de données. Les défis du contrôle des données sont importants. En effet, selon le Data Literacy Index
(Jung, 2002), les entreprises qui réussissent à améliorer leur capacité à utiliser les données pourraient
augmenter leur valeur de 3% à 5%. En outre, il semble avoir une forte correspondance entre
l'amélioration de la capacité du personnel à utiliser les données pour la prise de décision d'une part et
les mesures de la performance commerciale, telles que la marge de production brute, le rendement des
actifs, le rendement des capitaux propres et la rentabilité des ventes. Un lac de données désigne un
espace de stockage global qui absorbe le flux de données présent dans un système de production 4.0.
Cela implique de le faire avec suffisamment de flexibilité pour interagir avec les données dans leur
format d'origine, qu'il s'agisse de données brutes, structurées et non structurées. En plus du stockage,
l'un des enjeux du data lake est de pouvoir traiter et transformer facilement les données afin d'en extraire
des informations utiles pour améliorer les performances des processus de production mais aussi pour
enrichir la phase de conception du produit. Le choix de traiter les données localement EdC ou dans un
cloud CdC dépendra de la finalité d'utilisation.
43
Le 5ème axe est sans aucun doute le plus complexe concernant l'intégration verticale et la
communication entre les différentes couches du système de production. Cet axe comprend des choix en
termes de services distribués, orientés objet ou de systèmes d'information conduite par évènement (EDA,
SOA) adaptés aux besoins des entités internes et externes qui interviennent, selon différents protocoles,
en temps réel ou différé. Une architecture de système d'information adaptée est nécessaire pour déployer
la flexibilité et l'adaptabilité du processus de production et améliorer une gestion décentralisée de la
prise de décision à travers le réseau distribué.
(Theorin et al., 2017) ont abordé la pertinence de l'intégration de la SOA au niveau de gestion des
opérations de fabrication et au niveau de contrôle. La plupart des outils SOA sont conçus pour les
processus d'entreprise, qui n'ont pas d'exigence stricte en matière de temps ou de ressources, tandis que
les périphériques qui s'exécutent au niveau de la surveillance et du contrôle ont souvent une mémoire et
une puissance de calcul strictement limitées. SOA est une architecture logicielle distribuée où les
applications autonomes s'exposent en tant que services, auxquels d'autres applications peuvent se
connecter et utiliser. Pour atteindre leur plein potentiel, les applications SOA doivent être
autodescriptives, découvrables et indépendantes de la plateforme informatique et du langage. Cela
conduit à un couplage lâche et une flexibilité élevée. L'adoption de la SOA dans une entreprise
commence généralement par une initiative informatique visant à améliorer l'efficacité de l'infrastructure
et peut ensuite évoluer vers une utilisation optimisée à des fins commerciales.
Dans le cas d'un système de production traditionnel existant, le modèle proposé doit prendre en compte
les machines et réseaux déjà installés. L'approche que nous proposons consiste à analyser le système
existant afin de le valoriser avec une couche technologique adaptée aux fonctionnalités souhaitées et au
potentiel de l'entreprise à investir dans les lignes de production en réponse aux objectifs business
souhaités.
Avant d'entamer le processus de transition, nous considérons que l'entreprise souhaitant évoluer vers un
système de production 4.0 bénéficie déjà d'une architecture CIM existante Figure 12 qui fonctionne sur
un modèle proche de 3.0 et son processus de fabrication fait déjà l'objet d'une amélioration continue,
notamment en utilisant le lean manufacturing. Ce point de départ est crucial pour une transition réussie
vers des systèmes évolués.
En se basant sur le modèle conceptuel introduit ci-avant et les axes identifiés par notre hypothèse, nous
considérons que la transformation de l'atelier comporte 5 aspects. L'aspect matériel considérant
l'acquisition d'équipements pour les transferts entre machines, le contrôle qualité intégré ou les machines
de fabrication additive. A cela s'ajoute la reconfiguration des machines existantes en termes de contrôles
afin de les rendre compatibles avec un système de production connecté. Le deuxième aspect concerne
la superposition d'une couche IoT comme des capteurs, des caméras avec des réseaux dédiés
interopérables avec le réseau existant.
Le troisième aspect concerne l'acquisition de logiciels, notamment pour gérer l'atelier. Par ailleurs, le
développement d'un environnement de gestion numérique et d'industrialisation spécifique est nécessaire,
auquel seront transférées les fonctions de contrôle, sans oublier les problématiques d'interfaçage logiciel
appelées à dialoguer lors de l'exécution d'un plan de production. Le 4ème aspect concerne les capacités
de calcul, déportées ou locales, nécessaires pour supporter un système de production 4.0 doté de
fonctionnalités avancées, notamment lors de l'utilisation de caméras avec des attentes de réactivité en
temps réel.
Le dernier aspect concerne la programmation des machines. En effet, en 3.0, les programmes machine
sont conçus pour avoir une configuration dédiée à la production de masse. Les machines deviennent une
entité unique qui assure le bon déroulement d'un plan de production bien organisé. En revanche, en 4.0
chaque machine est une entité distincte et peut être considérée comme un CPS, ou incluse dans un CPS,
connectée, intelligente et autonome. Cette modularité au niveau de l'atelier nous conduit à concevoir des
45
programmes de fabrication spécifiques pour un plan de production modulaire, autonome, adaptable et
reconfigurable.
En conclusion, la modélisation des systèmes de production 4.0 explore différents concepts combinés et
déployés pour construire un système intégré et intelligent. Les concepts interfèrent avec différents
domaines et expertises et leur interopérabilité constitue un défi technique majeur. Une grande majorité
des articles consultés évoquent une complexité multi-échelles, incluant mais sans s'y limiter
l'interfaçage, la sécurité et les compétences. Les stratégies industrielles d'évolution sont également
confrontées à des problématiques de conseil délivrées par les éditeurs de solutions de plateformes
numériques et les fabricants de systèmes technologiques 4.0. Cette concurrence, sous une enveloppe
marketing bien développée, entraîne souvent des difficultés dans notre choix de solutions et dans leur
mise en œuvre. Un Framework de transition vers les systèmes de production de l'industrie 4.0 est proposé
dans le chapitre suivant afin d'appliquer concrètement le modèle proposé et d'orienter les entreprises qui
souhaitent faire évoluer leurs systèmes de production.
46
6.3 Framework de la transition numérique des systèmes de fabrication 4.0
Un nouveau Framework a été développé afin d’'exploiter ce modèle et de générer une structure de
support visant à guider l'intégration du 4.0 dans le système de production existant. Comme identifié
dans la dernière section, la gestion de projet de cette intégration est une clé essentielle à considérer.
Un Framework de principes de conception assurant principalement l’intégration verticale et
permettant l’horizontale dans la structure de l’entreprise est introduit et expliqué.
L'industrie 4.0 est un système dynamique et intégré permettant de contrôler l'ensemble de la chaîne de
valeur du cycle de vie des produits. L'intégration verticale et horizontale et la fusion des composants
physiques et virtuels sont essentielles pour la mise en œuvre de l'industrie 4.0 et des tendances
technologiques telles que les systèmes cyber-physiques et l'Internet des objets. En d'autres termes, le
concept de base de l'Industrie 4.0 est de construire un système de fabrication intelligent et auto-optimisé
basé sur les données traitées et de s'appuyer sur l'exploration de données collectées et supervisées en
temps réel. Ce processus d'exploration de données doit être bien implémenté dans le système de
fabrication et capable d'optimiser la configuration du processus de production en cours (système de
fabrication flexible). Cette mise en œuvre impose une liste d'exigences et de changements radicaux dans
tout le système de production, ciblant chaque couche du système de fabrication.
Alors que les processus de fabrication intelligents et connectés introduits par l'Industrie 4.0 attirent de
plus en plus les entreprises qui s'efforcent de s'adapter aux demandes des clients. A cet effet, nous
proposons une nouvelle approche conçue et appliquée avec une nouvelle architecture qui favorise la
transition de la fabrication numérique.
47
Fig. 13 Framework des principes de la transition
Notre Framework de transition est composée de cinq couches, assurant principalement l’intégration
verticale, et permettant l’horizontale dans la structure de l’entreprise. L'intégration verticale implique la
flexibilité au niveau de l'atelier et celle au niveau du contrôle commande. La flexibilité opérationnelle
assurée par les deux premières dimensions devrait être renforcée par le fait de prendre en compte la
dimension de gestion des données puisque l'objectif global de la transition numérique est d'améliorer la
performance opérationnelle par le data mining. Le déploiement complet de cette intégration est assuré
par la dimension d'intégration modulaire des services au niveau du système d'information de production.
Enfin, l'ouverture pour une communication au niveau business de l'entreprise et des lignes de production
distribuées consiste à assurer la communication du système de production avec les couches business de
l’entreprise. L’idée est de garantir au système de production une intégration horizontale tout au long de
la chaîne de valeur industrielle et surtout assurer sa communication avec les systèmes ERP des
entreprises.
En abordant techniquement la 1ère couche du Framework ’atelier flexible’, plusieurs aspects concernant
les machines doivent être pris en compte. Tout d'abord, un système de fabrication flexible FMS dans le
contexte de l'industrie 4.0 est requis. Un FMS est basé sur des machines flexibles telles que des robots
- Robotic Process Automation (RPA). La flexibilité des machines nécessite au moins de résoudre des
problèmes tels que l'interopérabilité, l'accès à distance aux machines et permettre une configuration
flexible du processus d'usinage. En plus des exigences susmentionnées, l'intégration des machines avec
un niveau décisionnel aiderait efficacement à transformer la machine en module CPS, idéalement
lorsqu'elle est capable d'exécuter des algorithmes et de communiquer des données à d'autres modules du
réseau. Deuxièmement, la modularité du système de fabrication est un autre aspect à considérer. La
48
modularité dans le FMS est un compromis optimal entre les autres systèmes de production traditionnels
(Kaushal et al., 2016). La flexibilité globale du système de production découle logiquement de la
flexibilité des modules et des flux qui le composent. La modularité dans ce contexte repose fortement
sur le système de contrôle d'atelier flexible et dépend essentiellement de sa configuration.
Afin de mettre en œuvre l’Industrie 4.0 et d'exploiter ses fonctionnalités innovantes, il est donc
nécessaire d'assurer la flexibilité des systèmes de contrôle de fabrication et des processus de
production figurant dans la 2ème couche du Framework. Comme discuté ci-avant, la flexibilité des
machines au niveau de l'atelier une fois configurées, doit être prise en charge par le système de contrôle
afin d'être déployée dans le système de production. Les machines flexibles et les CPS modulaires
signifient des opérations et des processus flexibles, où leur exécution et leur séquencement nécessitent
un système de contrôle d'atelier flexible. Par conséquent, une configuration différente de nos systèmes
de production est nécessaire pour permettre d'abord la planification dynamique du flux de travail et des
processus, et ensuite l'optimisation autonome en temps réel des processus de production, ce qui n'est pas
pris en charge par les systèmes de contrôle conventionnels en raison de sa configuration rigide.
La gestion intégrée des données doit d'abord prendre en compte le traitement local des données pour
les informations de bas niveau, notamment les machines, les appareils et les CPS, en particulier les
mesures de sécurité, la disponibilité des machines et la gestion globale de l'atelier. Les données locales
et de bas niveau doivent être prises en charge par des modules de traitement et des unités de calcul de
périphérie si nécessaire. Deuxièmement, l'agrégation d'informations de niveau supérieur telles que les
informations de planification de la production et les données historiques, les tendances et les KPI est
essentielle pour améliorer les processus. De plus, la large utilisation des capteurs et la numérisation par
acquisition de données fournissent des informations aux analystes et au processus de prise de décision.
Par conséquent, la gestion intégrée des données devrait permettre le stockage de ces données dans le
cloud, ce qui permettra leur accessibilité et leur traitement à tout moment à faible coût. La dimension de
gestion des données est essentielle à la numérisation des systèmes de production. Les données doivent
être classées et bien agrégées, transférées et stockées, afin d'améliorer les performances des systèmes de
fabrication modulaires et flexibles. Ainsi, l'évolution de l'architecture informatique doit prendre en
compte les flux de données et le stockage (lac de données) assurant une structure appropriée pour
différents types de données.
Enfin, l'ouverture pour une communication au niveau business de l'entreprise et des lignes de
production distribuées considère la dimension de l'intégration horizontale qui couvre le cycle de vie
du produit depuis le fournisseur jusqu'à le client. Cette intégration doit être autorisée à travers le réseau
et prise en compte. Cela implique la conception et le développement de produits, la planification de la
production, le démarrage et la gestion de la production, la logistique, les ventes et les distributions.
50
7. Expérimentations, application à la plateforme industrielle 4.0
Notre approche a été expérimentée sur une plateforme de production industrielle 4.0 installée dans le
laboratoire de recherche Fig. 14. Cette plateforme porte un projet de recherche global piloté par M.
Benfriha (Benfriha, 2020) et qui s’articule en 2 axes. Mes travaux de recherche s’inscrivent dans le 1er
axe qui a pour finalité la modélisation de système de production 4.0.
Fondamentalement, cette plateforme est construite à partir d'un robot fixe - chargé de gérer les pièces
brutes entre le magasin initial et le magasin intermédiaire, ainsi que de transférer les pièces finies vers
la table d'accumulation par le biais du convoyeur. Le robot mobile - responsable du transfert des pièces
entre le magasin intermédiaire et les machines. On retrouve aussi, un système de caméra fixe pour
l'identification et la classification des bruts, un tour, une fraiseuse et une machine de découpe laser.
Au tout début, la plateforme peut fabriquer, assembler et transférer des composants avec un processus
de contrôle automatique fermé et peu propice à la flexibilité, aboutissant à la production d'un seul produit
de type amortisseurs en polyoxyméthylène (POM) composés de 3 éléments (piston, ressort et corps). Ce
mode de fonctionnement correspond au mode de production de masse classique proche du 3.0. Nous
introduisons le nombre d’amortisseurs souhaité, ensuite l’automate exécute en boucle le plan de
51
production sous forme de grafcet déployé avec différentes opérations d’usinage, de transfère et
d’assemblage. Ce mode de fonctionnement classique est loin d’être dynamique, du fait que l’automate
ne réagit pas à aucune anomalie saufs les incidents de sécurité qui figurent dans les conditions qui
peuvent causer l’interruption du cycle. Dans cette configuration conventionnelle, cette plateforme de
production était assez représentative du mode de fonctionnement des entreprises.
L’objectif est d’utiliser cette plateforme afin d’expérimenter et déployer un processus de fabrication
intelligent et flexible dans le but de valider notre hypothèse ‘L'évolution des systèmes de production
vers la numérisation dans le contexte de l'Industrie 4.0 nécessite d'agir sur trois axes principaux : le
système de production, y compris le système de contrôle de l'atelier SFCS, les machines et technologies
avancées, et enfin le traitement et le déploiement des données’, Tout en intégrant plusieurs technologies
et en mettant en œuvre un nouveau système de production flexible.
En s’appuyant sur le Framework de conception que nous avons proposé comme le montre le protocole
expérimental défini par la Figure 15, une nouvelle architecture du système de production a été déployée
pour accroître la flexibilité et la modularité au niveau de l'atelier en tenant compte du système de contrôle
et du système informatique. Plusieurs technologies ont été ajoutées à la plateforme industrielle 4.0
comme le FARO 3D Scanner et des caméras embarquées sur les robots. Une fois les différents niveaux
d’intégration déployés, la justification et la démonstration de notre modèle conceptuel et du Framework
de transition seront établis à travers l’implémentation d’un nouveau programme de production
modulaire. Les méthodologies innovantes intégrées permettront l’émergence de nouvelles
fonctionnalités, telles que la flexibilité opérationnelle, la personnalisation, la production dynamique, le
rendement en lead-time, la décentralisation, et la portabilité. Finalement, une application démonstrative
52
sera menée afin de valider la dimension de la flexibilité et du rendement en lead-time du système de
fabrication évolué.
Le Framework de conception proposée ci-avant est appliqué sur la plateforme expérimentale et enrichi
d’une nouvelle architecture. Dans ce qui suit, le déploiement du Framework est expliqué étape par étape.
Concernant la couche machines, deux niveaux de modularité ont été développés, la modularité des
machines et la modularité opérationnelle. La modularité des machines consiste à transformer des
machines existantes en entités CPS et / ou en machines flexibles. En premier, les robots ont été
transformé en CPS intelligents et connectés en mettant en œuvre des configurations innovantes et en
développant une base de données interne intégrée dans le système de production. Cette base de données
est reliée à des programmes de trajectoires qui permettent les fonctions de transfert entre différents
magasins de stockage des pièces et les machines. Des points de bouclage par zone sont créés pour
éliminer toute collision entre les robots et les machines / obstacles. A ces trajectoires nous avons associé
plusieurs paramètres tels que l’identifiant pièce, le diamètre de la pièce, la hauteur de prise de la pièce
53
(distance entre pince du robot et la base de de la pièce) et enfin la vitesse de déplacement et le numéro
d'outil (pince).
Les machines CNC n’ont pas été conçues pour travailler dans un environnement connecté. Pour résoudre
ces problèmes, une couche IoT a d'abord été ajoutée en superposant les machines de fabrication par des
capteurs et des IoT afin d'exploiter les données et les informations utiles pour améliorer et superviser
les états de fonctionnement. Par conséquent, des capteurs de vibration sont ajoutés aux machines avec
des capteurs de température et de position IoT en fonction des exigences fonctionnelles de chaque
machine afin de collecter le maximum d'informations d'usinage utiles.
Deuxièmement, le contrôle des machines a été étendu par des modules Entrée / Sortie supplémentaires
pour permettre la sélection du programme de fabrication souhaité avec une capacité de paramétrage,
activée via un G-code flexible principal. L'accès à distance aux données de la machine via un adaptateur
réseau USB avec la CNC a permis le transfert de fichiers des programmes d'usinage vers les machines
CNC à partir du MES. L'opération d'usinage CNC est d'abord paramétrée par l'ID correspondant à la
référence de sous-programme associé au produit, la profondeur de coupe et la vitesse de coupe.
Troisièmement, un scanner 3D est installé pour une fonction de contrôle de qualité avancée et
automatisée. Le MES active le scanner 3D via OPC UA. Le scanner 3D génère un nuage de point et le
compare au modèle 3D natif du produit sélectionné et renvoie les résultats de conformité au MES sous
forme de rapport et de décision d’acceptation ou de rejet du produit. En plus de cela, un autre système
de caméra est intégré aux robots et dispose de deux fonctionnalités principales, un contrôle de
conformité supplémentaire in-situ qui signifie que les pièces seront contrôlées sur les machines avant
d'être récupérées par les robots et réajustées si nécessaire, et un contrôle de sécurité supplémentaire pour
éliminer toute collision entre robots vs machines et robots vs pièces.
La modularité des machines présentée ci-dessous permet de générer tous les scénarii d'usinage possibles
et par conséquent la personnalisation des processus de production devient une réalité. La production est
donc organisée à travers des modules fonctionnels qui peuvent être lancés dans un ordre différent. Cela
nous permettra de profiter du plein potentiel de la plateforme et de gagner en flexibilité. Chacun de ces
modules est composé d'opérations standardisées et sera ensuite intégré dans le processus de contrôle
d'automatisation. Pour cela, nous avons créé plusieurs opérations paramétriques qui représentent
l'ensemble des opérations de production adaptées au potentiel de notre plateforme industrielle 4.0:
54
- Opérations de contrôle qualité, y compris les opérations de scanner 3D pour le contrôle de la qualité,
et la qualité dimensionnelle des pièces fabriquées à l'aide d'une caméra intégrée au robot
- Opérations d'inspection / monitoring à l'aide des caméras fixes et embarquées incluant les contrôles de
sécurité, le stockage et les contrôles d'anomalies.
Pour conclure, disposer d'un système de production flexible et adaptatif est essentiel pour assurer la
transition numérique du système de production. Pour atteindre cet objectif, plusieurs changements
affecteraient l'organisation conventionnelle de l'atelier, notamment au niveau opérationnel et matériel.
La transformation de machines conventionnelles en CPS représente la clé de voute d’une transition
réussie vers la numérisation des systèmes de production. Plusieurs technologies avancées introduites par
l’I4.0 Fig. 16 ont été déployées telles que des machines connectées, un scanner 3D, des systèmes de
caméras embarquées intelligentes et des capteurs IoT, ainsi que le déploiement d'opérations
paramétriques standardisées, dans la perspective de concrétiser la fonction « flexibilité » opérationnelle
dans le cadre de « l’Industrie 4.0 ».
Afin de déployer la flexibilité opérationnelle créée au niveau des machines, le système de contrôle de
l'atelier doit prendre en charge la flexibilité et la modularité introduites. Pour surmonter la limitation de
contrôle pour un système de fabrication amélioré et plus flexible, l'approche que nous proposons est
55
similaire à la méthode d'appel (Call up) en déportant le contrôle des fonctions logiques d'automatisation
de la couche API vers la couche supérieure de la pyramide CIM, afin qu'elles soient accessibles à
l’ingénieur de production et des opérateurs de l'atelier. Pour cela, des fonctions logiques ont été adaptées
pour être utilisées sans nécessiter l'intervention d'un travail de redéveloppement de la couche
automatique (des automates). Nous n'utilisons plus de master grafcets pour gérer les commandes des
opérations, c'est le Manufacturing Execution System MES (supervision & exécution) qui assurera cette
fonction. Concrètement, nous générons de petites séquences de production paramétriques que l'ingénieur
peut utiliser séparément comme opérations standards ou assemblées pour former des modules
fonctionnels afin de construire un plan de production complet. Ces opérations et modules standardisés
activent les fonctions logiques du contrôleur, en utilisant les protocoles d'automatisation de commandes
standardisés pour contrôler les machines.
Cette solution concernant le système de contrôle n'est pas unique. Tout d’abord, un système de contrôle
d'atelier adapté doit prendre en compte les exigences de base, en particulier la commande standardisée,
le contrôle-commande des systèmes industriels : standards ISA-88, 95,106, fournissant un contrôle
direct via de petits microcontrôleurs ou des automates standards. L’utilisation des automates standards
ou des microcontrôleurs sera muni par le déploiement des fonctions logiques ou des blocs fonctionnels.
Par la suite, et dans l’objectif d’établir une flexibilité opérationnelle, une fonction d'ordonnancement
déportée vers un orchestrateur contribuera à libérer et agrandir le processus décisionnel pour configurer
un plan de production flexible d’un produit. Par conséquence, le processus d’exécution du plan de
production pourra être dynamique et optimisable grâce à la flexibilité implémentée au niveau
commandes. Enfin la possibilité d'intégrer l'IA pour les petites tâches grâce à des modules de contrôle
intelligents est désormais possible grâce à des nouveaux modules automatiques intelligents niveaux bas
récemment présents dans les marchés industriels.
L'objectif principal de cette couche est de déployer un système de fabrication orienté services capable
de fournir des fonctionnalités de fabrication telles que la planification de la production, la gestion de la
production, l'optimisation et la supervision de la production. Ce système informatique modulaire doit
héberger des applications de service de fabrication et prendre en charge l'intégration des modules CPS
en tenant compte des contraintes d'interopérabilité. En conséquence, cette structure permettra une
décentralisation des décisions au travers de différents modules (applications / logicielles) basée sur une
architecture distribuée au niveau du site de production.
Le développement est porté sur l'implantation de la fonction flexibilité dans cette ligne de production
afin qu’elle puisse fabriquer plusieurs produits et en suivant différents cycles de fabrication dans une
configuration rapide. Un MES avancé et orienté services a été déployé dans la plateforme, où toutes les
56
données issues des machines et capteurs ont été collectées. Le système de production est donc composé
de différents logiciels et modules dédiés à des fonctions précises comprenant l'historisation et les
tendances, le module serveur d'intégration de la communication, la planification de la production, le
traitement des données et le traitement des images en IA.
Ces logiciels ne sont pas nécessairement hébergés par le MES et pour certains ont été développés en
interne notamment les modules de traitement IA qui sont connectés au système de production et peuvent
lui fournir les informations interprétées. Le système d'inspection visuelle intelligent est l’un des
exemples de modules que nous avons intégrés, qui a pour fonctionnalité d’augmenter l’autonomie du
système de production en assurant la sécurité des personnes présentes dans la zone de production. A cet
effet, une caméra d'inspection a été ajoutée à la plateforme de production et contrôlée via un algorithme
d'apprentissage automatique pour la détection humaine (El Zant et al., 2021). Lorsqu'un humain est
détecté à l'approche de machines, le système de caméra envoie une notification au MES pour interrompre
la production et assurer la sécurité des opérateurs. Ce module est faiblement couplé au système de
production, et peut être utilisé pour d’autres applications et générer des opportunités pour développer
différents modules parallèles pour le traitement des images et les intégrer dans le système de fabrication
dans un but d’optimisation. Deux modules supplémentaires ont été intégrés dans l’objectif de valider
l’architecture informatique modulaire de notre système de production. Le premier module est le système
de vision intelligent présenté ci-avant qui sert à détecter la présence humaine dans le champs de travail
des robots en faisant référence à l’article publié ‘Enhanced Manufacturing Execution System “MES”
Through a Smart Vision System’ (El Zant et al., 2021). Le deuxième module est l’intégration et le
déploiement d’une application jumeau numérique avec deux modes fonctionnels monitoring et
simulation. Le développement et l’implémentation fonctionnelle du module jumeau numérique sera
porté par la thèse de doctorat de mon collègue CHARRIER Quentin.
L’interopérabilité entre la plateforme numérique et la couche matériel a été adapté grâce au serveur
d'intégration de communication du MES qui permet une grande variété de protocoles de communication
via des pilotes de communication. L'automate par exemple communique avec l’MES via le serveur
d'intégration des opérations Sidirect, et le scanner 3D communique via OPC UA.
Enfin, le contrôle en temps réel des opérations et l'ordonnancement se font via une application web IHM
développée via le MES. L’objectif est de fournir une interface interactive qui communique avec les
machines et les robots via leur identifiant, Cette dernière peut lancer les opérations d’usinage, les
transferts, les services de contrôle et de surveillance et enfin organiser les processus de production dans
l’ordre souhaité ainsi que de superviser et surveiller l’installation de production.
57
7.2.4 Gestion intégrée des données :
Comme évoqué précédemment, la gestion des données, y compris l'agrégation de données et leur
exploration, est un objectif essentiel de la transformation numérique des systèmes de production. Pour
ce faire, la couche atelier lui a été superposée par une couche de capteurs IoT pour collecter les données
et les informations potentielles des machines. On retrouve les capteurs de vibrations pour la maintenance
prédictive, des capteurs de consommation d'énergie pour l'optimisation énergétique des processus de
production, les capteurs de température pour l'assurance qualité et sécurité. Certaines des données
collectées sont cruciales pour la planification et la gestion de la production et peuvent être adressées au
niveau business via l’ERP. D'autres données de bas niveau sont traitées localement sur un CPS désigné,
assurant la communication machine à machine et permettant un processus décisionnel local. À titre
d'exemple, les robots avancés gèrent les magasins de transition et le transfert de produits en cours
d’usinage au niveau local à l'aide d'une base de données locale intégrée, qui est également partagée avec
le MES pour le suivi des produits. Les données du système de contrôle de l'atelier telles que les
disponibilités des machines sont bien partagées avec le MES via le réseau terrain puisque le système de
production sera en charge de la gestion du flux de processus et de la planification globale de la
production. Les données de contrôle sont également essentielles pour améliorer la sécurité des processus
de production.
Sur le plan technique, deux réseaux TCP / IP ont été mis en œuvre. Le premier concerne le niveau de
gestion et de pilotage, en connectant le MES avec d'autres modules d'applications, y compris Historian,
l'IHM et une IHM portable, ainsi qu'un lien vers le cloud. Le deuxième réseau représente le bus atelier
qui concerne les équipements et machines où deux flux de données peuvent être définis. Les données de
contrôle et de surveillance à traiter via les itinéraires des contrôleurs et le flux de transfert de données
en utilisant différents protocoles de communication qui sont collectés à partir de machines, de capteurs
et d'actionneurs. Le niveau de management et pilotage fournira au système de production des données
précieuses, en particulier des données historiques, des tendances et des indicateurs de performance clés.
Ce niveau est également conçu pour collecter des données brutes et des données exploitées à des fins
d'optimisation. Parallèlement, les modules de service d'exploitation de données sont connectés
directement au système de production en lui fournissant des données traitées, tel que le module de
caméra d’inspection intelligente. En outre, la gestion des données comprend le développement et la mise
en œuvre de plateformes cloud où les données peuvent être stockées et traitées en parallèle avec les
machines locales. La plateforme cloud peut fournir du cloud computing pour les données brutes et
apprendre pour enrichir les algorithmes d'IA locaux. Enfin, le système de production est bien connecté
au niveau business via la plateforme ERP une fois déployée où les données peuvent être partagées dans
deux directions dans un périmètre bien défini prenant en compte les questions de cybersécurité.
58
La prise en compte de la dimension de gestion intégrée des données a permis d'exploiter les données
dans l'organisation et le suivi du système de production et contribuera à la performance et à
l'optimisation puisque les flux de données sont assurés à la fois dans des directions ascendantes et
descendantes. En outre, la gestion intégrée des données permet de définir le lac de données de la
plateforme et d'agréger différents types de données, y compris les données de bas niveau (données brutes
des machines incluant les disponibilités des machines), les capteurs et les données IoT, les données non
structurées telles que les images pour le traitement d'images, le nuage de points générés le contrôle de
la qualité et finalement les données d’usinage, y compris les programmes d'usinage.
7.2.5 Ouverture pour une communication au niveau business de l'entreprise et les lignes de
production distribuées :
L'intégration des systèmes MES avec les systèmes ERP ou d'autres applications de pilotage est devenue
une mission critique pour les opérations de fabrication. En particulier, cela offre une flexibilité dans la
planification de la production. Il améliore également la visibilité de la chaîne d'approvisionnement et
l'aide à la décision en mettant à jour les applications métier avec des informations de production en
temps réel.
Ce qui comprend des capacités d'intégration et d'échange, y compris des messages XML, des fichiers
plats, des requêtes SQL qui permettent l'intégration entre plusieurs applications d'atelier et ERP ou
d'autres systèmes d'entreprise. Les extensions permettent via des plug-ins personnalisés d'étendre la
connectivité de l'ERP avec n'importe quel système ou de standardiser l'utilisation du produit dans un
scénario multi-sites. Les caractéristiques d'un tel ERP distribué portent sur une intégrité des échanges
de données, des échanges de messages flexibles le tout supporté nativement par le MES. Les utilisateurs
peuvent également bénéficier d'un historique complet des échanges de messages, de la fonctionnalité
Store & Forward, d'extensions pour une gestion personnalisée de la transformation des données ainsi
que d'interfaces web.
59
architecture informatique conduit à assurer la modularité du système et résoudre les contraintes
d'interopérabilité. Par conséquent, grâce au système informatique modulaire et à la flexibilité assurée au
niveau de l'atelier, une nouvelle configuration des niveaux de prise de décision des différentes machines
et modules a été ensuite permis, tout en assurant la fonctionnalité de décentralisation du système de
production.
L’architecture du système de production ainsi que son architecture informatique sont représentées dans
cette section. Enfin, et avant de discuter le changement de la nouvelle configuration du système de
production dans la section 7.4, le niveau de décision des machines est présenté en 7.2.3, introduisant la
décentralisation à travers la nouvelle architecture du système, et une synthèse rétrospective sur les
fonctionnalités attendues du système évolué grâce à cette intégration sera présenté en 7.2.4.
Comme le montre la Figure 17, l'architecture du système a clairement été évolué. La couche de
machines composée du tour, de fraiseuse et de découpe laser en plus des robots a une communication
de contrôle directe avec la couche PLC / API. Cette communication est établie via I / O, Profinet et
d'autres protocoles de communication industriels pour le transfert de données de bas niveau et les
commandes automatiques. Les données partagées entre les machines et l'automate sont dirigées vers le
MES via une communication établie entre le MES et l'automate en utilisant le driver Sidirect pour
Siemens Profinet, grâce au serveur d'intégration du MES. D'un autre côté, la couche IoT superposée aux
machines, y compris les accélérométriques d'Intellinova sur le tour, les capteurs de consommation
énergétique de Socomec installés sur la source d'alimentation des machines et le boîtier Sick IoT
connecté aux capteurs de position et de température sur la découpe laser communiquent tous directement
au MES via le protocole de communication Modbus. Le scanner 3D de FARO installé à proximité de la
machine de calcul communique via le protocole de communication OPC UA avec le MES sur le même
réseau. Les caméras embarquées Sensorpart sont connectées au réseau local et aux robots en Profinet
tandis que le système de caméras fixes de Cognex est intégré au système des robots.
60
Fig. 17 Architecture système de la plateforme industrielle 4.0
Une autre communication de transfert de données est établie directement à travers le réseau entre le
MES et les Machines pour le partage de programmes Gcode. Les programmes Gcode générés par l'outil
CAO / FAO SprutCam sont accessibles par le MES et partagés via le réseau afin de mettre à jour les
programmes actifs sur les machines CNC. Une caméra d'inspection IDS est connectée au système de la
plateforme sur le réseau. Le système d'inspection est installé sur la machine de calcul et se communique
avec le MES via OPC UA.
Le MES présenté dans cette architecture héberge plusieurs applications orientées services, notamment
l'IDE d'environnement de développement intégré d'Aveva System Platform, le serveur d'intégration,
l'IHM développée via Intouch OMI et l'historien. Un serveur DNS est installé sur le serveur principal
afin d'activer l'application Intouch Access Anywhere jouant le rôle d'une application IHM déportée pour
une utilisation portable via un réseau 4G séparé du local. L'application de jumeau numérique simulé est
intégrée au système de production avec le système d'inspection visuelle intelligente. Enfin, les modules
de cloud computing ne sont pas déjà déployés mais activés d'abord via l'application Azure et Aveva
Insight connectée à Aveva System Platform via le réseau 4G, et activés également sur la machine de
calcul si nécessaire. La configuration des plateformes de cloud computing sera menée par la thèse de
doctorat de mon collègue EL HELOU Marwan, travaillant sur l'intégration du Cloud Computing dans
les systèmes de fabrication industrielle.
Afin de bien clarifier l'architecture du système, l'architecture informatique mise en œuvre à la plateforme
est présentée par la Figure 18. L'architecture informatique est organisée à travers deux réseaux
61
principaux, le réseau de bas niveau (ou bus terrain) et le réseau de haut niveau. Le réseau de bas niveau
est un réseau TCP / IP incluant la couche de connexion des machines, la couche de contrôle et la couche
de gestion présentée par le serveur principal en tant que réseau de bus de terrain, où différents protocoles
de communication sont déployés selon les besoins pour chaque entité/équipements. Le deuxième réseau
connecte la couche de gestion composée de différentes machines adaptées à différentes utilisations en
fonction du rôle d'application implémenté, et permet la connexion au cloud et à l'application IHM
portable. Ce réseau est connecté à la couche business ainsi représentée par le système ERP qui n'est pas
déployé pour l'instant puisque nous nous concentrons sur la dimension du système de production niveau
atelier.
7.3.3 Modèle basé sur le niveau de décision des machines de la plateforme industrielle 4.0
62
système PLC est responsable des commandes automatiques des machines, les caméras embarquées sont
chargées de confirmer la bonne exécution des commandes telles que l'ouverture des portes des machines.
Les modules de traitement mis en œuvre sur la machine de calcul tels que le système d'inspection
intelligent et le contrôle qualité ne sont pas supervisés que par le MES, mais ces modules de calculs
échangent et partagent plutôt les informations traitées et les décisions prises localement avec le MES.
Fig. 19 Modèle basée sur le niveau de décision des machines de la plateforme industrielle 4.0
7.3.4 Synthèse
Le Framework basé sur le modèle conceptuel proposé contribue à considérer les différentes couches du
système de production d'une entreprise et recommande des nouvelles modifications avancées afin
d'assurer une évolution robuste vers la numérisation du système de fabrication. Ce Framework qui a été
63
bien appliqué à l’architecture traditionnelle existante de la plateforme et a contribué à évoluer
l’architecture du système de fabrication notamment en assurant les fonctionnalités fixées en début de ce
chapitre, la flexibilité le rendement en lead-time, la modularité, la décentralisation, et la portabilité.
La flexibilité des opérations et des processus de production a été bien évoqué dans la section 7.1.1 et
7.1.2. La nouvelle configuration des machines niveau atelier flexible ou shop floor qui est de transformer
les machines en CPS aura un effet principal sur la flexibilité opérationnelle, suivie par la nouvelle
configuration de la couche commande niveau SFCS. La flexibilité du système sera bien développée et
démontrée dans les 2 sections suivantes en déployant une nouvelle configuration du programme de
production. Enfin, l’application démonstrative présentée par la suite sera menée principalement pour
justifier l’évolution de la flexibilité du système. En plus, l’application quantitative permettra de valider
le rendement en lead-time notamment dans la phase d’industrialisation qui porte sur le développement
d’un nouveau plan de production. La simplicité en termes de définition d’un nouveau plan de production
et le gain en temps réalisé dans cette tâche est dû aussi à l’évolution de l’architecture du système ainsi
qu’à la nouvelle configuration du programme de production implémentée ensuite. La modularité du
système informatique expliquée dans la section 7.1.4 est dû à l’intégration d’une architecture de type
SOA et l’intégration des modules de traitements d’images. La portabilité de l’IHM fait partie de ces
modules informatiques implémentées dans le système sous forme d’une application informatique
Intouch Access Anywhere qui déporte l’IHM de configuration et de commandes via le réseau 4G. La
dimension de traitement de données se configure dans le fait d’intégrer les capteurs de vibrations
installés sur le tour, de collecter les données générées afin d’être traitées par des modules algorithmiques
à développer grâce à la prise en considération du niveau gestion de données intégrées. Finalement, la
décentralisation des niveaux de décision de différentes entités du système de production a été bien
présenté dans la section précédente
64
le scanner 3D pour le contrôle de la qualité et l'intégration de capteurs Iot et de systèmes de caméras
embarquées. Cette approche modulaire vise à déployer le potentiel de la nouvelle architecture système
sans changer l'ensemble du processus de contrôle existant mais en déplaçant le processus de prise de
décision vers la couche MES grâce à l’implémentation d’une plateforme numérique avancée et
interopérable.
Le potentiel du MES réside dans la capacité de créer facilement des programmes de production en
utilisant des langages de programmation standard autorisés pour développer des scripts complexes, ce
qui est très difficile à développer sur des contrôleurs logiques ou même parfois impossible. Les
fonctionnalités visées par le déploiement de l’MES ou de cette plateforme numérique sont la gestion de
la production y compris l’exécution des opérations de fabrication, et le management de la production y
compris le pilotage des plans de production modulaires, en répondant aux exigences de qualité et de
quantité des clients, et aux besoins fonctionnels du système en termes de maintenance et de management
de flux. Cela permet d'aller plus loin dans l'optimisation des processus de production, ainsi de faciliter
le développement pour les opérateurs et les ingénieurs de production sans avoir à effectuer de
modifications en continu au niveau des contrôleurs logiques et de la couche de développement de
l'atelier. Nous pouvons identifier deux cas principaux d'utilisation avec ce nouveau processus modulaire:
- Pour l'industrialisation d'un nouveau produit : la capacité à réduire les délais de mise sur le marché
pour la production d'un nouveau produit en réutilisant les modules existants.
- Pour l'ordonnancement de la production : la capacité de lancer une nouvelle production sans perte de
temps et de passer facilement d'une gamme à une autre.
65
opérations de base dans le MES est organisée en deux étapes. La première se déroule sur des systèmes
spécifiques à la machine, tels que les outils FAO dédiés à la génération des Gcode pour les machines
d'usinage CNC. En effet, toutes les opérations nécessaires à un plan de production produit sont validées
en amont. Vient ensuite la deuxième étape, qui consiste à préparer leur intégration dans le MES à travers
une action d'encapsulation afin de les rendre compatibles avec son environnement opérationnel en
s’appuyant sur des connaissances techniques avancées de la ligne de production. Un travail de
codification spécifique est mis en place pour bien déployer les opérations d’usinage et de transfert dans
les programmes de production niveau de l’MES. Par exemple, les opérations de transfert concernant les
robots fixe et mobile sont définis par 33 trajectoires utilisables de manières individuels et combinées.
Ces trajectoires de transferts sont développées pour couvrir les fonctions de transferts nécessaires entre
les différentes zones de travail incluant les différents magasins et les différentes machines présentes à la
plateforme.
La 3ème phase appelée "Orchestration du Plan de Production d'un Produit" est dédiée à la finalisation
d'un plan de production pour un produit donné. Cette phase consiste à orchestrer les modules pour
générer le flux de production d'un produit, définir les contraintes d'antériorité entre les modules, et mettre
en place le contrôle qualité du produit. L'ingénieur de production rassemble tous les modules nécessaires
66
à l'exécution d'un ordre de fabrication donné et impose les contraintes d'antériorités inter-modules.
Certains modules sont configurés, notamment les modules de contrôle qualité, qui doivent être activés
proportionnellement au nombre de composants fabriqués. Dans cette phase, le MES doit proposer
plusieurs options, la première consiste à prendre en compte les contraintes imposées entre les modules
et à proposer tous les scénarios possibles d'ordonnancement des modules de production pour un produit
défini.
Enfin, la 4ème phase « Planification des processus de production » ou « pilotage » est une suite logique
de la phase précédente. L'objectif ici est de préparer le lancement et l’exécution de la production et de
s'assurer qu'elle est gérée en temps réel.
La plateforme est contrôlée via Wonderware System Platform WSP - produit Schneider («Wonderware
System Platform - Fondation de l'Industrie du Futur», 2017), où sont collectées les alarmes, les tendances
et toutes les données des machines et capteurs Fig. 22. La ligne de production, grâce au déploiement de
l’intégration verticale et la nouvelle architecture flexible, est capable de produire différents produits et
de piloter différents plans de fabrication.
Grâce à la configuration flexible des machines dans l'atelier, le développement des opérations
paramétriques standardisées devient possible. Plusieurs opérations paramétriques configurées ont été
créées représentant l'ensemble des opérations de production adaptées au potentiel de notre plateforme
industrielle 4.0. La dimension paramétrique exprime une variabilité qui permet à l'opération de s'adapter
au contexte, en particulier à la variabilité des composants fabriqués. Les opérations peuvent être définies
de différentes manières, tout dépend de la standardisation recherchée. En ce qui concerne notre
plateforme, nous avons choisi de regrouper les opérations en 4 catégories, résumées dans la première
colonne du Tableau. 2. L'intérêt ici est d'identifier toutes les opérations nécessaires à la fabrication, aux
contrôles, aux transferts, à l'assemblage, etc. qui permettent d'assurer la production d'un composant, à
partir du stock initial jusqu'à la livraison du produit fini, et du produit contrôlé où il pourrait être récupéré
à partir de la table d'accumulation. En d'autres termes, nous organisons les processus de production en
opérations de production multi-tâches comme suit :
- Opérations d'usinage incluant les opérations d'usinage CNC (codes G pour les procédés d'usinage) et
les opérations de découpe laser.
- Surveillance des opérations à l'aide de caméras fixes et embarquées incluant les contrôles de sécurité,
le stockage et les contrôles des anomalies. Le système de caméra fixe au-dessus du magasin initial et du
magasin de transition aide au contrôle du stockage et à la gestion des pièces brutes et fabriquées. La
caméra de supervision permet de surveiller la ligne de production en détectant toute présence humaine
à proximité des robots pour alerter l'opérateur et mettre les machines en pause. Enfin, des contrôles de
sécurité complémentaires sont effectués en complément grâce aux caméras embarquées sur les robots
pour assurer la disponibilité d'un espace de travail pour déposer une pièce ainsi que l'ouverture des portes
des machines pour éviter toute collision.
68
operations Functions Materials
Robotic transfer between the transition store and the machines Kuka Agilus 1100
Robotic transfer between machines and the 3D Scanner Kuka Agilus 1100
Robotic transfer between initial and transition store Kuka Agilus 900
Milling Tormach
Universal Laser
Engraving - ULS
System
Supervision
Monitoring Anomaly control
Camera
Certaines opérations peuvent nécessiter une configuration spécifique, notamment en ce qui concerne la
flexibilité de la plateforme. Par exemple, la pince du robot en termes de course et de forme doit gérer
différentes dimensions de pièce. En effet, la typologie des différents composants qui seront fabriqués
suggère que l'on conserve la "pince" comme paramètre d'entrée pour transférer les programmes. Les
trajectoires du robot pour le transfert et le chargement des machines sont créés comme des entités de
trajectoire qui peuvent être assemblées et utilisées dans n'importe quel processus de production. Pour
cela, nous avons défini des zones au sein de la plateforme, qui seront les points de départ et d'arrivée
des trajectoires du robot. Ces points sont appelés points de bouclage et seront utilisés au début et à la fin
de chaque trajectoire et garantiront que chaque trajectoire peut être effectuée avant ou après toute autre
trajectoire. En effet, les chemins entre chacun de ces points de bouclage ont été testés pour éviter tout
69
risque de collision avec l'équipement de la cellule. Des trajectoires supplémentaires dédiées à des
fonctions spécifiques ou à des questions de sécurité ont également été ajoutées, notamment des
séquences de contrôle de sécurité de base, telles que le contrôle de l'ouverture des portes des machines
vers lesquelles elles se dirigent. Les trajectoires choisies pour la programmation correspondent à chaque
chemin possible reliant différentes zones Fig. 23 et chaque machine.
Tableau. 3 illustre dans la deuxième colonne un exemple d'opérations de production standardisées qui
peuvent être demandées dans différents programmes de production. La dernière colonne du tableau
présente quelques paramètres dédiés aux opérations correspondantes. Une trajectoire de robot est
paramétrée par l’ID de la pièce, le diamètre de la pièce, la hauteur de la pince du robot depuis le bas de
la pièce et la vitesse du robot. L'opération d'usinage CNC est d'abord paramétrée par l'ID correspondant
à la référence de sous-programme, la profondeur de coupe et la vitesse de coupe.
Robot K900 W_RF_Pick_InitStore Pick the part from the Initial Store ID, Ø, H, S
Robot K900 W_RF_Drop_TransStrore Drop the part in the Transition Store ID, Ø, H, S
Robot K1100 W_RM_Pick_TransStore Pick the Part from the Transition Store ID, Ø, H, S
70
Robot K1100 W_RM_Drop_Tour Drop the part to the Lathe ID, Ø, H, S
Embedded ID, Ø, H, S
Camera W_RM_Control_Tour Presence check of the part in the Lathe
Robot K1100 W_RM_Pick_Tour Pick the part from the Lathe ID, Ø, H, S
Frequency
°
Scanner FARO Prog_Frao_1 Start Scanning (referenced part program) /°
Robot K1100 W_RM_Drop_TransStore Drop the part to the Transition Store ID, Ø, H, S
Robot K900 W_RF_Pick_TransStore Pick the part from the Transition Store ID, Ø, H, S
Robot K900 W_RF_Drop_Conv Drop the part to the final conveyor ID, Ø, H, S
7.5.2 Modules
En effet, le MES transforme les opérations de production de base des machines interconnectées, robots,
scanners, automates, actionneurs et capteurs, en modules qui encapsulent des programmes machine mais
aussi des informations et des commandes liées à son environnement opérationnel. L’ingénieur de
production et l’opérateur ont l’opportunité de créer des modules fonctionnels qui regroupent jusqu’au 6
opérations paramétriques standardisées. Cependant, et afin de mieux conserver la flexibilité du système
et augmenter son autonomie, c’est potentiellement préférable de concevoir les modules en une opération
d’usinage menée d’une opération de transfert pour alimenter ou récupérer la pièce de la machine et une
opération de contrôle au maximum. Cette stratégie de la configuration des modules fonctionnels permet
d’augmenter la flexibilité du système de production notamment par sa capacité à réorganiser le plan de
production afin d'éviter la survenue d'événements indésirables sur la plateforme, et offre la possibilité
d'une optimisation renforcée dans la phase d'ordonnancement des modules de production. Les
contraintes préalables d'interopérabilité facilitent leur regroupement en un module ; ainsi, le paramétrage
de certaines opérations est transféré au module qui renvoie la valeur du paramètre à l'opération
concernée en phase de production.
On voit dans le Tableau 4 que certains modules regroupent une seule opération, tandis que d'autres
modules regroupent plus d'une opération pour diverses raisons. Nous distinguons trois catégories de
71
modules. La première catégorie considère les modules de transfert, qui peuvent regrouper une ou
plusieurs opérations de transfert concernées par les robots. Les modules de transfert tels que le premier
et le quatrième module répertoriés dans le tableau 4, concernent le transfert de pièces entre les machines
et les différents magasins ainsi que les fonctions d'assemblage. Le deuxième type considère les modules
de process concernés par les opérations d'usinage, la contrôle qualité et le contrôle de surveillance. Ces
modules se veulent uniques et spécifiques à l'usinage d'un composant précis. De plus, les modules de
process tels que le module "Usinage de tour - Pion" sont censés être indépendants des modules de
transfert en termes de fonctionnalités afin d'améliorer la flexibilité du système et de libérer les robots
pour effectuer d'autres tâches. D'autres modules de process ont des paramètres de pourcentage qui
définissent la fréquence de la demande d’un module concernant une famille de produits telle que le
module de contrôle de présence présenté dans le tableau 4. La dernière catégorie prend en compte les
modules personnalisés. Cette catégorie permet à l'opérateur de développer un module personnalisé en
regroupant plusieurs opérations de transfert et de process pour des besoins fonctionnels spécifiques.
La possibilité de créer de tels modules séparés, grâce au processus modulaire développé, offre au
système de fabrication un niveau une flexibilité considérable puisque plusieurs modules pourraient être
lancés en parallèle en fonction de la disponibilité des machines d'abord et de la gestion des files d'attente
générée par le système de fabrication basé sur le critère d'optimisation (temps, consommation électrique,
priorité machines, …). Le travail de développement de ces modules nécessite des compétences
multidisciplinaires en automatisation, robotique et usinage, d’où l’intérêt de ces travaux d'ingénierie, car
une fois terminé, aucun développement supplémentaire ne sera plus nécessaire.
Machining lathe -
Prog_Tour_1 Start the machining program Lathe
Pawn
72
7.5.3 Orchestration du plan de production
La 3ème phase de cette section est celle de la planification des modules de production pour obtenir un
plan de production numérique et ainsi orchestrer les modules de production. Nous distinguons dans notre
approche 3 versions de plans de production. Le primaire appelé "initial" qui prend en compte les
contraintes précédentes entre les modules. Ce premier plan n'est pas bien optimisé puisqu’il vient
enclencher les modules définis d’une façon séquentielle. La deuxième version du plan de production,
dite « dynamique », devra être soumise aux contraintes de capacité de la plateforme 4.0 et aux
disponibilités des machines. Par conséquent, le MES sera capable de piloter la production en exécutant
plusieurs plans de production en parallèle en gérant le management des flux. Le système MES pourra
planifier son exécution en tenant compte de la charge de production actuelle en gérant les flux d'attentes
(Queuing Management). La 3ème version du plan de production général consiste à optimiser les
opérations d’usinages et de transferts en fonctions de temps, de cout ou d’énergie en se basant sur le
traitement de données collectées des différents capteurs et des IoT superposées sur les machines.
L’optimisation des différents processus de production peut être mono-objective qui minimise la
consommation énergétique par exemple, ou multi-objective qui minimise le temps de la production tout
en optimisant la consommation énergétique.
La figure 24 illustre un plan de production pour un pion organisé en plusieurs scénarios de flux
possibles. Le flux primaire transfère la pièce usinée du mandrin de tour au scanner 3D, active le scanner
pour vérifier les écarts dimensionnels, puis la transfère du scanner au convoyeur. Le deuxième flux
transfère la pièce usinée directement du mandrin de tour au convoyeur sans parcourir le scanner 3D. Ces
flux alternatifs sont conditionnés par la fréquence de contrôle dimensionnel définie pour ce produit
spécifique. Ces paramètres, contrôlés par le MES sont souvent fixes ou variés en fonction de l'évolution
des écarts dimensionnels observés dans le temps. L'historisation des écarts dimensionnels des pièces
usinées par rapport à leur modèle natifs CAO peut être une véritable source d'optimisation pour prendre
en charge la production à une cadence optimale. En effet, le contrôle interne dynamique permet d'une
part de limiter les interruptions de réglage et d'autre part de limiter le nombre de pièces rejetées pour
non-conformités dimensionnelles et géométriques
73
Fig. 24 Exemple d’un plan de production du pion
Enfin, le contrôle en temps réel des opérations et l'ordonnancement se font via une IHM développée via
Wonderware System Platform WSP. Les applications logicielles WSP offrent une intégration améliorée
et fournissent une plateforme de services d'applications industrielles commune et stratégique basé sur la
technologie d'architecture orientée services SOA en temps réel, telle que le module IDE « ArchestrA »
et le module IHM « OMI ».
Cette intégration fournit un environnement de développement pour MES, SCADA, IHM, historisation
et autres applications de services de fabrication à l'aide d'une plateforme logicielle SOA unifiée. Ces
nouvelles solutions logicielles unifiées sont conçues pour aider les fabricants à augmenter la capacité
des actifs existants grâce à des optimisation dans l’exploitation des moyens (assets) (“Wonderware
System Platform - Fondation de l’Industrie du Futur,” 2017).
74
Fig. 25 IHM de production de la plateforme 4.0
Les plans de production, sélectionnés et validés lors de la phase précédente, seront chargés et positionnés
par ordre d'exécution prioritaire. Le MES gère la synchronisation des plans de production avec la charge
de capacité de la plateforme.
Dans notre approche, la synchronisation se fait au niveau des modules, apportant la flexibilité
opérationnelle à la flexibilité modulaire. Dans sa phase de préparation, le MES prend en compte le plan
de charge de chaque machine, les arrêts programmés, les stocks, l'ordonnancement initial, la
disponibilité des documents de référence, la constitution des lots, etc. équilibrage temporel des flux, des
pannes, des écarts mesurés, des contingences, de la traçabilité et de la libération des lots, etc.
Pour cela, le MES développé ici sera doté d'un niveau d'autonomie et doit donc pouvoir prendre des
décisions basées sur des données ascendantes presque en réel temps. Ainsi, la commande des modules,
ou regroupement de modules, devient dynamique et sa mise à jour dépend d'une part de l'évolution de
la situation en temps réel et d'autre part des aléas rencontrés.
75
7.6 Application démonstrative
Suite au développement de l'interface expérimentale, une phase de mise en service a été organisée pour
vérifier le bon fonctionnement des opérations modulaires. Concernant les fonctions de transfert, toutes
les combinaisons possibles ont été testées afin de valider le passage sécurisé et réussi des robots entre
toutes les machines et les étapes / zones de production. En ce qui concerne les machines CNC, le G code
flexible est testé aussi bien sur le tour que sur la fraiseuse. Le pilotage est confirmé par la possibilité de
modifier la sélection du sous-programme (à partir d'une liste de 8 sous-programmes code G actifs sur la
machine) correspondant à un produit spécifique.
Afin de valider l'architecture proposée et de la comparer à l'approche traditionnelle, plusieurs tests ont
été menés par un groupe de stagiaires et d'étudiants en ingénierie de production. Le 1 er objectif des
expérimentations menées est de valider la simplicité de la création d'un nouveau plan de production pour
un nouveau produit commandé ainsi que les délais importants réservés en phase d'industrialisation pour
configurer un plan de production. Le 2ème objectif est de valider la fonctionnalité de la flexibilité de la
nouvelle architecture de notre système de production en comparaison avec l’ancienne architecture. Pour
cela, les expérimentations sont dirigées en deux étapes, considérant deux scénarii. Le 1 er scénario
concerne la définition d’un nouveau plan de production d’un Pion « A ». Le 2ème scénario concerne la
réception d’une nouvelle commande d’un Pion « B » avec une variante dimensionnelle par rapport au
pion « A », sachant que les pinces montées sur les deux robots fixe et mobile sont identiques et acceptent
un intervalle de diamètre entre 20 et 35 mm. Le Pion « A » et le Pion « B » ont le même diamètre de
base de 25 mm.
76
7.6.1 Etape n. 1 :
L'expérimentation est soutenue par une interface utilisateur créé sur l'IHM pour guider l'opérateur dans
la configuration des opérations, des modules d’usinage et de transfert avec des descriptions et des
annotations. Une pièce d’échecs ‘Pion A’ a été identifié pour être fabriquée et produite par la plateforme
dans sa nouvelle configuration. L'ingénieur de production utilise l'IHM pour créer un nouveau produit
appelé Pawn A. La première étape consiste à attribuer les différentes opérations de transfert nécessaires
à la manipulation de la pièce par les robots, ainsi qu'à sélectionner les paramètres appropriés
correspondant à la dimension de la pièce brute, sachant que toutes les opérations de transfert possibles
sont déjà définies, répertoriées et bien indexées sur l'IHM. Concernant le paramétrage des opérations de
transfert identifiés pour le Pion A sont identiques pour toutes les trajectoires, comme suivant :
- ID = 1001, l’identifiant 1001 a été assigné à ce Pion A qui désigne le produit numéro 1 de la
série 1
- Ø = 25, le diamètre du Pion A est de 25 mm
- H = 5, la hauteur de la prise de la pièce entre la base de la pince et la base de la pièce est de 5
mm. Cette valeur correspond également à la hauteur de la prise en la récupérant du mandrin du
tour, sachant que le mandrin porte la pièce à 3 mm de sa base
- S = 15%, correspondant à 15 % de la vitesse maximale des robots y compris les mouvements
linéaires et Point to Point (PtP)
Concernant les opérations d'usinage, l'opérateur génère les programmes d’usinage sous forme de G-code
correspondant au produit en fonction de son modèle 3D à travers une application FAO. Les programmes
d’usinage créés et bien indexés sont importés vers le module IHM de l’MES et partagés par la suite avec
les machines CNC en mettant à jour la liste des 8 sous-programmes actifs. Une fois les fonctions de
transfert et les opérations d'usinage configurées et affectées au produit, l'opérateur génère des modules
fonctionnels regroupant les opérations configurées comme expliqué précédemment. Ensuite, l'opérateur
crée le plan de production comme présenté par la Figure 24, en ajoutant un ensemble de modules
présentés par une zone de liste, et en orchestrant les modules générés sur la base du produit sélectionné.
La dernière phase consiste à choisir le produit configuré dans l'interface du plan de production où
l'opérateur peut ajouter la quantité requise du produit et lancer la production.
7.6.2 Etape n. 2 :
La 2ème étape consiste à créer un nouveau plan de production d’un Pion B. Une fois le plan de production
du Pion A est bien configuré, nous supposons à présent l’arrivée d’une nouvelle commande de type Pion
« B ». Les deux pions demandés ont la même forme géométrique, cependant, le Pion B marque une
différence de hauteur de 10 mm comparativement au Pion A, voire Figure 26. L’objectif de choisir les
mêmes pièces avec des hauteurs différentes répond au besoin de changement que nous souhaitons
77
impulser dans le plan de production associé. Ce qui démontre de la capacité à répondre aux demandes
personnalisées des clients. La hauteur du Pion A, X1 vaut 50 mm, tandis que la hauteur du Pion B, X2
vaut 40 mm. Par conséquent, la hauteur Y1 de la base cylindrique du pions A est de 8 mm et celle du
Pion B est de 6.5 mm.
Une nouvelle tâche de création du plan de production du Pion B est requise au travers de l’IHM de
l’MES développé. Les opérations de transfert requises pour l’usinage du Pion B sont les même que
celles du Pion A :
- RF_Pick_InitStore : correspondant à la prise de pièce brut du magasin initiale par le robot fixe
- RF_Drop_TransStore : correspondant à placer la pièce brute dans le magasin intermédiaire
toujours par le robot fixe
- RM_Pick_TransStore : correspondant à la prise de la pièce brute du magasin intermédiaire par
le robot mobile
- RM_Drop_Lathe : correspondant à alimenter la pièce brute dans le mandrin du tour
- RM_Pick_Lathe : correspondant à récupérer la pièce usinée du mandrin du tour
- RM_Drop_TransStore : correspondant à retourner la pièce usinée dans le magasin intermédiaire
- RF_Pick_TransStore : correspondant à la prise de la pièce usinée par le robot fixe
- RF_Drop_Conv : correspondant à remettre la pièce usinée et finie sur le convoyeur des produits
finis vers la table d’accumulation par le robot fixe
En revanche, les changements requis au niveau des opérations de transfert se trouvent dans les
paramètres. En effet, le diamètre et la vitesse des robots restent identiques par rapport au Pion A, tandis
78
qu’un nouvel identifiant (ID2 = 2001, pièce numéro 1 de la série numéro 2) du Pion B sera généré par
l’MES et associé à ce nouveau produit, et un changement de la hauteur de prise sera requis
spécifiquement pour la trajectoire RM_Pick_Tour (ID2, Ø, H2, S) qui récupère le Pion B usiné dans le
tour. Puisque le mandrin tient la pièce sur une profondeur de 5 mm, la cote restante de la base cylindrique
du Pion B sera de 1.5 mm, ce qui entraine une difficulté de la prise de la pièce et un risque de faire
tomber la pièce dans la machine. En conséquent, la hauteur de la prise de la trajectoire RM_Pick_Tour
sera réajustée à la valeur 15 (H2 = 25 mm), dans le but de récupérer le Pion B en tenant le corps de la
pièce avec le profile sphérique en haut de la pièce. Par la suite, un deuxième g-code sera généré par
l’outil FAO, importé à l’MES et envoyé directement au tour en mettant à jour la liste des programmes
d’usinage actifs sur la machine. Le plan de production du Pion B sera créé en assemblant les modules
fonctionnels de transfert et d’usinage en suivant la même logique du regroupement des opérations de
base que le Pion A.
La dernière phase consiste à choisir les deux produits configurés Pion A et Pion B successivement dans
l'interface de pilotage de la production. Le volume de production pour cette expérimentation est limité
à pièce par produit. La production a été exécutée avec succès suivant le plan de production ‘initial’
correspondant à une exécution séquentielle automatique.
En effet, la quantification pour ce type de projet est difficile à réaliser notamment les travaux
d’intégration et de l’évolution des différents niveaux du système de production. A cet effet et afin de
quantifier notre contribution, nous avons proposé un référentiel estimatif de comparaison des
performances pour valider les fonctionnalités du système de production. Ce référentiel estimatif de
comparaison considère principalement la flexibilité et le rendement en lead-time pour la phase
industrialisation et configuration du plan de production d’un nouveau produit ainsi que la phase
d’exécution de la production. Le tableau 5 représente une estimation de mesure proposée pour pouvoir
évaluer la nouvelle performance du système de production à la plateforme 4.0 en termes de flexibilité et
de rendement en lead-time en comparant l’architecture évoluée en 4.0 avec l’approche traditionnelle
3.0.
79
Tableau. 5 Référentiel de mesure proposé pour l’évaluation de la nouvelle performance
La création du plan de production d’un produit comprend des éléments de mesures qui concernent
plusieurs taches :
- La configuration des opérations d’usinage consiste à générer les g-codes associés au produit ce
qui est requis dans les 2 approches. L’estimation du temps requis pour effectuer cette tâche est
correspondant à l’outil FAO utilisé en termes de simplicité et aux compétences de l’opérateur
(estimé à une heure) ; tandis que la possibilité de paramétrage du programme d’usinage dans
l’approche 4.0 marque un avantage par rapport à l’approche traditionnelle. La possibilité de
paramétrer les programmes d’usinage dans l’approche 4.0 exige une tache supplémentaire
d’encapsulation des programmes générés suivant un protocole bien défini. Le temps
supplémentaire nécessaire pour effectuer cette configuration est estimé à 30mn
- La configuration des opérations de transfert consiste à générer les trajectoires robots associées
à la fabrication d’un nouveau produit en tenant compte de ses spécifications dimensionnelles.
Cette tâche est nécessaire à être effectuer dans l’approche traditionnelle vue du manque de
80
paramétrage des trajectoires possible, ce qui exige des couts supplémentaires grâce à
l’intervention d’un roboticien et des pertes de temps estimées à une journée de travail. Dans le
cas d’un atelier 4.0 flexible, les trajectoires sont conçues pour servir une diversité de produits et
sont disponibles avec une configuration flexible et paramétrique ce qui marque un avantage
important par rapport à l’approche traditionnelle en termes de temps et de couts.
- La configuration des modules fonctionnels ainsi que l’orchestration du plan de production d’un
produit et la planification de la production sont regroupées dans la 3ème tache concernant la
création du plan de production. La configuration de cette tâche est requise dans l’approche
traditionnelle par le fait de créer des équations de sorties et des grafcets associées à un produit
menées des conditions de passage en termes de paramétrages ce qui exige le besoin de
l’intervention d’un automaticien marquant une majoration de cout et de temps estimé à une
journée de travail pour la création des nouveaux programmes automatiques de la production.
Dans la nouvelle architecture 4.0 il suffit d’effectuer ces configurations à travers l’IHM dédiée
à la production sans le besoin d’un effort supplémentaire concernant la couche automatique et
sans aucune modification requise pour la couche de commandes.
- L’accès à distance aux machines d’usinage et aux robots est limité à l’exécution automatique
des machines à distance estimé à un taux d’accessibilité maximum de 50% dans l’approche
traditionnelle, tandis que ce taux s’élève à 80% - 90% pour les machines d’usinage et les robots
respectivement dans la nouvelle approche interopérable. Cette estimation s’est basée tout
d’abord sur la possibilité de paramétrer les fonctions de transfert des robots et les programmes
d’usinage grâce à la nouvelle configuration de ces machines, la possibilité de sélectionner et
exécuter une liste de programmes échangés à distance avec les machines, et la création d’une
base de données embarquée sur les robots qui gère les locations des produits à la plateforme et
échange ces informations avec le système de production. Du point de vue fonctionnel, le fait
d’évoluer les configurations machines / robots au niveau shop floor vers un une configuration
CPS tout en résolvant les contraintes d’interopérabilité comme expliqué dans la section 7.2.1 a
contribué à améliorer le fonctionnement et le niveau d’accès des machines.
- La tache mise en service est pré-requise à l’exécution de la production dans le cas de l’approche
traditionnelle pour valider les nouveaux programmes automatiques développés et vérifier
l’assignation des variables des entrées / sorties. Cette tâche est estimée à une demi-journée de
travail supplémentaire et peut s’élever à une journée complète contre une heure de validation
générale de fonctionnement des machines dans la nouvelle approche 4.0.
81
- L’exécution automatique à distance est désormais possible dans les 2 approches, mais
l’exécution successive automatique de plusieurs produits n’est pas autorisée dans l’approche
traditionnelle grâce à la nécessité du déploiement manuel des programmes automates produit
par produit. La possibilité d’optimisation globale des processus de production marque une
différence pour l’approche 4.0 estimée à un minimum de 80% grâce à la génération des données
et la possibilité de les traiter.
- La supervision de la production et la traçabilité des produits sont estimées à 99% en mode 4.0
contre 50% dans l’approche traditionnelle, grâce à l’intégration de l’MES et l’échange de
données générées des différentes postes de travail en cours de la production.
L'approche traditionnelle qui permet de générer un nouveau plan de production pour un nouveau produit
consiste à créer toutes les opérations d'usinage et de transfert correspondantes (et parfois à partir de zéro)
nécessaires à la fabrication du produit. Ensuite, un technicien ou un ingénieur de contrôle automatique
est nécessaire pour attribuer des fonctions logiques aux opérations configurées et par conséquent créer
un nouveau programme de contrôle / par produit, et le déployer sur le contrôleur / automate PLC, suivi
d'une phase de mise en service.
En reprenant la 1ère étape de l’application menée, la simplicité des différentes tâches et sous tâches,
mesurées dans le référentiel, de la création et de la configuration d’un nouveau plan de production a été
bien validée. En effet, la réalisation d’un nouveau plan de production en se basant sur l’ancienne
architecture traditionnelle de la plateforme, nécessite tout d’abord le développement d’un nouveau
programme au niveau de l’automate sous forme d’un grafcet qui vient déployer les différentes fonctions
d’usinages et de transfert robotique, associées à un produit bien défini. Une phase de mise en service
sera toujours exigeante afin de vérifier les assignations des variables logiques correspondantes aux
différentes opérations, et de vérifier les nouvelles opérations configurées en termes de fonctionnement.
La 2ème étape de l’application valide de nouveau la simplicité et la rapidité de la création d’un nouveau
plan de production. Elle a mis en évidence la flexibilité de la nouvelle configuration du système de
production au niveau Planification / Pilotage ainsi qu’au niveau shop floor. En réalisant le même
scénario sur les deux produits Pion A et Pion B avec l’ancienne approche traditionnelle, trois dimensions
techniques se rajoutent. La 1ère dimension concerne la contrainte issue du besoin de récupérer la pièce
usinée à une hauteur de prise spécifique au Pion B, et différente de celle de base configurée. Cette
contrainte exige deux taches techniques dans la configuration traditionnelle de la plateforme. La 1 ère
tache technique requise est tout d’abord la création d’une nouvelle trajectoire spécifique pour récupérer
ce produit du tour qui reprend les mêmes déplacements de la trajectoire basique en changeant le point
d’arrivée sur le mandrin de la machine en axe Z. la 2ème tache technique concerne l’assignation d’un
82
nouveau cycle code à cette trajectoire et son déploiement à travers la création d’une nouvelle fonction
logique à implémenter dans le grafcet correspondant à ce produit. Cette difficulté a été surmontée dans
la nouvelle architecture de la plateforme en paramétrant les fonctions de transfert robotique. La 2 ème
dimension technique concerne le besoin d’activer un nouveau programme d’usinage g-code sur le tour.
Le chargement du nouveau g-code dans la machine était manuel, ainsi que sa sélection. Cette contrainte
a été résolue grâce à l’interopérabilité déployée au niveau de la plateforme qui a facilité l’échange entre
le MES et les différentes machines d’usinage. La 3ème dimension considère le rôle avancé de l’MES dans
la planification et le pilotage de la production d’une diversité de produit. Toujours en comparant
l’application menée avec l’ancien fonctionnement de la plateforme, il fallait lancer le 1er programme de
production du Pion A qui figure dans le programme de l’automate, ensuite déployer le 2ème programme
de production du Pion 2 au niveau de l’automate et lancer le cycle de la fabrication du 2 ème produit
simultanément. Le fait qui marque une flexibilité accrue dans la nouvelle architecture et un niveau de
pilotage considérable en termes d’autonomie.
Les fonctions de transferts sont bien configurées et la conception flexible de ce système de production
a bien fonctionné, donnant à l'opérateur la possibilité de piloter facilement le plan de production dans
diverses configurations et combinaisons via une simple IHM sans aucun codage supplémentaire ni effort
technique requis comme le montre le référentiel de mesures. Trois dimensions principales méritent d'être
mentionnées en termes de comparaison de la nouvelle approche avec l'approche traditionnelle.
Principalement les compétences techniques, y compris la dimension « complexité » et la dimension
« gain de temps en phase d’industrialisation ». Dans la première dimension l'opérateur ou l'ingénieur de
production n'a pas besoin d'avoir des compétences spécifiques en contrôle, robotique et autres domaines
techniques pour la génération d’un nouveau plan de production, réduisant ainsi les coûts de l'intervention
des intégrateurs externes. Le gain de temps est assuré en réduisant le temps de configuration nécessaire
à la phase d'industrialisation et d’exécution de la production avec une estimation de 14 heures en moins.
Ainsi, La réduction du temps réalisée est liée aussi à minimiser le besoin d'une phase de mise en service
puisque la nouvelle approche est bien flexible et conçue en amont pour gérer une diversité de produits
en tenant compte des capacités machines. La troisième dimension se concentre sur la flexibilité du
système de production évolué. Cette flexibilité déployée dans le système, est due grâce à l’intégration
du Framework et l’évolution des différentes couches de notre système de production y compris la couche
shop floor mais aussi bien à l’évolution la couche de la commande, de pilotage et de planification.
Les travaux présentés dans cette section montrent que le MES joue un rôle important dans les processus
de fabrication intelligents, et démontrent également que la notion de la flexibilité opérationnelle doit
être prise en compte dans la phase de conception du système de production avancé, en particulier lorsque
l'on travaille dans un système de fabrication plus complexe avec plus de machines, d'appareils et de
capteurs. Le Framework réalisé a montré la pertinence de nos choix méthodologiques et de nos
83
orientations de recherche. Les perspectives en termes de développement consistent à développer
l'autonomie de la fonction « pilotage » ainsi que déployer des algorithmes d'optimisation de production
multicritères.
84
8. Résultats et discussion générale
Le travail effectué lors de cette thèse de doctorat est une nouvelle approche permettant une flexibilité
accrue d'une ligne de production. Cela nous permet de contrôler facilement la fabrication de divers
produits via une IHM plutôt que de reprogrammer les contrôleurs industriels normalement requis pour
permettre un nouveau processus de fabrication de produit. La contribution de cette thèse se situe d'abord
dans l'intégration réussie du système de production basé sur le modèle proposé et le Framework
développé. Deuxièmement, le développement de cette approche flexible et modulaire basée sur plusieurs
technologies avancées telles que la robotique avancée, les systèmes de caméras embarquées et les
machines-outils flexibles, ainsi que la mise en œuvre d'une nouvelle architecture de contrôle qui consiste
à faire évoluer le processus de prise de décision vers un système MES avancé et interopérable. Le
fonctionnement de notre architecture a été testé avec succès. Malgré des difficultés d'interopérabilité,
nous avons réussi à développer un système de reconfiguration de l'espace de production automatisé pour
fabriquer une large gamme de produits à partir des mêmes matières premières.
Suite au développement de l'interface MES expérimentale, toutes les combinaisons possibles ont été
testées, afin de valider le passage sécurisé des robots dans les différentes zones et réussi entre toutes les
machines et les étapes de production. La conception modulaire de ce système de production a bien
fonctionné, donnant à l'opérateur la possibilité de piloter facilement le plan de production dans diverses
configurations et combinaisons via une simple IHM sans aucun codage supplémentaire ni effort
technique. Cela valide des bénéfices considérables et intéressants au niveau de la flexibilité du système
de production et un délai remarquablement réduit pour intégrer un nouveau produit au système de
production dans la phase d'industrialisation.
Le nouveau module proposé pour un système visuel intelligent bien intégré dans le système d'exécution
de fabrication ainsi que le module jumeau numérique contribuent à améliorer le système de production.
Cette intégration est une contribution concrète qui peut être utilisée pour plusieurs applications et assurer
l'optimisation du processus de production et du contrôle. Ces modules nous offrent de larges
85
opportunités pour développer différents modules parallèles pour le traitement des données et des images
et les intégrer dans le système de fabrication afin d'améliorer et d'optimiser l'ensemble des processus de
production.
Enfin, la dimension des données est bien prise en compte lors de la phase de développement. Cette
intégration de la gestion des données en parallèle avec la nouvelle configuration flexible de la plateforme
permettra la possibilité d'une auto-optimisation continue des processus de production à travers le
traitement des données et l'exploitation à effectuer afin d’aboutir à une configuration dynamique de
notre système de production.
En outre, et sur la base de la méthodologie développée, l'architecture proposée peut être appliquée à des
autres systèmes de fabrication avec des exigences spécifiques impliquant des robots contrôlables et
programmables jouant des fonctions de transfert. Les exigences spécifiques susmentionnées concernent
tout d'abord le degré de flexibilité des machines en atelier et leur capacité à effectuer des opérations
multiples à partir d'un nombre limité de ressources. La flexibilité des machines nécessite au moins de
résoudre des problèmes tels que l'interopérabilité, l'accessibilité ou l'accès à distance aux machines, et
permettre une configuration flexible du processus d'usinage en termes de capacité dimensionnelle pour
réaliser une marge de pièces, et en termes de paramétrisation. Deuxièmement, une attention particulière
doit être portée à la configuration des fonctions logiques au niveau de la couche API en tenant compte
de la flexibilité opérationnelle des machines. Troisièmement, la plateforme MES numérique développée
ou adaptée pour jouer le rôle de système PPC de planification et de contrôle de la production doit faire
face aux nouveaux défis créés par l'industrie 4.0. Le MES doit devenir logiquement décentralisé et
composé d'objets découplés ou d'applications avec un service externe chargé du couplage des machines
connectées et du CPS. Enfin, il convient de mentionner que l'application de la méthodologie de
conception proposée doit être guidée par une nouvelle vision de la configuration de l'atelier en tenant
compte des fonctionnalités attendues du système de production. Nous pensons que notre travail est une
base qui peut être déployé aux niveaux des entreprises de fabrication manufacturière, notamment pour
les petites et moyennes entreprises, en particulier lorsque la flexibilité opérationnelle joue un rôle
important dans un marché en constante évolution.
En termes de limitations, nous distinguons deux axes qui les catégorisent : les limitations scientifiques
de recherche et les limitations techniques concernant le développement et les travaux expérimentales.
Concernant les limitations de recherche, une attention spécifique au principe des systèmes multi-agent
pourrai enrichir ce travail de recherche. D’une autre part, la définition des compétences requises pour la
numérisation des systèmes de fabrication ainsi que la réalisation d’une synthèse de l’évolution des
métiers dans le contexte de l’industrie 4.0 pourrai améliorer ce travail. En ce qui concerne les limitations
expérimentales, il faut citer au tout début que la réalisation des travaux d’intégration et le déploiement
86
de la nouvelle architecture est un challenge principal en termes de gestion de projet d’intégration. Le
fait de générer une feuille de route à partir de cette démarche expérimentale devra être bien apprécié.
87
9. Conclusion et perspectives
Les futurs systèmes de fabrication doivent être plus flexibles et intelligents afin de répondre aux
demandes des clients et du marché. L'industrie 4.0 permet d’amener le monde industriel à un niveau
avancé de personnalisation de masse de produits et permettant un niveau élevé de la fabrication
numériques. Pour bien gérer la transition numérique, plusieurs étapes de développement sont
nécessaires, des technologies innovantes doivent être intégrées et l'entreprise doit faire face à un large
éventail de défis affectant différents composants de la structure de l'entreprise. Le défi principal est de
savoir comment guider les entreprises dans l'évolution de l'Industrie 3.0 vers le 4.0. L'objectif de cette
thèse est de proposer un Framework de transition basée sur un modèle approprié pour aider les
entreprises à réussir la transformation numérique des systèmes de fabrication Fig. 27. Ce Framework
prend en considération trois axes principaux le système de production, y compris le système de
contrôle de l'atelier SFCS, les machines et technologies avancées, et enfin le traitement et le
déploiement des données.
88
Dans cette thèse, une étude bibliographique détaillée a été menée sur l’industrie 4.0 et la numérisation
des systèmes de fabrication pour les entreprises de production. La définition et le concept de l'industrie
4.0 sont présentés, ainsi que les changements et avantages affectant le secteur manufacturier en
transformation numérique. Les axes de recherche sont bien définis et un résumé des tendances, principes
et fonctionnalités technologiques de l'Industrie 4.0 ont été fournis. De plus, les modèles d'intégration
existants ont été présentés et les défis auxquels sont confrontées les entreprises à l'ère de l'Industrie 4.0
sont bien discutés, afin de répondre à notre problématique de recherche centrale :
Quel modèle pour une transition réussie vers la numérisation des systèmes de fabrication 4.0 ?
En conséquence, l'hypothèse du doctorat a été identifiée comme l'évolution des systèmes de
production vers l'Industrie 4.0 nécessite d'agir sur trois axes principaux : le système de
production, y compris le système de contrôle de l'atelier SFCS, les machines et les technologies
avancées, et le traitement et le déploiement des données.
Cette proposition s’appuie sur la littérature, les défis et problématiques de recherche connexes abordées
dans le cadre d’un projet de recherche global porté par K. Benfriha, qui nous a proposé un nouveau
modèle conceptuel qui permet de décrire d’une manière générique un système de production 4.0. Le
modèle proposé est ensuite déployé sur un système de production conventionnel existant (plateforme
4.0) afin d'identifier les évolutions requises et de définir en conséquence le Framework de transition
concernant les actions à entreprendre pour l'objectif d'intégration numérique. La conception d’une
nouvelle architecture d'intégration vers la transition de la fabrication numérique est introduite. Le
Framework proposé contribue à considérer les différentes couches du système de production d'une
entreprise et recommande des modifications innovantes et avancées afin d'assurer une évolution robuste
vers la numérisation du système de fabrication. Ce Framework est bien appliqué à une architecture
traditionnelle existante implémentée sur la plateforme industrielle 4.0.
Finalement, une application démonstrative a été menée, au niveau la plateforme industrielle 4.0 des Arts
et Métiers, afin de valider la nouvelle architecture du système de production facilitant la transition
numérique vers l’Industrie 4.0. Cette architecture a permis l’émergence de nouvelles fonctionnalités
avancées telles que la flexibilité, et le rendement en lead-time.
Plusieurs autres travaux de recherche et de développement sont identifiés comme perspectives. Sur la
base du développement mentionné ci-dessus, nous constatons que la configuration du système de
production définissant la planification des processus doit être prise en considération avec le système
flexible évolué dans la phase de conception de l’MES ou de la plateforme numérique responsable de la
planification et de pilotage. La conception du programme de production est un domaine distinct de
recherche et de développement, ainsi une attention particulière doit y être accordée dans le contexte de
l'Industrie 4.0. La planification des processus dans le système de production numérisé est une tendance
89
cruciale et a un lien étroit avec le Lean 4.0, en particulier lorsque l'on considère les aspects logistiques
intelligents.
De plus, les développements ultérieurs se concentreront sur les opportunités d'optimisation par rapport
aux critères de temps, de coût et de consommation d'énergie grâce à des capteurs intégrés déjà montés
sur les machines. La dimension data mining et la possibilité d'avoir une auto-optimisation continue des
processus de production grâce au traitement des données activé est une piste riche de recherche. Suite à
la configuration flexible du système de production, le data mining contribue à bien quantifier et profiter
de la numérisation des processus de fabrication et valoriser les performances et les fonctionnalités visées
des systèmes de production 4.0.
Enfin, il est nécessaire d'exprimer la nécessité d'une approche multidisciplinaire dans les projets de ce
type. La transition vers l'Industrie 4.0 est le paradigme de l'intersection de multiples domaines
d'expertise et de compétences en ingénierie. Capitaliser les connaissances et les compétences requises
de chacun de ces domaines pour transformer l'industrie est, à notre avis, l'un des enjeux majeurs.
90
References
Bajic, B., Cosic, I., Lazarevic, M., Sremcev, N., Rikalovic, A., 2018. Machine Learning Techniques for
Smart Manufacturing: Applications and Challenges in Industry 4.0 11.
Benfriha, K., 2020. Conception, Modélisation, Architecture, Technologie et Pédagogique (Mémoire
HDR). ENSAM Paris.
Bloch, H., Hensel, S., Hoernicke, M., Stark, K., Menschner (neé Hahn), A., Fay, A., Urbas, L., Knohl,
T., Bernshausen, J., 2018. State-based control of process services within modular process plants.
Procedia CIRP 72, 1088–1093. https://doi.org/10.1016/j.procir.2018.03.037
Block, C., Lins, D., Kuhlenkötter, B., 2018. Approach for a simulation-based and event-driven
production planning and control in decentralized manufacturing execution systems. Procedia
CIRP 72, 1351–1356. https://doi.org/10.1016/j.procir.2018.03.204
Bodrow, W., 2017. Impact of Industry 4.0 in service oriented firm. Adv. Manuf. 5, 394–400.
https://doi.org/10.1007/s40436-017-0196-3
Bonnard, R., Vieira, K.M.M., Arantes, M.D.S., Lorbieski, R., Nunes, C., Mattei, A.P., 2019. A BIG
DATA / ANALYTICS PLATFORM FOR INDUSTRY 4.0 IMPLEMENTATION IN SMEs
10.
Borangiu, T., Trentesaux, D., Thomas, A., Leitão, P., Barata, J., 2019. Digital transformation of
manufacturing through cloud services and resource virtualization. Computers in Industry 108,
150–162. https://doi.org/10.1016/j.compind.2019.01.006
Dalenogare, L.S., Benitez, G.B., Ayala, N.F., Frank, A.G., 2018. The expected contribution of Industry
4.0 technologies for industrial performance. International Journal of Production Economics 204,
383–394. https://doi.org/10.1016/j.ijpe.2018.08.019
El Zant, C., Charrier, Q., Benfriha, K., Le Men, P., 2021. Enhanced Manufacturing Execution System
“MES” Through a Smart Vision System, in: Roucoules, L., Paredes, M., Eynard, B., Morer
Camo, P., Rizzi, C. (Eds.), Advances on Mechanics, Design Engineering and Manufacturing
III, Lecture Notes in Mechanical Engineering. Springer International Publishing, Cham, pp.
329–334. https://doi.org/10.1007/978-3-030-70566-4_52
Erol, S., Schumacher, A., 2016. Strategic guidance towards Industry 4.0 – a three stage process model
8.
Frank, A.G., Dalenogare, L.S., Ayala, N.F., 2019. Industry 4.0 technologies: Implementation patterns
in manufacturing companies. International Journal of Production Economics 210, 15–26.
https://doi.org/10.1016/j.ijpe.2019.01.004
Ghobakhloo, M., 2018. The future of manufacturing industry: a strategic roadmap toward Industry 4.0.
Jnl of Manu Tech Mnagmnt 29, 910–936. https://doi.org/10.1108/JMTM-02-2018-0057
91
Goerzig, D., Bauernhansl, T., 2018. Enterprise Architectures for the Digital Transformation in Small
and Medium-sized Enterprises. Procedia CIRP 67, 540–545.
https://doi.org/10.1016/j.procir.2017.12.257
Gorecky, D., Weyer, S., Hennecke, A., Zühlke, D., 2016. Design and Instantiation of a Modular System
Architecture for Smart Factories. IFAC-PapersOnLine 49, 79–84.
https://doi.org/10.1016/j.ifacol.2016.12.165
Gürdür, D., El-Khoury, J., Seceleanu, T., Lednicki, L., 2016. Making interoperability visible: Data
visualization of cyber-physical systems development tool chains. Journal of industrial
information integration, 4, 26-34.
Henning, K., Wolfgang, W., Johannes, H., 2013. Recommendations for implementing the strategic
initiative INDUSTRIE 4.0. Final report of the Industrie, 4, 82.
Hermann, M., Pentek, T., Otto, B., 2016. Design Principles for Industrie 4.0 Scenarios, in: 2016 49th
Hawaii International Conference on System Sciences (HICSS). Presented at the 2016 49th
Hawaii International Conference on System Sciences (HICSS), IEEE, Koloa, HI, USA, pp.
3928–3937. https://doi.org/10.1109/HICSS.2016.488
Jasperneite, J., 2012. Was hinter Begriffen wie Industrie 4.0 steckt. Computer & Automation, 19.
Jung, J.Y., 2002. Manufacturing cost estimation for machined parts based on manufacturing features.
Journal of Intelligent Manufacturing, vol. 13(4), pp. 227-238.
Kamble, S.S., 2018. Sustainable Industry 4.0 framework: A systematic literature review identifying the
current trends and future perspectives. Process Safety and Environmental Protection 18.
Kamble, S.S., Gunasekaran, A., Gawankar, S.A., 2018. Sustainable Industry 4.0 framework: A
systematic literature review identifying the current trends and future perspectives. Process
Safety and Environmental Protection 117, 408–425. https://doi.org/10.1016/j.psep.2018.05.009
Karnouskos, S., Colombo, A.W., Jammes, F., Delsing, J., Bangemann, T., 2010. Towards an architecture
for service-oriented process monitoring and control, in: IECON 2010 - 36th Annual Conference
on IEEE Industrial Electronics Society. Presented at the IECON 2010 - 36th Annual Conference
of IEEE Industrial Electronics, IEEE, Glendale, AZ, USA, pp. 1385–1391.
https://doi.org/10.1109/IECON.2010.5675482
Kaushal, A., Vardhan, A., Rajput, R.S., 2016. Flexible Manufacturing System A Modern Approach To
Manufacturing Technology 8.
Khan, W.Z., Rehman, M.H., Zangoti, H.M., Afzal, M.K., Armi, N., Salah, K., 2020. Industrial internet
of things: Recent advances, enabling technologies and open challenges. Computers & Electrical
Engineering, 81, 106522.
Kumar, K., Zindani, D., Davim, J.P., 2019. Process Planning in Era 4.0, in: Industry 4.0, SpringerBriefs
in Applied Sciences and Technology. Springer Singapore, Singapore, pp. 19–26.
https://doi.org/10.1007/978-981-13-8165-2_2
92
Lasi, H., Fettke, P., Kemper, H.G., Feld, T., Hoffmann, M., 2014. Industry 4.0. Business & information
systems engineering, 6(4), 239-242.
Lin, Y., Lan, C.-B., Huang, C.-Y., 2019. A Realization of Cyber-Physical Manufacturing Control
System Through Industrial Internet of Things. Procedia Manufacturing 39, 287–293.
https://doi.org/10.1016/j.promfg.2020.01.449
Lina, Y.J., Lana, C.B., Huanga, C.Y., 2019. A Realization of Cyber-Physical Manufacturing Control
System Through Industrial Internet of Things Through Industrial Internet of Things. Presented
at the 25th International Conference on Production Research Manufacturing Innovation: Cyber
Physical Manufacturing Cyber Physical Manufacturing, Chicago, Illinois (USA).
Mes, M., Gerrits, B., 2019. Multi-agent Systems, in: Zijm, H., Klumpp, M., Regattieri, A., Heragu, S.
(Eds.), Operations, Logistics and Supply Chain Management, Lecture Notes in Logistics.
Springer International Publishing, Cham, pp. 611–636. https://doi.org/10.1007/978-3-319-
92447-2_27
Moeuf, A., 2018. The industrial management of SMEs in the era of Industry 4.0. International Journal
of Production Research 20.
Mourtzis, D., Vlachou, E., Xanthopoulos, N., Givehchi, M., Wang, L., 2016. Cloud-based adaptive
process planning considering availability and capabilities of machine tools. Journal of
Manufacturing Systems 39, 1–8. https://doi.org/10.1016/j.jmsy.2016.01.003
Ollinger, L., Abdo, A., Zühlke, D., Heutger, H., 2014. SOA-PLC – Dynamic Generation and
Deployment of Web Services on a Programmable Logic Controller. IFAC Proceedings Volumes
47, 2622–2627. https://doi.org/10.3182/20140824-6-ZA-1003.02189
Rojko, A., 2017. Industry 4.0 Concept: Background and Overview. Int. J. Interact. Mob. Technol. 11,
77. https://doi.org/10.3991/ijim.v11i5.7072
Rüßmann, M., Lorenz, M., Gerbert, P., Waldner, M., Justus, J., Engel, P., Harnisch, M., 2015. Industry
4.0: The Future of Productivity and Growth in Manufacturing 20.
Santos, C., Mehrsai, A., Barros, A.C., Araújo, M., Ares, E., 2017. Towards Industry 4.0: an overview
of European strategic roadmaps. Procedia Manufacturing 13, 972–979.
https://doi.org/10.1016/j.promfg.2017.09.093
Schierholt, K., 2001. Process configuration: Combining the principles of product configuration and
process planning. Artificial Intelligence for Engineering Design, Analysis and Manufacturing:
AIEDAM 15, 411–424.
Schlapfer, R.C., Koch, M., Merkofer, P., 2015. Industry 4.0 Challenges and solutions for the digital
transformation and use of exponential technologies.
SenTryo, 2017. The 4 industrial revolutions [WWW Document]. URL https://www.sentryo. net/the-4-
industrial-revolutions/ (accessed 6.17.19).
93
Strategy&, 2015. Small automation, big benefits [WWW Document]. PwC. URL
https://www.strategyand.pwc.com/gx/en/functions/technology-strategy/big-benefits.html
(accessed 3.5.20).
Sung, T.K., 2018. Industry 4.0: A Korea perspective. Technological Forecasting and Social Change 132,
40–45. https://doi.org/10.1016/j.techfore.2017.11.005
Tao, F., Qi, Q., Liu, A., Kusiak, A., 2018. Data-driven smart manufacturing. Journal of Manufacturing
Systems 48, 157–169. https://doi.org/10.1016/j.jmsy.2018.01.006
Thames, L., Schaefer, D., 2016. Software-defined cloud manufacturing for industry 4.0. Procedia cirp,
52, 12-17.
Theorin, A., 2012. Service-oriented Process Control with Grafchart and the Devices Profile for Web
Services 6.
Theorin, A., Bengtsson, K., Provost, J., Lieder, M., Johnsson, C., Lundholm, T., Lennartson, B., 2017.
An event-driven manufacturing information system architecture for Industry 4.0. International
Journal of Production Research 55, 1297–1311.
https://doi.org/10.1080/00207543.2016.1201604
Wonderware System Platform - Fondation de l’Industrie du Futur [WWW Document], 2017. .
Wonderware. URL https://www.wonderware.fr/produit/supervision-et-controle/ihm-
supervision-et-controle/system-platform-2 (accessed 5.5.20).
Xu, L.D., Xu, E.L., Li, L., 2018a. Industry 4.0: state of the art and future trends. International Journal
of Production Research 56, 2941–2962. https://doi.org/10.1080/00207543.2018.1444806
Xu, L.D., Xu, E.L., Li, L., 2018b. Industry 4.0: state of the art and future trends. International Journal
of Production Research 56, 2941–2962. https://doi.org/10.1080/00207543.2018.1444806
Zdravković, M., Zdravković, J., Aubry, A., Moalla, N., Guedria, W., Sarraipa, J., 2018. Domain
framework for implementation of open IoT ecosystems. International Journal of Production
Research 56, 2552–2569. https://doi.org/10.1080/00207543.2017.1385870
Zhang, L.L., Rodrigues, B., 2009. A tree unification approach to constructing generic processes. IIE
Transactions 41, 916–929. https://doi.org/10.1080/07408170903026049
Zheng, L.Y., Dong, H.F., Vichare, P., Nassehi, A., Newman, S.T., 2008. Systematic modeling and
reusing of process knowledge for rapid process configuration. Robotics and Computer-
Integrated Manufacturing 24, 763–772. https://doi.org/10.1016/j.rcim.2008.03.006
Zhong, R.Y., Xu, X., Klotz, E., Newman, S.T., 2017. Intelligent Manufacturing in the Context of
Industry 4.0: A Review. Engineering 3, 616–630. https://doi.org/10.1016/J.ENG.2017.05.015
Zhou, K., Taigang Liu, Lifeng Zhou, 2015. Industry 4.0: Towards future industrial opportunities and
challenges, in: 2015 12th International Conference on Fuzzy Systems and Knowledge
Discovery (FSKD). Presented at the 2015 12th International Conference on Fuzzy Systems and
94
Knowledge Discovery (FSKD), IEEE, Zhangjiajie, China, pp. 2147–2152.
https://doi.org/10.1109/FSKD.2015.7382284
95
96
ÉCOLE DOCTORALE SCIENCES DES MÉTIERS DE L’INGÉNIEUR
[Laboratoire Conception de Produits et Innovation – Campus de Paris]
THÈSE
présentée par : Chawki EL ZANT
soutenue le : 16 Juillet 2021
et co-encadrée par :
[Mr BENFRIHA Khaled]
97
98
Table des matières
1. Introduction ........................................................................................103
99
3.3.2.3 Virtualization .......................................................................................................... 128
3.3.2.4 Modularity .............................................................................................................. 128
3.3.2.5 Service orientation ................................................................................................. 129
3.4 Industry 4.0 Axes of Research ............................................................................................... 129
100
8. Proposed transition framework toward industry 4.0 production systems
180
8.1 Existing Integration Frameworks ........................................................................................... 180
8.2 Advanced Manufacturing System, A Design principles Framework ...................................... 183
8.2.1 Flexible Shop Floor ................................................................................................... 184
8.2.2 Flexible Shop Floor Control System SFCS................................................................ 185
8.2.3 Integrated Data Management ..................................................................................... 186
8.2.4 Modular Service Integration ...................................................................................... 188
8.2.5 Openness for Business Level Communication and Distributed Industries ................. 189
101
9.6.1 Smart Visual Inspection System Module ................................................................... 219
9.6.2 Digital Twin Application Module .............................................................................. 221
9.7 Implemented System Ontology and Data Flows .................................................................... 223
9.8 Demonstrative Application ..................................................................................................... 225
9.8.1 Phase 1 ....................................................................................................................... 226
9.8.2 Phase 2 ....................................................................................................................... 227
9.8.3 Proposed measurement frame of reference for the quantitative assessment of the new
performance ........................................................................................................... 229
9.8.4 Analysis and discussion of results ............................................................................. 231
102
1. Introduction
Europe's industry is facing essential economic challenges due to an increasing pace of societal and
technological developments, such as decreasing availability of natural resources, increasing energy
prices, increasing the age of employees and globalization of markets. Moreover, consumers are
increasingly demanding for improved product-service innovation, product variety, quality standards,
support services, and immediacy or order satisfaction. These challenges drive industrial enterprises to
be capable of managing their whole value-chain in an agile and responsive manner.
Industrial production is nowadays driven by global competition and the need for fast adaptation of
production to the ever-changing market requests. These requirements can be met only by radical
advances in current manufacturing processes, as well as the integration of all actors in the company's
value chain (suppliers and customers). Technical aspects of these requirements are addressed by the
application of the genetic concepts of Cyber-Physical Systems CPS, industrial Internet of Things IoT,
and the integration of the new functionalities and technologies to the industrial production systems. The
understanding of industry 4.0 is the application of the internet of things to the industrial processes,
referring especially to the representation of physical objects such as machines, tools, workpieces, and
their equipment in the digital world, with the ability to communicate with other objects. These resulting
smart objects related to each other and together they build a system that allows for more flexible, more
efficient, and transparent planning control, and execution of production and logistics. To be more
specific: companies need both - virtual and physical structures that allow for close cooperation and rapid
adaption along the whole lifecycle from innovation to production and distribution (Erol and
Schumacher, 2016). This also requires changes along different layers of the company’s production
system from control systems to the information system at the planning and execution level as well as at
the business and management level.
Thus, all manufacturers need to ready themselves to embrace this potential industrial revolution to
remain competitive in the turbulent and hyper-competitive market. Technological innovations and
changes in business environments affect both firms' short-term performance and long-term
sustainability. When future directions and options in technology are obscure and uncertain, firms need
to formulate an appropriate technology strategy to support their planning for interacting with upcoming
future technological developments such as Industry 4.0. From both strategic and technologic
perspectives, the transitioning toward Industry 4.0 requires a comprehensive strategic model that
visualizes every further step on the route toward an entirely digital manufacturing enterprise.
Contemporary firms use technology road mapping extensively as a framework for supporting research
and development of future technologies that could sustain a competitive advantage. Road mapping is an
important method that has become integral to creating and delivering strategy and innovation in many
103
organizations. It is, therefore, obvious that an accurate technological and strategic model is indispensable
for securing success in the digital transformation process needed by Industry 4.0. Most research on
Industry 4.0 is focused on the Reference Architecture Model for Industry 4.0 RAMI 4.0 framework. The
conversation is around how to implement Industry 4.0 business models and how these will increase
productivity (Santos et al., 2017).
A major problem occurred, when a company decides to shift from 3.0 to 4.0 functionalities, and the
question is how to create this transformation in terms of integration level, tools, and strategy. Well,
trying to meet new technologies and upgrade the existing production lines with respect to the industry’s
actual structure must be deeply studied. An accurate integration process/remodelling Framework needs
to be developed carefully. Furthermore, the initial point is to identify the integration concept and
boundary.
The present PhD thesis aims to offer an integrative model that can be used as a steppingstone tool by
academicians and industrials toward the development of a detailed strategic and technological model
for the successful transformation from traditional manufacturing to digital manufacturing systems in the
context of Industry 4.0. It is obvious that having a deep understanding of the particularities of Industry
4.0 is a prerequisite for the development of the strategic and technological roadmap. Our scientific
approach consists in developing this conceptual model and applying it to a traditional production system
in order to identify the required changes and to define a framework for the digital transition accordingly.
The proposed Framework will be applied to the 4.0 industrial platform of Arts et Métiers of Paris in
order to experiment and validate our scientific approach.
The agility offered by Industry 4.0 allows plants to handle a more diverse product range with shortening
lead time, and quickly customize products to specific requirements, assuring higher responsiveness to
customers’ needs. While Industry 4.0 offers enormous, new opportunities for businesses; the scale of its
impact also brings new challenges. The dynamic issue of the traditional vertical structure of the
production is to be investigated since production systems need to be capable of adapting to new products
and responding quickly to fluctuations in demand first. Second, production systems must allow
deploying data collected from machines, and sensors in order to learn and self-optimize its processes.
This PhD thesis is organised into eleven chapters. Following the introduction, chapter 2 introduces the
industrial context of Industry 4.0 regarding the production system evolution through the four-industrial
revolutions. Chapter 3 reviews Industry 4.0 definition, benefits, design principles and functionalities,
and explores the potential of Industry 4.0 technology trends. Chapter 4 identifies the research positioning
and the main objectives of this project. Chapter 5 addresses the challenges facing the integration process
of Industry 4.0 into the manufacturing systems based on a state of art study. In chapter 6 the hypothesis
104
of this PhD is proposed, before presenting the model in chapter 7. Chapter 8 discusses the integration
framework proposed assuring mainly the vertical integration and allowing the horizontal one into the
company’s structure. In chapter 9, we experiment a novel integration architecture toward the digital
manufacturing transition developed and explained based on the literature review findings in order to
help companies toward the digital transformation. Chapter 10 will discuss results and findings, while
chapter 11 concludes the PhD thesis.
105
2. General Context of the 4th Industrial Revolution
Over the course of history, mankind has perfectly improved its industry by not only relying on technical
evolution but also by reinventing new resources which have created new technical means. Therefore,
the industry has benefited from qualitative advancements that have sometimes been so ingrained in a
certain time period and have had such an overwhelming impact that we have dubbed them revolutions"
(Derbel et al., 2019). The industrial needs and requirements of evolution drive the industrial revolution.
This is to say that the need for business evolution has conducted to the 4th industrial revolution along
with the emergence of new technologies and functionalities.
While the recent industrial revolution came up with the discovery of a novel technology or a new
source of power, Industry 4.0 offers a combination of technologies and functionalities (El Zant et al.,
2021) to be implemented into the whole value chain of an enterprise. Industry 4.0 is not offering a
mechanized steam machine to be added to the production line or a functionality aiming to optimize
the production time, it is a full transformation proposal, interconnected industry, a smart factory, and
a whole new approach to enhance businesses.
The objective of this chapter is to establish the link between the industrial needs vis-à-vis the digital
transition of companies and the objective of our scientific research in order to specify the general
positioning of this PhD thesis. The general context upon industrial revolutions is introduced in section
2.1. Industry 4.0 is introduced in section 2.2 as well as the economic and industrial aspects that
accompany the emergence of Industry 4.0. The aim of section 2.2 is to discuss industrial needs in the
context of the digitization of companies and production systems. Section 2.3 presents the research need
in the context of the 4th industrial revolution while positioning among the research axes of the Product
Design and Innovation Laboratory (LCPI), and by introducing the research project of the platform 4.0.
Section 2.4 identifies the objective of the PhD thesis among the defined research axes.
Explaining the Industrial Revolution, the global context for this epochal economic transformation lies
in a very substantial increase in human numbers from about 375 million people in 1400 to about 1 billion
in the early nineteenth century. Accompanying this growth in population was an emerging energy crisis,
most pronounced in Western Europe, China, and Japan, as wood and charcoal, the major industrial fuels,
became scarcer and their prices rose. In short, global energy demands began to push against the existing
local and regional ecological limits. In broad terms, the Industrial Revolution marks a human response
to that dilemma as fossil fuels replaced the earlier reliance on wind, water, wood, and the muscle power
of people and animals. All of those had derived from recently captured solar energy, but now human
ingenuity found the means to tap the anciently stored solar energy of coal, oil, and natural gas as well
(Strayer and Nelson, 2009).
106
Hartwell (Hartwell, 2017) listed in his book several causes of the industrial revolution, as the increase
in the capital rate and the world trade, the technological revolution, and the rise of the rational ethic and
culture which liberalized the context of enterprise.
Following a slow period of proto industrialization, the first industrial revolution spans from the end of
the 18th century to the beginning of the 19th century. It witnessed the emergence of mechanization, a
process that replaced agriculture with industry as the foundation of the society economic structure.
This transition included going from hand production to machines, improved efficiency of waterpower,
new industrial manufacturing, and iron production processes.
The Birth of the industrial Revolution in Great Britain provided the legal and cultural foundations that
enabled entrepreneurs to pioneer the industrial revolution. The absence of the trade barriers between
England and Scotland, the coal availability all over Britain especially in Wales, and free market
economy were key factors enabling the rise of that new industrial era. In its earliest form, the
manufacturing in XVIII century was usually carried out by a single skilled artisan with assistants.
Before the Industrial Revolution, most manufacturing occurred in rural areas to serve the agricultural
activities. In XVII century, entrepreneurs organized several manufacturing households into a single
enterprise through the putting out system. The most important techno-developments were Textiles -
Mechanized cotton spinning powered by steam or water that increased the output of a worker, Steam
power - The efficiency of steam engines increased greatly, the adaption of stationary steam engines
made them suitable for industrial use, Iron making - The substitution of coke for charcoal, lowered the
fuel cost of iron and wrought iron production. Mass extraction of coal along with the invention of the
steam engine created a new type of energy that thrusted forward all processes thanks to the
development of railroads and the acceleration of economic, human and material exchanges. Other
major inventions such as forging and new know-how in metal shaping gradually drew up the blueprints
for the first factories and cities as we know them today.
Mass production played a lead role in the second Industrial Revolution, inspired through many
inventions. Right about 1890’s many more inventions were created making this era unique. The newer
inventions advanced the speed and quality of thread spinning overriding most other machinery and
products. A novel strategy came to the industrial field, one of the best production ways was the
assembly line. Workers would stand at fixed stations and a pulley system would bring the meat to each
worker and they would complete one task. Henry Ford and others have written about the influence of
this American slaughterhouse practice on the later developments at Ford Motor Company. In the
assembly line, one person would do a specific task, while others are doing different ones, this way
107
made making cars much easier than usual.
As a result, the development of the combustion engine set out to use these new resources to their full
potential. Furthermore, the steel industry began to develop and grow alongside the exponential
demands for steel. Chemical synthesis also developed to bring us synthetic fabric, dyes, and fertilizer.
Methods of communication were also revolutionized with the invention of the telegraph and the
telephone and so were transportation methods with the emergence of the automobile and the plane at
the beginning of the 20th century. All these inventions were made possible by centralizing research
and capital structured around an economic and industrial model based on new “large factories” and
the organizational models of production as envisioned by Taylor and Ford.
In the second half of the 20th century, a third industrial revolution appeared with the emergence of a
new type of energy whose potential surpassed its predecessors: nuclear energy. This revolution
witnessed the rise of electronics with the transistor and micro-processor but also the rise of
telecommunications and computers along with the internet and the renewable energies. This new
technology led to the production of miniaturized material which would open doors, most notably to
space research and biotechnology. For industry, this revolution gave rise to the era of high-level
automation in production thanks to the information technology, the programmable logic controllers
PLCs and the robots.
Trade began to be easier with new innovations, companies started to produce in foreign countries. This
system now is called” offshoring” and it is used to reduce labour costs and raw materials. A new
process started in this century, which is now called” globalization”. By this process the world is
becoming interconnected as a result of increased trade and cultural exchange. It has increased the
production of goods and services and the biggest companies now are called multinational corporations.
To sum up, the key invention in Europe’s industrial revolution was the steam engine, taking advantage
of the energy potential of coal. Later industrial revolutions also used electric and internal combustion
motors (developed by the 1870s) and petroleum as well as coal. Before the industrial revolution,
almost all production in manufacturing and agriculture relied on equipment powered by people or draft
animals, with some assistance from waterwheels. Except for waterwheels, used mainly to mill grain,
almost all tools were designed for manual use. The industrial revolution progressively introduced
steam or other power to the production process and steadily increased the proportion of the process
accomplished by equipment without direct human guidance” (STEARNS, 2020). Some notable
inventions of the industrial revolution include:
Atmospheric steam engine - invented by Thomas Newcomen in 1712, Cotton gin - invented by Eli
Whitney in 1793, Telegraph - invented by Samuel F.B. Morse in 1836, Sewing machines - invented
108
by Elias Howe in 1844, Telephone - patented by Alexan- der Graham Bell in 1876, Phonograph -
invented by Thomas Edison in 1877, Induction electric motor - invented by Nikola Tesla in 1888,
Roller Coaster - patented by Edwin Prescott in 1898, Airplane - invented by Orvile and Wilbur Wright
in 1903, Model T Ford and the large-scale moving assembly line - invented by Henry Ford in 1908
and 1913.
Industrial revolutions have systematically been accompanied with an unquiet increase in productivity.
the rise in productivity has consequently enhanced wages and, the standard quality of life of people.
The evolution of Industry 4.0 is progressive; however, its impact can have an unquiet impact on
company performance, productivity and overall markets and industries (Santos et al., 2017).
The economical context and challenges of Industry 4.0 are presented in section 2.2.1. In section 2.2.2,
Industry 4.0 is introduced as a unique concept among other industrial revolutions. Finally, section
2.2.3 presents the origin of Industry 4.0.
Three main economical and industrial challenges are driving the emergence of Industry 4.0. Decreasing
availability of natural resources, increasing energy prices, increasing the age of employees and the
globalization of markets are key challenges in the economical context. Alongside, customer demands
are getting more and more challenging from improved product-service innovation to product variety
with high quality standards and reinforced support services as well as immediacy or order satisfaction.
Stages within the development of industrial manufacturing systems from manual work towards
Industry 4.0 concept is given as a path through the four industrial revolutions. The first industrial
revolution began with the mechanization and mechanical power generation, broke out only in a certain
country, England. It brought the transition from manual work to the first manufacturing processes,
principally in textile business. The 1st industrial revolution is characterized by the dominance of the
textile industry and the use of coal. The symbols are the steam engine and trains. The second industrial
109
revolution was triggered by electrification that enabled industry and production. The 2nd industrial
revolution is marked by a wide range of sectors entirely new as the chemical industry, mechanical
engineering, food. If the energy in the 1st industrial revolution was coal, in the 2nd industrial
revolution were the electricity, oil and gas. In the 2nd industrial revolution era, binomial fundamental
science and industry also arises. The results of science flow together in the industries especially in the
chemical industry that made her mistress in the pharmaceutical and food industries. The symbol of the
2nd industrial revolution is steel and Eiffel Tour construction. Typically mentioned in this context a
quote of Henry Ford, saying about the Ford T-Model car “You can have any colour as long as it is
black”. The quote introduces well the mass production but without the possibility of products
customization. The third industrial revolution is characterized by the automation with introduction of
microelectronics (Rojko, 2017). Industry 3.0 evolved from the second generation of industrial
revolution. It is the Information Technology (IT) era. Integrated Management Systems IMS were
introduced to reduce considerably the management costs. Industry 1.0 and 2.0 enable the production
as a service in industry 3.0. The whole company structure has been affected and reorganized into more
specialized and separated functions defining their core of business such as Research and Development
department, Finance, and Marketing (Santos et al., 2017).
Nowadays, we are living the fourth industrial revolution, called Industry 4.0. Industry 4.0 differs from
previous industrial revolutions in terms of concept and actuators. It is the first industrial revolution to
deviate from the energy-greed trend in terms of non-renewable resources since we have been
integrating more and more possibilities to power our production processes with alternative resources.
Tomorrow, factories 4.0 will be embedded in smart cities and powered by wind, sun, and geothermal
energy (SenTryo, 2017).
From the technological side, industry 4.0 is not about discovering a novel mechanical, electrical, or
even communication phenomenal, but a true combination between many novel technologies. Thanks
to the wide use of internet of things (IoT), the introduction of Cyber Physical Systems (CPS), Cloud
Manufacturing (CM), Additive Manufacturing, and many other technologies and actuators, a new level
of Industry and manufacturing could be reached, forming together the core technologies of the Industry
4.0. This new industrial revolution is mainly characterized by the connectivity and the interoperability,
the flexibility, the intelligence through the implementation of smart industries and smart products, the
prediction, the modularity, and the mass product customization. In another word, industry 4.0 helps
companies to produce increasingly individualized products with a short lead-time to market while
enhancing quality requirements. Intelligent manufacturing plays an important role in Industry 4.0.
Typical resources are converted into intelligent objects so they can act, sense, and behave within a
smart environment (Zhong et al., 2017).
Industry 4.0 is the first industrial revolution rooted in a new technological phenomenon - digitalization
110
- rather than in the emergence of a new type of energy. This digitalization enables us to build a new
virtual world from which we can steer the physical world. Today’s industries aim to connect all
production means to enable their interaction in real time. Factories 4.0 make communication among
the different actors and connected objects in a production line possible thanks to technologies such as
Cloud, Big Data Analytics, and the Industrial Internet of Things (SenTryo, 2017).
Traditional industry is currently undergoing a digital transformation, and industrial processes are
increasingly merging with modern information technologies. Industry 4.0 was developed as a picture
of the future that stretches out for 20 years or more. It is based on many available and ongoing
development technologies as the American William Gibson once wrote “I’d suggest that the future
has already arrived, it’s just not very evenly distributed yet”. The digital enterprise of the future will
be based on 4 key pillars: data backbone consistent across the organization, data connectivity through
networks, data mining, and finally data security. Between 20 and 50% savings might be achieved at
the top end of high agility. A lot could be saved in the reconfiguration reuse of the production
equipment or indeed dynamic changes in the product (Schlapfer et al., 2015).
The continuous progress of science and technology support the development of industrialization all
around the world, taking the manufacturing sector to an advanced stage (Belvedere et al., 2013). In
recent years, the Internet of Things and Cyber- Physical Systems have received an increased attention
from researcher, governments and industries, acting seriously to benefit from what this new industrial
revolution could provide (Liao et al., 2017).
Industry 4.0 is a strategic initiative of the German government, which was first announced at Hannover
fair 2011. Germany has one of the most potential economy and manufacturing sector presence
worldwide, especially in machinery and automotive sector (Rojko, 2017). This announcement
presented a strategic start point of the German ambitions in the manufacturing development. In 2012,
the German government passed the ‘High-Tech Strategy 2020’ action plan, investing billions of euros
for the development of cutting-edge technologies (Liao et al., 2017). In 2011, series of national level
discussions called ‘Advanced Manufacturing Partnership (AMP)’ were launched on behalf of the
United States government, aiming to prepare its industries for the next manufacturing generation
(Science and Technology (U.S.), 2014). In France, the concept ‘Industrie du futur’ was introduced and
initiated as a strategic review in 2013. It is based on the cooperation of industry and science and
defined as France’s industrial policy priorities. The United Kingdom government proposed a long-
term strategy in 2013 until 2050, called the ‘Future of Manufacturing’, in order to reinforce and support
the UK manufacturing in the 4.0 era (Foresight, 2013). The ‘Innovation in Manufacturing 3.0’ was
launched in 2014 by the South Korean government (Kang et al., 2016), the ‘Factories of the Future’
111
(FoF) launched in 2014 by the European Commission is funded by 80 billion euros for a 7-year plan.
The Chinese government issued the ‘Made in China 2025’ strategy alongside the ‘Internet Plus’ plan
in 2015, accelerating the industrialization in many manufacturing field (Kennedy, 2015).
2.3 Research objectives in the context of Industry 4.0 and the digital
transition of companies
Based on the industrial context introduced below, a critical need for a guide allowing the company to
adopt a successful digital transition towards the digitization of manufacturing systems is identified.
Taking into account the existing production system, the major question that arises is not anymore more
why, but how?
This research project subscribes in the field of industrial engineering and is particularly interested in the
digitization of manufacturing systems, based on a scientific approach.
In the revolutionary context of Industry 4.0, (Benfriha, 2020) proposed a research initiative which
considers 2 main axes:
Axis 1 involves the development of a generic approach to the digital transition of production systems.
The main objective of this axis is to help companies in their digital transition based on a gradual
approach (Benfriha, 2020). The digital transition of the company requires an evolution of the production
system and the integration of new technological pillars offered by Industry 4.0. This approach takes into
consideration the desired functionalities and the level of integration and evolution targeted by the
company, which defines the scope of this digital transition in terms of technological investment.
Axis 2 addresses the design optimization phase in the context of Industry 4.0, based on generated data
from digital production processes in order to optimize the design attributes of certain components.
Digitized manufacturing contributes to a critical generation of data and information from different
production processes and therefore different phases of the product life cycle. Data processing provides
useful information for the optimization of various production processes of a product in terms of costs,
energy consumption and time (lead-time), which brings considerable improvements of design by
creating a digital continuity of the product.
The integration of 4.0 concepts and technologies into manufacturing systems helps to create this digital
continuity of the product through its different life cycle phases. The digital continuity of the product is
a research axis of the Product Design and Innovation Laboratory (LCPI). The research activities of the
112
Product Design and Innovation Laboratory (LCPI, EA 3927) of the HESAM, Paris, fall essentially in
the field of industrial engineering as shown in Figure 1. Consequently, our research focused on the
integration of 4.0 concepts and technologies into manufacturing systems, in order to develop a scientific
decision-making support process for the transition to the digitalization of manufacturing systems.
Fig. 1 Laboratoire Conception de Produits et Innovation LCPI research fields and activities
Continued advances in science and technology are supporting the development of industrialization all
over the world, bringing the manufacturing sector to an advanced stage (Belvedere et al., 2013). In recent
years, the IoT Internet of Things and CPS cyber-physical systems have received increased attention
from researchers, governments, and industries, taking serious action to benefit from what this new
industrial revolution could offer (Liao et al., 2017). Along with government actions, many research and
development studies point in the same direction. Universities, research centres and companies are
increasingly interested in discovering industry 4.0 initiatives, contributing to the implementation and
integration of these new technologies. The goal behind all these efforts is to harness the potential of new
technologies and concepts, in particular the use of IoT, the integration of technical and business
processes, virtualization, flexibility, connected factory and smart product (Rojko, 2017).
113
In this context, the 4.0 industrial platform was developed to integrate the CPI laboratory scopes within
the Arts et Métiers school of Paris, in partnership with the Apprenticeship Training Center (CFA)
Ingénieurs 2000. This platform allows the development of experimental research in design,
maintenance, LCC, and production. This platform makes it possible to experiment models and
architectures of complex production systems. The second objective is educational. Indeed, on the basis
of the research work carried out, it becomes possible to offer innovative teaching and simulation courses
for learning the principles of Industry 4.0.
Among the two-axis identified in section 2.3.1, this thesis is the first thesis carried out at the platform
4.0 by considering the 1st axis: Digital transition towards 4.0 production systems. The objective of this
axis is to develop a scientific process of gradual transition to the digitization of production, thus
offering new functionalities that break with traditional production models. It consists of proposing a
general approach adapted to the company’s needs, taking into account the desired performances, the
expected technological deficit and the desired investment of the digital integration.
From new functionalities identified and targeted by the company, in terms of operational flexibility,
integrated quality, decision autonomy, predictive maintenance, optimization of energy consumption
and the desired level of portability, the idea is to propose a new production system model which defines
and integrates a distributed system architecture, cyber-physical systems (CPS), and connected 4.0
technological layers. The originality of this research lies in the consideration of all the layers of a
production system including the technological level but also the IT architecture and the control layer.
Obviously, the technological aspect is not the only factor of success, other factors such as the
management, the competencies, ... contribute to the outcome of such a project which marks a break in
industrial practices. However, only the technological aspect will be discussed in this thesis.
In what follows, chapter 3 is reserved to a general review concerning Industry 4.0 definition, benefits,
design principles and functionalities, and explores the potential of Industry 4.0 technology trends,
while chapter 4 identifies the research problematic in the perimeter of the digital transition of
manufacturing systems in the context of Industry 4.0, based on an analysis of the literature review.
114
3. Industry 4.0 Reviewed
Traditional industry is currently undergoing a digital transformation, and industrial processes are
increasingly merging with modern information technologies. There are many different descriptions
and characterizations of industry 4.0. In this chapter, the definition of Industry 4.0 is discussed, what
changes and benefits could the integration of Industry 4.0 offers, Industry 4.0 tools and functionalities
are introduced, and finally, Industry 4.0 main axes of research are identified from the literature.
According to (Dalenogare et al., 2018), Industry 4.0 can be seen as a matter of technology diffusion
and adoption. The Industry 4.0 concepts are proposed to enable companies to have flexible
manufacturing processes and to analyse large amounts of data in real time, improving strategic and
operational decision-making. (Duckworth, 2019) claims that Industry 4.0 is about a constellation of
technologies that perform optimally when interconnected and intelligently coordinated across the
115
parties and across the cyber-physical divide. It is this communication and collaboration that drive
much of the business value that Industry 4.0 offers. An industrial revolution is a fundamental shift in
the way industry works that forever changes the way manufacturing is done. However, besieged by
bright shiny objects of Industry 4.0, manufacturers should focus on long term goals instead of being
swept up by the abundance of new technologies, such as cloud, big data,…
(Radziwon et al., 2014) defines the smart factory as a manufacturing solution in a world of increasing
complexity, where flexible and adaptive production processes can solve dynamically production
issues. This solution is provided through the implementation of advanced automation saving exceeded
labour and waste of resource, and the collaboration between different industrial and nonindustrial
partners, forming a dynamic organization. (Zhou et al., 2015) considers innovative aspects of Industry
4.0 as a move from resource consumption, pollutant emissions and more extensive manufacturing to
green manufacturing changes, and development from production manufacturing to service-oriented
manufacturing. (Erol and Schumacher, 2016) proposed four main areas of actions to realize the novel
approaches to business and manufacturing operations, vertical integration of all layers of a production
system, horizontal integration of all partners in a value chain, lifecycle-engineering across the entire
value chain, and acceleration through exponential technologies. The digital enterprise of the future
will be based on four key pillars: data backbone consistent across the organization, data connectivity
through networks, data mining, and finally data security.
To sum up, Industry 4.0 is based on the integration of the business and manufacturing processes,
alongside with all actors in the company’s value chain including suppliers and customers. Three main
aspects are required in order to well achieve a successful integration and to apply potentially the
promised functionalities of Industry 4.0. Starting from the application of Cyber-Physical Systems CPS
and Industrial Internet of Things IoT to the production systems, by implementing CPS blocks with
decentralized control through an advanced network. Secondly, adapting the Manufacturing Execution
Systems MES and Enterprise Resource Planning ERP to integrate the manufacturing and business
processes adding higher levels of flexibility in control and connectivity is required. The third important
aspect is handling of a big amount of data collected from the processes, machines and products.
Typically, the data is stored in a cloud storage. This data requires extensive analytics that lead from
the ‘raw’ data to the useful information and, finally to the concrete actions that support an adaptive
and continuously self-optimizing industrial production process.
According to (Bodrow, 2017), the economic survival of each company is dependent on its willingness
to change. Industry 4.0 applies substantial changes in business affecting small, medium, and big sized
enterprises, specifically in three areas.
116
Changes in production
The production process changes are expected due to the quality improvement of products, and
production means, when assigned to information and data management methods. These changes will
be visualized through enabling digital value stream image and the real-time control, process flexibility
and smart assistance systems, online supervision and energy consumption, quality assurance
improvement. Based on these changes new perspectives for business will appear or strengthen their
position on the market.
Changes in logistic
The essential component of Internet of things is their autonomy. The same idea is fundamental in
respect to logistic in Industry 4.0. Within the vision of Internet of things planning, controlling,
connecting tasks are mainly delegated to employees. Otherwise, trucks transport the goods
autonomously; cars and forklifts organize themselves based on particle swarm approach; containers
organize their cargo and exchange with others in the logistic networks.
According to the McKinsey and Company, Industry 4.0 is the integration of the manufacturing sector
with the IoT framework using the latest sensors, integrating production systems with the back-end
supply chain and with the front-end market demand functions using analytic powers, and further
reducing contractual risks by using blockchain technologies and the latest manufacturing techniques,
eliminating human intervention all together (Santos et al., 2017). Management structures are pretended
to be more decentralized and flatter, albeit more specialized. Newer organizational functions such as
expertise in data analytics are prominently getting updated into management.
Regarding IT support, software tools are crucial for operating of the Industry 4.0 smart factory. Figure
2 depicts the well-known pyramid structure of support software of modern production systems. On the
business level, the Enterprise Resource Planning ERP tool is implemented. ERP supports enterprise-
wide planning such as business planning, supply chain management, sales and distribution,
accounting, human resource management and similar. Usually commercially available solutions are
implemented. In traditional ERP tools, the decision process is centralized on the highest level in the
automation pyramid. Most of the available ERP solutions do not support fast adaptation in production
planning due to the unplanned events. The second level in the traditional automation pyramid is
117
Manufacturing Execution System MES. It supports production reporting, scheduling, dispatching,
product tracking, maintenance operations, performance analysis, workforce tracking, resource
allocation and similar. It covers aspects such as management of the shop floor and communication
with the enterprise (business) systems. Most of the software solutions available on the market are
centralized and not distributed to the shop floor elements. This is a major limiting factor when
flexibility is needed due to the dynamics of customers’ order flow and/or changing production
environment, including shop floor configuration. The next operative level is process level control
based on Supervisory Control and Data Acquisition SCADA control system architecture followed by
controllers on machine/device level such as Programmable Logic Controllers PLCs, robot controllers
and other controllers. The last level of the automation pyramid is a machine/device level. In opposition
to the top two layers, this level has a naturally distributed control level. ERP and MES tools represent
basic software in the company and are used since the nineties. Both systems have typically a modular
structure but are centralized in their operation and thus have limited capability for dynamic adaptation
of the production plan.
Nevertheless, already implemented conventional ERP and MES systems should not be seen as main
obstacles to the introduction of the Industry 4.0 concept but more as a step towards it. Namely already
the introduction of a common MES tool requires advanced IT infrastructure on the shop floor level and
this is also a precondition for further development towards smart factory. The next important issue is
information integration among ERP, MES and other software tools used in the company such as, for
example, Customer-Relationship-Management (for help at managing relationships with the outside) and
118
Business Intelligence (for business analysis purposes). The problems such as database integration and
communication protocols need to be resolved. It can be concluded that for the Industry 4.0 the classical
automation structure does not present the best solution as it is not flexible enough for adapting to the
dynamic changes in the order flow and at the shop floor. Distributed MES solution, where most of the
functions are decentralized, is expected to be more suitable for the reconfigurable production systems,
Figure 3. For full support of reconfigurable systems, a continuous flow of information (vertical and
horizontal integration) between all elements should be realized (Rojko, 2017).
Figure 3 Industry 4.0 structure of IT support and operative level control (Rojko, 2017)
According to (Schlapfer et al., 2015), between 20 and 50% savings might be achieved at the top end of
high agility. A lot could be saved in the reconfiguration reuse of the production equipment or indeed
dynamic changes in the product.
(Santos et al., 2017) listed the main benefits of Industry 4.0 in four principal objectives, developing
green and sustainable industrial processes and supply chains, creating new strategic advantage in the
market, reducing costs using autonomous machines, and ensuring faster product development.
(Rojko, 2017) mentioned that Industry 4.0 could result in decrease by 10 to 30% of production and
logistics costs, and 10 to 20% of quality management costs. Furthermore, many advantages are
acquired:
- A shorter time-to-market for the new products.
- An improved customer responsiveness.
119
- Enabling a custom mass production without significantly increasing overall production costs.
- More flexible and friendlier working environment.
- More efficient use of natural resources and energy.
Based on the concept of Industry 4.0 and its full pack approach, this fourth industrial revolution has
unique design principles. Regarding our field of research and the aim of proposing clear integration
strategy and models, it is necessary to well define Industry 4.0 technologies and design principles and
functionalities. The successful integration does not rely only on the implementation of advanced
technologies but on the real improvement of the production process and the company performance.
Industry 4.0 is a complex combination of innovative functionalities such as Intelligence, Prediction,
Interoperability, and Modularity, assured and well implemented to the manufacturing systems through
the integration of a bunch of different new technologies and techniques offered by the 4.0 as CPS,
Advanced Robotics, IoT, and AI, in order to enhance the production performance, assure the security
means of workers and machines, improve product quality, and develop smart products (El Zant et al.,
2021).
The review identifies various Industry 4.0 technologies that are used to assist sustainable manufacturing
decisions. The survey identifies big data technologies that include virtual reality, robotic systems, cloud
computing, internet of things, simulation and prototyping, and 3D printing as the emerging technologies
of Industry4.0. Advanced industrial technologies aim to connect machines, tools, devices, components,
products and logistics ensuing real-time communications between themselves in such a way the system
develops the potential to generate and feed information, adding value to the manufacturing process.
3.3.1.1 3D printing
The 3D printing is a novel manufacturing concept. It differs from the conventional subtractive
manufacturing using machining techniques such as cutting, drilling or grinding based on the removal
method (Kamble et al., 2018). 3D printing or Additive Manufacturing technology makes it possible to
turn 3D modelled designs into custom solid objects. It is achieved using an additive process, where
successive layers of material are laid down, avoiding the need for parts and components assembly.
Generated model from any Computer-Aided Design CAD software can be used to the launch the 3D
printing process. The individual customer can now get exactly what he wants at a low price. Initially
used only in industries, this new technique is now available to a large number of people, and it will
soon change the global economy. We will not need to transport goods all over the world because they
will be printed anywhere at any time. According to (Horst et al., 2018), Freedom of design, mass
120
customization, waste minimization and the ability to fabricate complex structures are the main benefits
of additive manufacturing.
3.3.1.2 Augmented Reality AR and Virtual Reality VR
Augmented Reality AR and Virtual Reality VR are advanced vision technologies that differ in their
level of immersion. The first of these, AR, allows reality to be augmented through the addition of
virtual information to the real physical vision. The second, VR, creates a new appearance of reality
giving the user the sensation of actually being present within it. Within AR and VR, there are different
categories depending on the segmentation criterion used. For example, by type of hardware-software
components, type of device or electronic support. The most frequently used in the manufacturing
systems is the AR technology, and it is considered as one of the 9 pillars technologies of industry 4.0,
from the point of view of direct application to the industrial environment.
Augmented Reality AR technology is a human-computer interaction that integrates the natural visual
perception of a human with computer-generated information (3D models, simulations annotation, and
text). A communication, design and training enabling technology, and an immersive environment
allows non domain experts to provide valuable input. It drives as well to the simplification of
maintenance tasks and assembly procedures aiming to reduce the operating time and training costs.
Another function of AR is the visualization of real time data and information along with registered
links for documentation, and quality control (Ghobakhloo, 2018). The integration of AR in the
manufacturing processes where needed is an added value, which enhances the manufacturing
processes as well as the product, reducing the production cost and time and improving quality.
(Mekni and Lemieux, 2014) mentioned the on growing interest of educational researchers in AR. AR
is supposed to play an important role in the educational field, where it can potentially explain
complicated processes and parts, providing a simple and an interacting way to teach and present things.
It enriches the practical educational part as well with wide creative openings.
AR implementation is a true challenge, many aspects should be taken into consideration to ensure the
implementation success and the usability of the system. Technology context and organizational fit are
key success factors for the AR implementation.
3.3.1.3 Cyber security and Blockchain
The need for security in the industrial field is growing extremely due to the increased connectivity of
all the value chain components. Standard communication protocols are susceptible for cyber-attacks
and security breaches. Lately, cyber-attack stories are becoming countless, especially in stealing and
exploiting personal data, health records and financial privacies. Unfortunately, cyber criminals are not
running after data and intellectual property only, they simply can have negative intention in affecting
business and disrupt operations. Industry 4.0 is a networked environment where machines, robots,
121
sensors and actuators are connected through the internet or amongst each other (Ghobakhloo, 2018).
Therefore, it has an attractive ground for cyber issues.
The security issues of the IoT based CPS in the manufacturing systems in form of cyber-attacks,
viruses or malicious software, and data thefts have the potential to span from machine to another
crossing the communication systems. As (Kamble et al., 2018) highly sensitive industrial data could
be affected causing modifications or damages to the manufacturing processes and leading to quality
defects or even a complete shutdown. According to (Yu et al., 2017) CPSs are liable for cyber-attacks.
Where production files and process parameters could be affected, the damage may reach the
operational safety and the product quality.
The need for better security has led the development of advanced solutions that can help protecting
businesses, processes, and data. (Yu et al., 2017) used object-oriented Petri nets OPNs to create a
formal model of Cyber Physical Manufacturing System CPMS more secured against cyber issues and
malicious software. The blockchain is one of the solutions. Blockchain is a Distributed Ledger
Technology (DLT) mainly used and associated to the cryptocurrency transactions like Bitcoins.
Blockchain technology is very useful in modern cyber-defense systems. It can be used in industrial
applications and any type of digitized transfer of information and not limited to the financial field
(Ghobakhloo, 2018). It was built with the prevention of fraudulent activities in mind. It takes
advantage of intelligent consensus mechanisms that instantly detect data tampering and drop such
attempts. The integration of blockchain to the manufacturing system will enhance the trust and the
security through the whole supply chain, enabling autonomous and secure operations between CPSs,
machine and any supply chain partner of the smart factory.
Since information about products, business strategies, and companies are highly critical, Cyber
Security is considered as an essential pillar of industry 4.0. Cybersecurity and Blockchain must be
taken into consideration and placed on the top of the digital transformation process, since they can
lead to a safer environment for all type of data and information exchange.
3.3.1.4 Cloud computing
Thanks to the advancements in hardware, virtualization techniques, distributed computing, and the
wide services offered by the internet, cloud computing is henceforth reachable. Cloud computing or
Cloud Manufacturing CM is one of the most valuable and interesting technologies promised by
industry 4.0. The integration of Cloud Computing to the manufacturing process enhances the
interconnectivity and the interoperability between physical things, data, services, and people, main
function of the smart factory (Kamble et al., 2018). The cloud computing applied in the manufacturing
field refers to use computing services over the internet including data storage and processing. Several
options could be offered by the cloud computing technology, such as web-based management,
122
software application, data mining and processing applications, and distributed manufacturing
resources (Ghobakhloo, 2018).
Cloud Manufacturing is a research trend nowadays, focusing on three main directions: data sharing
architecture and platform design, resource allocation and management, and the creation of new
business models and CM applications (Esmaeilian et al., 2016). Five main characteristics are
recommended by the National Institute of Standards and Technology NIST in order to build an ideal
cloud: on-demand self-service, broad network access, resource pooling, rapid elasticity, and measured
service. The cloud model is delivered to the customers in three service dimensions, software, platform
and infrastructure (Zhong et al., 2017).
In the context of industry 4.0, the cloud technologies are strongly useful for enhancing the advanced
manufacturing systems. CM has many benefits, starting from increasing the system performance and
flexibility by increasing the data sharing across the company boundaries (Kamble et al., 2018). In fact,
a cloud computing system can execute all the applications performed with a normal computer and
more efficiently. Everything from word processing software to customized business programs
designed and developed for an organization can potentially perform on a cloud system (Zhou et al.,
2015). Furthermore, the use of CM is significant in reducing costs through bringing online systems.
Companies are no longer obliged to invest in licensing new software, building new infrastructure and
training personnel (Zhou et al., 2015), which increases the small enterprises potential to invest in other
development directions. Besides, with the Cloud Manufacturing enabled there in no need for adapting
data storage optimization, since storage spaces issues will be resolved. From another side, it is worth
to mention that the cloud computing has many challenges to overcome, especially the privacy and
security subjects, migration to cloud and compatibility, data management and resource allocation.
To sum up, Cloud computing has a great value in the realization of the smart factory, enabling greater
flexibility, cost reduction, and optimal resource availability.
3.3.1.5 Robotic systems
Robotic systems are playing an interesting role in enabling Industry 4.0 implementation. Modern
robots are characterized as systems offering autonomy, flexibility, cooperation, and reconfigurability.
Robots nowadays are working safely in the production environment, interacting with one another and
with human workers and even learning from them (Kamble et al., 2018). The integration of these new
robotic systems into the manufacturing processes offers an increased flexibility and reconfigurability,
especially when doing multi-tasks and interchanging with machines and parts and using integrated
camera systems with advanced image processing facilities. The use of robotics helps protecting
humans from handling dangerous tasks or areas, along with complex missions. In Industry 4.0 era, the
integration of robots in a smart network has the great added value. First, advanced robotic systems are
able now to communicate directly to SCADA systems bypassing the Programmable Logic Controllers
123
systems PLCs which enhances the flexibility of the production process and therefore the mass
customization aspect. Adding to this, another layer of flexibility and intelligence is offered through
the integration of machine learning techniques to the image processing stage, and the interconnection
between robots and machines which extend the decentralization aspect. It is worth to mention the
significant advancement and evolution of the robotics simulation software, especially that it can be
used to create digital twins of the robots by collecting the real time data from robots and predicting
possible collusions in a complex environment. From another side, industrial robotics and automation
promise numerous benefits such as reduced part cycle time, lower defect rate, higher quality and
reliability, reduced waste and better floor space utilization, making it indispensable to world-class
manufacturers (Ghobakhloo, 2018).
3.3.1.6 Simulations and prototype
Simulations and prototyping techniques are key factors in creating real-time data to mirror the physical
world in a virtual model, including machines, products, and humans. This allows operators to test and
optimize the machine settings for the next product in line in the virtual world before the physical
changeover, thereby driving down machine setup times and increasing quality. Simulation and
modelling not only enable manufacturers to prevent errors at an early stage that might otherwise result
in substantial costs for plant operators, but they can be used to optimize a manufacturing plant during
ongoing daily operation (Ghobakhloo, 2018).
The simulation technique was used to study cyberattacks using the bifurcation control method with an
objective of generating the expected dynamic behaviours and guaranteeing the trustworthiness of
cyber-physical manufacturing systems at run time. Simulation studies were also used for observing
the behaviour of machines: movements, machining operations, connectivity with robotic arms, real-
time tracking of manufacturing cost, energy efficiency of wireless sensor networks, deadlock
prevention, and the use of virtual engineering objects for industrial design and manufacturing. Use of
digital twins (prototypes) to simulate the process behaviour of the actual machines interactively is a
potential growing area. Digital twins are based on the Ethernet/IP communication protocol and provide
information like movements, machining operations, and connectivity with robotic arms (Kamble et
al., 2018).
3.3.1.7 The Internet of Things IoT
The IoT is the pervasive access to entities on the Internet for the extension of the physical world using
a variety of sensing, detection, identification, location tracking, and monitoring equipment. In general,
IoT is able to offer advanced connectivity of physical objects, systems, and services, enabling object-
to-object communication and data sharing. The IoT is now envisioned as a larger convergence of
cutting-edge technologies such as ubiquitous wireless standards, data analytics, and machine learning.
RFID technology provides one such example. It has been reported that nearly 20.8 billion devices will
124
be connected and making full use of RFID by 2020 (Zhong et al., 2017).
(Ghobakhloo, 2018) discussed the lack of a universal application protocol of Industrial Internet of
Things IIoT, which prevents the integration of machinery from various manufacturers, and therefore
proposed Semantic Technologies. The use of semantic technologies facilitates the information
exchange between different actor of Industry 4.0 in the large scale, providing a common standard
communication through semantic Web of Things WoT and enhancing the communication between
heterogenous partner of Industry 4.0 in the large scale.
3.3.1.8 Horizontal and vertical transformation
The defined goal of Industry 4.0 is the implementation and utilization of individual interdependency
matrix fitted to every single business process. In this respect, horizontal and vertical integration of IT-
systems are required. The horizontal integration covers the product life cycle from supplier until it
reaches the customer. It involves the design and development of product, production planning,
production rump-up and management, logistics, sales and distributions (de Man and Strandhagen,
2017). On the other side, vertical integration is concerned with all hierarchy levels in the enterprise.
This dimension comprises the enterprise planning (regarding production planning), enterprise
management (regarding detailed planning, production data acquisition, and quality management),
process management (regarding monitoring and management of production), and operation (regarding
the interface to production process using sensors and actors) (Bodrow, 2017).
Industry 4.0 concept will be in most companies realized by using already available equipment and
technologies. Only when a new production system is planned, there is an opportunity to design the
production system already from the beginning as Industry system. Therefore, one of the challenges is
how already available standards will be integrated into the new concept (Rojko, 2017). The concept
of Horizontal and vertical integration is proposed by the German Plattform Industrie 4.0 which
introduced the Reference Architecture Model for Industry 4.0 RAMI4.0 as a reference framework for
digitalization Figure 4 This is a meta-model so it describes the aspects that play an important role in
the Industry 4.0 production system. It is based on the internationally accepted Smart Grids architecture
model introduced in year 2014, however with two additional bottom layers to address specific aspects
of Industry 4.0. The three-dimensional RAMI4.0 should enable:
The first dimension of the RAMI4.0 addresses two elements, type and instance. As long as an idea, a
125
concept, or a product is still a plan and is not available/realized yet, it is called type. The second
dimension of the model deals with location, functional hierarchy from the product to the connected
world (as the last stage of Industry 4.0 development with all enterprises, customers and suppliers
connected). The third dimension of the RAMI4.0 model is organized in functional layers as follows:
- An assets layer includes physical components such as robots, conveyer belts, PLCs,
documents, archives, but also non-physical objects such as software and ideas.
- An integration layer provides information for assets in a form that can be digitally processed.
It includes elements connected to IT such as sensors, integration to HMI and computer-aided
control of technical processes.
- An information layer is processing and integrating available data into useful information.
- A functional layer includes formal descriptions of functions. Also, ERP functions belong to
this layer.
- A business layer includes mapping of the business model and links between different business
processes.
126
3.3.1.9 Big Data
The advances in the internet technology, internet of things, cloud computing, big data, and artificial
intelligence have profoundly impacted manufacturing. The volume of data collected in manufacturing
is growing. Big data offers a tremendous opportunity in the transformation of today’s manufacturing
paradigm to smart manufacturing. Big data empowers companies to adopt data-driven strategies to
become more competitive.
Research and studies about Industry 4.0 design principles are numerous, which makes this area rich
while the findings are pretty similar. The main design principles of Industry 4.0 as identified in the
state of the art and agreed by researchers are: Interoperability, Decentralization, Virtualization,
Modularity, and Service orientation (Carvalho et al., 2018; Ghobakhloo, 2018; Hermann et al., 2016;
Kamble et al., 2018; Xu et al., 2018a).
3.3.2.1 Interoperability
The interoperability is one of the offered functionalities of Industry 4.0. It is the consequence of
digitizing and enabling the interconnection between different actors of the company. The concept of
interoperability is based on the ability of all participants such as human, resources, and machines to
communicate and interact through the horizontal and the vertical integration required for Industry 4.0
(Ghobakhloo, 2018; Xu et al., 2018a). This network authorizes the operational, managerial, and
technical interoperability, based on the system flexibility and strength of analysing data and its
exploitation, improving consequently the collaboration between departments and different value chain
actors, and offering a great flexibility in a wide meaning.
3.3.2.2 Decentralization
Decentralization is a whole new concept of Industry 4.0, based on the decentralized control network.
Instead of depending on the centralized computing system of decision-making body, the
decentralization offers the ability of companies, operations staff, and even the machines to make
decisions (Carvalho et al., 2018; Kamble et al., 2018). It enables different actor of the smart factory to
127
work independently and make decisions autonomously, thanks to the Self-regulating systems and
intelligent control mechanisms such as CPS, and IoT.
The basic principle of Industry 4.0 is that by connecting machines, workflows, and systems, businesses
implement intelligent networks along the entire value chain to control one another autonomously
(Sung, 2018). Decentralization helps to achieve quick decision making and offers more flexibility
decentralization thanks to the simplified planning and coordination of different processes.
Decentralization acts as well as a perfect organizational setup to handle the growing needs of highly
customized products (Carvalho et al., 2018; Kamble et al., 2018). An example of the decentralization
is the Autonomous Guided Vehicles AGV in the smart warehouses and their efficiency of
synchronization which significantly reduce the complexity of central planning by providing the
freedom of decision making. Another example is the predictive maintenance, when a machine can
predict failure and trigger maintenance processes autonomously or self-organized logistics that react
to unexpected or unusual changes in operation. This example would show the rationalization of the
driving force behind the Internet of Things.
3.3.2.3 Virtualization
Virtualization means creating a virtual copy of the physical world. Virtualization enables the
development of a “digital twin” of the entire value chain (smart warehouse, smart factory, all related
equipment and machinery, and even smart products) by merging sensor data acquired from the
physical world into virtual or simulation-based models. The virtual twin of the smart factory would
enable process engineers and designers to enhance existing processes and optimize the efficiency of
production lines in complete isolation, without disrupting the physical processes in the smart factory
they have virtualized. Alternatively, the digital twin of a smart product would enable manufacturers
to have a complete digital footprint of their existing or new products all throughout their lifecycle,
from design and development to the end of the product. This not only would enable a better
understanding of the performance of the product at the consumption stage, but also allows companies
to virtually evaluate the system that builds the product. Virtualization is heavily dependent upon the
real-time capability. In general, Industry 4.0 is centred on cumulative, real-time, real-world data across
an array of dimensions such as smart warehouse, smart factory, smart product and smart business
partners, meaning that real-time capability is deeply supported by internet of everything. Real-time
capability is not just about collecting data, as it involves real-time data analysis, real-time decision
making according to the new findings, and even real-time cyber-security attack detection
(Ghobakhloo, 2018).
3.3.2.4 Modularity
Modularity is another design principle of Industry 4.0, referring to the modular production systems
that can be adapted by replacing and expanding the individual modules in a much more comfortable
way. Modularity is concerned with shifting from linear manufacturing and planning, rigid systems and
128
inflexible production models toward an agile system that can adapt to an ever-changing circumstances
and requirements. Modularity involves the entire production and manufacturing levels, and builds on
agile supply chain, flexible material flow systems, modular decision-making procedures and flexible
processes. Modularity is complemented by product personalization, which is another design principle
of Industry 4.0. Product personalization is indeed a more customer-oriented implication of mass
customization. The introduction of modern technology trends such as responsive CPS, IoT, open
product architecture, automation and additive manufacturing has enabled product reconfiguration
based on the continually changing customer preferences, mostly identified via assessment and
prediction of consumers’ behaviour. This means manufacturers not only should meet customers’
existing demands and preferences, but also benefit from Internet of People IoP simulation and big data
analytics to forecast upcoming market trends and customers’ needs (Ghobakhloo, 2018).
3.3.2.5 Service orientation
Service orientation in the context of Industry 4.0 mostly refers to the concepts of Manufacturing as a
Service MaaS and Production as a Service PaaS. MaaS business model refers to the collective use of
a networked manufacturing infrastructure to produce goods. Interconnectivity between manufacturers
and the widespread of IoT and cloud computing has offered new manufacturing ecosystems by
allowing companies to communicate their manufacturing needs and capacities automatically. In this
environment, complex manufacturing tasks can be accomplished collaboratively by several
manufacturing services from different companies. This means, instead of the physical product, the
production capacity of manufacturers can be regarded as the primary good. In the PaaS business
model, products are delivered as a service or virtualized experience, and instead of a single upfront
payment, customers subscribe to the product and pay a recurring fee on a perpetual per-outcome basis.
This business model is particularly enabled by Internet of Service IoS technologies that can be built
into products to monitor when and how they are used (Ghobakhloo, 2018).
The process integration brings in the sense of service orientations all the entities in the production
system are interconnected facilitating the creation of the product-service system. The flexibility and
agility achieved as an outcome of the service orientation enable the organizations to respond to market
changes more quickly. This allows the various stakeholders of the organizations to come together and
partner to co-create value for the consumers (Kamble et al., 2018).
In the realm of manufacturing, the advances of science and technology continuously support the
development of industrialization all around the world. In recent years, the Internet of Things IoT and
Cyber-Physical Systems CPS have received an increased attention from researcher, governments and
industries, acting seriously to benefit from what this new industrial revolution could provide (Liao et
al., 2017).
129
(Yang et al., 2018) identified in their review the international research trends related to the smart
factory: IoT or ICT, technology innovation and environmental conservation, operations management,
and manufacturing and logistics research. This review listed the main research topics related to smart
factory, relying between developing smart factory frameworks using IoT, developing methods to
measure the smart factory performance and energy usage, maximizing productivity and efficiency in
smart factories, studying current additive manufacturing systems, minimizing the operational and
resource cost, and improving equipment efficiency, reliability, and product quality of manufacturing
systems.
(Zhong et al., 2017) reviewed 165 recent papers and posted the future research trends related to
different key technology in industry 4.0, starting from IoT, to CPS, cloud computing, big data
analytics, and Information and Communication Technologies ICT. From a conceptual perspective, the
main areas of research will focus on the generic framework for intelligent manufacturing, data-driven
intelligent manufacturing models, human-machine collaboration, and the intelligent manufacturing
applications.
In (Xu et al., 2018a), a systematic research perspective is presented based on a deep review. Several
research trends were proposed: the integration of CPS including designing interfaces, and integrating
heterogeneous components, and tools, verification and testing CPS developing uniform standards and
specifications, Blockchain technology implementing blockchain into industry 4.0 manufacturing
process, Smart devices creating smart objects using artificial intelligence and IoT, resilient smart factory
developing a resilient industrial ecosystem with high reliability, Enterprise Resource Planning ERP
evolution due to the rise of CPS and IoT in industry 4.0, future ERP systems must evolve to address
needs and new challenges, and industry 4.0 career impacts.
(Ghobakhloo, 2018) focused on identifying industry 4.0 design principles and technology trends. A
sum of 14 technology were found in this review along with 9 design principles:
Technology trends - IoT, Internet of Services IoS, Internet of People IoP, Internet of Data IoD, Cloud
computing, Big data, Blockchain, Augmented reality AR, industrial robotics, Cybersecurity, Additive
manufacturing, Simulation and modelling techniques, CPS, and Semantic technologies.
Design principles - Service orientation, Smart product, Smart factory, Interoperability, Modularity,
Decentralization, Virtualization, System integration, and Corporate social responsibility.
A growing interest about industry 4.0 is pointed out by (Kamble et al., 2018) through the number of
papers published in relation to the topic between 2012 and 2017. A significant increasing in the
published research papers is observed in the year 2016 to 2017. Based on the type of research, the
main research direction is still conceptual discussing the concept of industry 4.0 and its topics,
followed by case studies papers, while other research direction such as simulation, experimentation,
130
survey and prototypes are minorities (Kamble et al., 2018). From another side, the corresponding
research categories were identified as following: high rated interest in industry 4.0 technologies, shop
floor equipment interactions, human-machine interfacing, and sustainability.
(Liao et al., 2017) recognized three essential integration aspects necessary for the transition toward
industry 4.0, along with eight priority areas of actions. First, the horizontal integration which covers
the integration of the IT systems to communicate various manufacturing and business planning
processes including logistics, production, marketing . . . and between several companies. the vertical
integration enables the integration of needed IT systems at the different hierarchical level such as
actuator, and sensor level, manufacturing and execution level, production management level, and
corporate planning levels) delivering an end-to-end solution. The End-to-End digital integration
addresses the engineering process so that the digital and real worlds are integrated across a product’s
entire value chain and across different companies, whilst also incorporating customer requirements.
This integration process requires several technical, managerial and organizational actions, such as
Standardization and Reference Architecture, delivering a Comprehensive Broadband Infrastructure,
Safety and Security, Work Organization and Design, Training and Continuing Professional
Development, Regulatory Framework, and Resource Productivity and Efficiency.
Though, according to (Liao et al., 2017), the realization of industry 4.0 requires more research efforts
especially in the direction of proposing the detailed roadmap to guide this transformation, a Regulatory
Framework, and End-to-End Digital Integration. Beside this, it is worth to mention that Industry 4.0
experiments done in laboratories are far from the industrial applications, in terms of implementation.
Moreover, and along with the existing research efforts related to the data analytics, some other area of
actions needs to be involved, such as the upgrade of existing technologies, like Programmable Logic
Controllers, Production Machinery, and Industrial Robots, to meet the needs of Industry 4.0. However,
many research efforts are undergoing in the context of Industry 4.0, such as Standardization and
Reference Architecture, Resource Productivity and Efficiency, Vertical and Horizontal Integration,
strategic management, decision-making, location tracking, reconfigurability and sustainability.
131
4. Scientific Context & Research Problematic
The objective of this chapter is to identify the research problematic in the perimeter of the digital
transition of manufacturing systems in the context of Industry 4.0, based on an analysis of the literature
review. The identification of aspects of Industry 4.0 is classified according to the different phases of the
life cycle of the product in order to well define this research positioning and orientation and therefore
specify our research problematic concerning the digitization of the manufacturing systems.
This chapter starts from discussing the business model evolution in the context of Industry 4.0
integration in section 4.1. Section 4.2 analyses the aspects of the evolution of manufacturing companies
towards the digitalization of industrial processes. Section 4.3 identifies the scope of the integration of
Industry 4.0 principles in the manufacturing companies. In section 4.4, a projection of the concepts of
Industry 4.0 to the product life cycle is introduced to clearly define the positioning of our problematic.
Section 4.5 presents the positioning of our research problematic and the research question of this PhD
thesis.
In the industrial context, the main industrial problematic occurs at the business and strategic level. In
the competitive and the ever-changing market, companies are driven by the need for digitalization and
flexibility. The aim of this chapter is to clarify the connection between the digitalization of the
manufacturing systems and the evolution of the business model of companies especially when
considering flexible and connected manufacturing systems. According to (Dalenogare et al., 2018), this
new industrial stage is affecting competition rules, the structure of industry and customers’ demands. It
is changing competition rules because companies business models are being reframed by the adoption
of IoT concepts and digitization of factories. From the market point of view, digital technologies allow
companies to offer new digital solutions for customers, such as internet-based services embedded in
products. From the operational perspective, digital technologies, such as CPS, are proposed to reduce
set-up times, labour and material costs and processing times, resulting in higher productivity of
production processes. With the service orientation of Industry 4.0 and horizontal integration, new
business models can be developed, with new ways to deliver and capture value from customers.
Three proposal are identified by (Ibarra et al., 2018) implicating Industry 4.0 in the business models’
components enabling the identification of different ways to transform business models. In the context
of Industry 4.0, service-oriented approach, network-oriented approach, and user-driven approach
contribute to the improvement of traditional Business Model due to an incremental innovation of both,
value creation and value delivery.
132
Based on this classification, (Ibarra et al., 2018) proposed a new Business Model typology based on the
smartization of products and services in order to study the implications of Industry 4.0 in Business
Model Innovation. Four ways to conduct the digital transformation in manufacturing companies has
been identified according to the innovation degree applied that goes from modifying such a few elements
of the business model through an incremental innovation, to the transformation of all the elements of
the business models due to a radical innovation:
A Business Model describes "the rationale of how an organization creates, delivers, and captures value”.
Each proposal is described thereby, showing the changes in the value creation (that takes into account
the key activities, resources and partnerships), the value delivery (which encompasses the product and
services offered, the distribution, communication and sales channels, the customer segments and the
established relationships) and the value capture (that describes the costs and revenues gained by the
firm).
Within the framework of our PhD thesis, we are interested by the internal and external process
optimization dimension proposed by (Ibarra et al., 2018) which concerns mainly the digital integration
into the manufacturing systems. This transformation represents an incremental innovation that optimizes
the actual business without involving big changes. New enabling technologies such as Big Data, Cloud
Computing, Collaborative Robots, Additive Manufacturing, Artificial Vision or Augmented Reality are
introduced just to optimize the value creation architecture (key resources and activities) due to increasing
efficiency and improving performance (reducing costs, time and failures, employee training, etc.). This
could be the first step for traditional manufacturing companies to embrace the Industry 4.0
without addressing high risks. The Table 1 below describes the impacts on each component of the
Business Model.
Table. 1. Changes in Business Model components to achieve the internal and external processes
optimization (Ibarra et al., 2018)
133
To sum up, the digitalisation of the manufacturing systems and the flexibility of the production processes
affect companies from several aspects:
The flexibility of the manufacturing systems enables the product diversity aspect through the
implemented flexibility into the production processes and assure the product customization functionality
regarding the clients demands. Moreover, the digitalisation and the flexibility integrated allow for an
advanced and continuous optimization aspect of production systems through the data deployment and
the ability to enhance manufacturing processes based on data mining. The third aspect concerns the
business level and the strong connection between the flexibility and the digitalisation of the
manufacturing systems with the evolution of the economic model of companies. The solid optimisation
ability aspect is crucial for companies willing to evolve their economic model, it strengthens the
evolution steps first and it guaranties their market stability against the evolution risks. Second, the
product diversity aspect is a part of the economical evolution itself when the company could enlarge the
variety of products but also this aspect is another key success of the economic model evolution since the
integrated flexibility assure important time and effort savings of the industrialization stage. The last
dimension affecting the economic model evolution indirectly is precisely the flexible production system
itself since it a prerequisite for companies aiming to evolve their economic model toward connected
products and service-oriented manufacturing, customer integration purposes and new ecosystems and
value networks. To conclude, the digitalized and flexible manufacturing system is the starting point of
the economic model evolution in the context of Industry 4.0.
Based on the above-presented literature review, and referring to the changes and benefits study as well
as the axes of research and the business model evolution identified previously, three main aspects could
be associated to the integration of Industry 4.0 into the companies as shown in Table. 2 below:
Table. 2 Aspects of Industry 4.0 Integration into the companies (Bodrow, 2017)
134
According to (Bodrow, 2017), changes in production is the application of Cyber-Physical Systems CPS
and industrial Internet of Things IoT, by implementing CPS blocks with decentralized control via an
advanced network. Changes in logistic is considered by Adapting MES and Enterprise Resource
Planning ERP manufacturing execution systems to integrate manufacturing and business processes by
adding greater levels of flexibility in terms of control and connectivity. Finally, the influence of big data
and cloud is processing and deploying big data collected from processes, machines and products using
edge computing and cloud computing.
The classification of these aspects proposed by (Bodrow, 2017) is supported by (Dalenogare et al.,
2018), considering three main advantages that characterize Industry from a technological integration
point of view: vertical integration, horizontal integration and end-to-end engineering. The vertical
integration refers to the integration of ICT systems in different hierarchical levels of an organization,
representing the integration between the production and the management levels in a factory, matching
and contributing to changes in production. On the other hand, the horizontal integration consists in the
collaboration between enterprises inside a supply chain, with resource and real time information
exchange, corresponding to changes in logistics. End-to-end engineering is the integration of
engineering in the whole value chain of a product, from its development until after-sales. The last
dimension End-to-End engineering is assured mainly by processing and deploying data and the influence
fact of big data and cloud integration into the manufacturing systems.
(Frank et al., 2019) claimed that the concept of Industry 4.0 represents a new industrial stage of the
manufacturing systems by integrating a set of emerging and convergent technologies that add value to
the whole product lifecycle. Referring to (Frank et al., 2019), Industry 4.0 technologies can be separated
into two different layers according to their main objective:
‘Front-end technologies’ of Industry 4.0 considers the transformation of the manufacturing activities
based on emerging technologies (Smart Manufacturing) and the way product are offered (Smart
Products). It also considers the way raw materials and product are delivered (Smart Supply Chain) and
the new ways workers perform their activities based on the support of emerging technologies (Smart
Working). We call ‘front-end technologies’ this technology layer because the four ‘smart’ dimensions
are concerned with operational and market needs. Therefore, they have an end-application purpose for
the companies’ value chain, as shown in the schematic arrow represented in Figure 5. It is worth noticing
that the central dimension of the front-end technology layer is the Smart Manufacturing, while the other
dimensions are interconnected to this one.
135
‘Base technologies’ which comprises technologies that provide connectivity and intelligence for front-
end technologies. This last layer is the one which enables the Industry 4.0 concept, differentiating this
concept from previous industrial stages. This is because base-technologies allow front-end technologies
to be connected in a complete integrated manufacturing system.
This new industrial stage demands a socio-technical evolution of the human role in production systems,
in which all working activities of the value chain will be performed with smart approaches (Smart
Working) and grounded in information and communication technologies (ICTs).
Industry 4.0 is rooted in the advanced manufacturing or also called Smart Manufacturing concept, which
means an adaptable system where flexible lines adjust automatically production processes for multiple
types of products and changing conditions. This allows to increase quality, productivity and flexibility
and can help to achieve customized products at a large scale and in a sustainable way with better resource
consumption.
Industry 4.0 also considers the exchange of information and integration of the supply chain (called Smart
Supply Chain), synchronizing production with suppliers to reduce delivery times and information
distortions that produce bullwhip effects. This integration also enables companies to combine resources
in collaborative manufacturing, allowing them to focus on their core competences and share capabilities
for product innovation in industry platforms, a joint effort to develop products and complementary assets
and services, with more value-added.
The technologies embedded in the final products (Smart Products) are also part of the broader Industry
4.0 concept. Smart products can provide data feedback for new product development as well as they can
provide new services and solutions to the customer. Thus, some scholars consider the smart products as
136
the second main objective of Industry 4.0, since they allow new business models such as the product-
service systems, which create new opportunities for manufacturers and service providers.
At the heart of the Industry 4.0 concept, smart manufacturing technologies form the central pillar of
internal operational activities, while smart products take into account the external added value of
products, when customer information and data is integrated into the production system. Smart
manufacturing considers technologies for manufacturing products (production system), while smart
products consider technologies related to product offering.
Therefore, we assume that smart manufacturing is the starting point and primary goal of Industry 4.0,
while smart product is its extension. This vision follows the recent chronological evolution of the
Industry 4.0 concept, which finds its roots first in advanced manufacturing systems and its connections
to other business processes. Regarding the Smart Manufacturing dimension, the main associated
technologies include six main objectives: vertical integration, virtualization, automation, traceability,
flexibility and energy management, and will be discussed the state-of-the-art chapter.
To sum up, Smart supply chain focuses on the connection of the manufacturing system with the supply
chain and therefore it is concerned by the horizontal integration toward the value chain, while Smart
working provides support for different needs of the Industry 4.0 production system and focused on
integrating the worker with the manufacturing system. Smart working environment has two dimensions.
First dimension considers supporting the working power toward the integration of new technologies and
helping them through maintenance and operation tasks. Seemingly, the second dimension of smart
working environment is strongly related to the smart manufacturing integration considering the
industrialization phase of new products and supporting engineers with operating the new manufacturing
system.
Based on the categorization proposed by (Frank et al., 2019), each dimension could be associated to a
lifecycle phase of the product from an engineering point of view. As presented in Figure 6, the
integration of Industry 4.0 aspects into the manufacturing companies are associated to the different
lifecycle phases of the product. The aim of this figure is to highlight our interest in the Manufacturing
dimension and therefore clarifying our research positioning focusing on the digitization of
manufacturing systems in the context of Industry 4. The Smart Product aspect concerns mainly the
design phase by adapting the Production-service-system PSS as design pattern and embedding sensors
and software to the product in order to integrate the concept of service orientation to the connected
products. The Smart Working Environment concerns basically the industrialization phase of the product
lifecycle, by integrating innovative support tools to the operators and assisting the production engineers
137
in the production and the process planning tasks. The Smart Production Systems aspect considers
directly the manufacturing phase, and is the main objective of this research which will be discussed in
the next section. The Smart Production System or the Flexible Manufacturing System is mainly
concerned with flexibility and vertical integration. Lastly, the Smart Supply Chain is associated to the
services related to the product across the whole value chain concerned mainly by the horizontal
integration.
In summary, our research focuses on the integration of 4.0 concepts and technologies into manufacturing
systems. Therefore, the main objective of our research is to develop a model as a decision-making
support tool to help companies aiming to drive toward the digital integration of manufacturing systems
and processes based on a scientific approach. The purpose with developing such model is to improve
the flexibility and performance of production systems and processes through digitization and 4.0
technologies implementation, and to enable the manufacturing of product diversity along with enhancing
the product quality.
Manufacturing companies can focus on different needs they may have when prioritizing the
implementation of the aforementioned “smart manufacturing” technologies. However, recent literature
has shown that the industry varies in terms of the expected benefits of these technologies and should
systematically consider implementing smart manufacturing functionalities through the deployment of
these technologies to achieve a level of maturity compatible with Industry 4.0 concepts.
This means that smart manufacturing technologies can be interdependent and create an active synergy
for Industry 4.0. The synergistic “Smart Manufacturing” integration supported by technological
integration as tools of 'digitization' potentially fits into the configuration and functional evolution of
138
'production processes', which has a direct effect on the evolution of the manufacturing system
architecture. New manufacturing systems are able to adjust their production capabilities and use
advanced features to meet customer demands on time. The flexibility of manufacturing systems applies
not only to a factory's hardware components such as robots and machines, but also to its information
and communication technology (ICT) infrastructure including networks, data transfer, storage and
processing.
The IoT results in CPS cyber-physical systems which integrate physical objects of the factory with the
virtual dimension of the factory, including integrated data, artificial intelligence, and simulation.
Therefore, manufacturing companies aiming for a higher maturity level of Industry 4.0 should
consistently adopt most smart manufacturing technologies, as these technologies are interdependent.
The adoption of this advanced technology in turn results in data generation, helping to improve and
optimize production processes once that data is processed and exploited.
This analysis leads us to define the positioning of our research problem ‘which model for a successful
digital transition of 4.0 manufacturing systems?’ This problematic lies between the three axes defined
by Figure 7, 'digitization', 'manufacturing system', and 'production processes'.
The evolution of the manufacturing system in the context of Industry 4.0 is considered the starting point
to answer the research question identified. This development must first take into consideration the
139
current state of the manufacturing system to subsequently establish its digital transition by providing it
with advanced functionalities. Digitization is one of the technological tools offered by Industry 4.0 to
apply the targeted functionalities on the production system. The functionalities and advanced
technologies integrated into the production system seem to overlap with the evolution and digitization
of the various production processes with the aim of improving their performance.
The definition of these three dimensions helps us adapting our research direction in order to respond to
the identified problematic. At this stage, our approach consists in exploring a broad state of the art study
in order to identify the fundamental concepts necessary for the construction of a 4.0 production system
model with the objective of guiding companies in the digital transition of their production systems.
These concepts of different but complementary natures should overlap, interface and interconnect in
order to create a coherent production system and detailed enough for pragmatic deployment. The
proposed model will be followed by a Framework for the digital transition of production systems as a
support tool aiming to guide the 4.0 digital integration into existing production systems.
140
5. Digitization of Manufacturing Systems in The Context of Industry
4.0 – State of the Art
Talking about successful integration models of Industry 4.0 holds doubts in the implementation
process and the use of new technologies. While the recent industrial revolution came with the
discovery of a novel technology or a new source of power, Industry 4.0 holds several aspects and a
combination of technologies to be implemented into the whole value chain of an enterprise. Industry
4.0 is not offering a mechanized steam machine to be added to the production line or a functionality
aiming to optimize the production time, it is a full transformation proposal, interconnected industry, a
smart factory, and whole new approach to enhance production processes performance.
In this chapter, we aim to clarify the main challenges following the implementation of Industry 4.0
from the literature point of view and from our perspective regarding the direction of our research.
Section 5.1 analyses general challenges regarding manufacturing systems evolution. The following
section 5.2 is devoted to the classification of issues related to the integration of technology 4.0 in
manufacturing systems that fit into the dimension of "vertical integration". This classification takes
into consideration the three areas identified in the previous chapter "Production System",
"Digitization" and "Production Processes" in order to respond to our research problematic. The main
concepts identified in section 5.2 specifically the flexibility, the modularity, information and
communication technologies (ICT), and the vertical integration are presented respectively in the
following sections. Then, a state-of-the-art analysis of recent architectures and models proposed in the
literature will be carried out in section 5.5. Finally, section 5.6 provides a summary of positioning
toward the state of the art.
Challenges regarding the concept of the fourth industrial revolution could be identified in two main
themes referring to (Erol and Schumacher, 2016), strategic and technical challenges. Respectively,
companies aiming to apply this transformation are in great need for support and guidance in finding a
proper strategy towards coping with the challenges imposed by Industry 4.0 due to its complexity and
the lack of existing roadmaps. (Erol and Schumacher, 2016) addressed the fact that Industry 4.0 is a
methodology more than being just a technology. As mentioned by the National Academy of Science and
Engineering in Germany (Kagermann et al., 2013), Industry 4.0 is a paradigm shift in business
operations, rather than a technology-based improvement of production capabilities. While companies
are trying to implement some technological innovations in order to attain short-term market advantages,
the methodological integration of Industry 4.0 into the whole value chain has a greater impact on the
company’s future business. Industry 4.0 applies substantial changes in businesses affecting small,
medium, and big-sized enterprises, specifically in three areas: changes in production, logistics, and data
deployment. (Bodrow, 2017) referred the issue of applying these changes to the information system and
141
the software tools when the classical automation structure does not present the best solution for industry
4.0 as it is not flexible enough for adapting to the dynamic changes in the order flow and at the shop
floor as at the business layer.
(Santos et al., 2017) negotiated the challenges of Industry 4.0 implementation by proposing different
political strategy and integration scenario, taking into consideration the market aspects. Based on cost-
benefit analysis, Successful implementation and absorption of Industry 4.0 is highly dependent on the
market width and depth of the company. The company has to decide whether it implements Industry
4.0 technologies internally (develop in house) by simply buying and installing them or engaging with
technologies providers and integrators. This decision has to be taken precisely regarding the overall
benefits of the company since outsourcing Industry 4.0 capabilities has a great impact on the internal
strategy, meanwhile it may decrease significantly the implementation costs. Industry 4.0 will increase
the enterprise performance if it is coupled with the new product development strategy. The challenge
of Industry 4.0 implementation is not only about the capital of investments. the transformation toward
Industry 4.0 affects many areas of interest of the companies. The engagement with the digital
transformation requires consequently developing strategies for exploring and exploiting new markets,
responding the mass production and customization. Otherwise, the production costs may increase
dramatically with the risks of underutilization of production systems.
From the technical implementation perspective, industries are still holding doubts about implementing
new technologies, because of unclear possible benefits, lack of clear implementation details, and the
seemingly large investments required (Theorin et al., 2017a). Several studies indicated that high
investments, the complexity and required know-how, as well as the unsuitability of existing IT-
Infrastructure and technologies which disturb the vertical and horizontal integration, are the main
restraints for realizing Industry 4.0. According to (Zhou et al., 2015), the implementation of Industry
4.0 faces many challenges in the way, from various areas such as scientific, technological, and
economic challenges. The main relative challenges are the development of smart factories, the
construction of the network environment, CPS modelling, and the integration of CPS which is an
enormous challenge for SME’s (Goerzig and Bauernhansl, 2018).
Well, the development of smart factory requires the implementation of smart devices into different
production stages, while the latter necessitate a long and deep work of development followed by the
testing process, beside its costs. The construction of the network environment in Industry 4.0 is another
challenged task. It requires the implementation and the integration of the CPS platform. To do so, first,
the cooperation between physical and computing systems is to be considered, in order to ensure the
real time communication and process control. Second, regarding the CPS modeling, the behavior
between the physical and computing processes needs to be well studied. The physical environment,
142
software and hardware platforms, and network models as well as any further scheduling software
required, network delays, power consumption and other functional and non-functional factors are to
be taken into consideration. At the end, the integration of CPS is to be examined carefully regarding
its complexity, along with developing a flexible interface which supports this heterogenous
combination of components. The big data analysis and processing is another layer of complexity and
development, which constitutes a serious challenge. Managing the large amount of data introduces the
information protection and privacy issues, as well as the real time integration and analysis. Finally,
the digital manufacturing involves all the value chain of an enterprise increasing the need for cyber
security(Zhou et al., 2015).
(Dalenogare et al., 2018) addressed the adoption of emerging technologies of Industry 4.0 from a socio-
technical perspective. They identified three socio-technical dimensions considering the digitization
process towards the Industry 4.0 implementation. First dimension considers the organization of work
since the implementation of new technologies requires a to rethink how the organization will operate.
The second dimension considers the human factors and the fact that new technologies require new
competences and skills from the workers. Thirdly, external environment is considered since the adoption
of new technologies are dependent of the maturity where they are implemented.
(Sung, 2018) identified the main challenges of Industry 4.0 based on the economic effects of the digital
transformation, including, IT security issues, Reliability and stability needed for critical machine-to-
machine communication M2M including very short and stable latency, Need to maintain the integrity
of production processes, Need to avoid any IT snags as those would cause expensive production
outages, Need to protect industrial knowhow (contained also in the control files for industrial
automation gear), Lack of ad- equate skill-sets to expedite the march toward the fourth industrial
revolution, Threat of redundancy of the corporate IT department, General reluctance to change by
stakeholders, Loss of many jobs to automatic processes and IT-controlled processes, especially for
lower educated parts of society.
(Xu et al., 2018b) considered four main challenges facing Industry 4.0 integration. Technical challenges
include the need to develop the existing Information and Communication Technology ICT
infrastructures, the scalability resulting from the network growth and the large volume of data, data
science and analytics, and IoT technical integration and connection. In addition to the technical issues,
(Xu et al., 2018b) addressed the need for standardization regarding the rapid growth of IoT, and finally
the information security and privacy protection.
The required extension of products by digital components to the point of cyber- physical systems and
their integration into service systems leads enterprises in mechanical engineering into a fundamental
change process – the digital transformation. Besides mechanics and electronics, they now need to
143
develop digital capabilities and implement them into fast-changing, cross-company processes and
structures. Especially for small and medium-sized enterprises SMEs this is an enormous challenge
(Goerzig and Bauernhansl, 2018).
To sum up, the challenges facing the successful integration of Industry 4.0 into the manufacturing
systems can be classified into two main dimensions, the business dimension, and the manufacturing
integration dimension. The first-dimension concerns company’s strategy and perspective for evolution
regarding its core business activities and its aim of evolution. Industry 4.0 and digitization are
dramatically driving companies to adapt to the saturated and globalized market, where producing in
quantity ‘Mass Production’ is no longer suitable to respond to the increasingly individualized demand
‘Mass Customization’. Faced with this problem, the responsiveness and adaptability of production
systems are major assets to adapt to fluctuations in demand (Erol and Schumacher, 2016). Therefore,
the evolution of the manufactured systems becomes necessary and imperative in order to scope with the
business challenges. Hence, technical challenges arise, several studies indicated that high investments,
the complexity and required know-how, as well as the unsuitability of existing IT-Infrastructure and
technologies which disturb the vertical and horizontal integration, are the main restraints for realizing
Industry 4.0. These challenges drive us far to negotiate the ability of the existing manufacturing systems
to be adapted into digitization when dealing with its architecture and IT infrastructure. The next section
is dedicated to classify the challenges regarding the integration of Industry 4.0 into the manufacturing
systems in order to conclude the main problematic of the research.
The detailed state-of-the-art findings regarding strategic and technical challenges led us to identify the
main concepts of the evolution of manufacturing systems in the context of Industry 4.0. These concepts
are about a superposition of principles and technological blocks related to different areas affected by the
digitization of production systems as shown in Figure 8.
144
Fig. 8. Main concepts of the digitization of production systems
Actually, the diversity of challenges and integration issues identified above, drive us to summarize these
overlayed challenges. Figure. 9 classified these overlayed challenges and assigned them to a central
topic, by considering the dimension of vertical integration as the main issue. First of all, ICT and the
production information system is an essential dimension identified in the literature, which considers
mainly the limitations of the existing ICT infrastructure and the required changes needed to fit the new
configuration of the digital manufacturing system (Xu et al., 2018b). Alongside, the information system
design is another challenge especially when coping with advanced technologies, implementing specific
services into the manufacturing system (Bodrow, 2017) and adapting cloud computing to enhance data
analysis (Kamble et al., 2018). Second, the modularity of the manufacturing systems is another key
challenge as identified by (Kaushal et al., 2016). The modularity concept includes several dimensions
such the CPS development and integration into the manufacturing system (Goerzig and Bauernhansl,
2018), and deploying modularity into the Shop Floor Control System SFCS (Mes and Gerrits, 2019).
145
Third, the Flexible Manufacturing System in the context of Industry 4.0 is a main challenge (Theorin et
al., 2017a), which comprises the advanced technologies implementation issue (Zhou et al., 2015),
flexible operations deployment at the shop floor (Erol and Schumacher, 2016), and connected machines
configuration matters. Lastly, these diverse key challenges could be all considered under the vertical
integration challenge (Dalenogare et al., 2018), which concerns in addition the IT infrastructure
dimension (Rojko, 2017), the IoT implementation (Kamble et al., 2018), and data Integration issues
(Rüßmann et al., 2015a).
Following this analysis of the concepts and challenges of the digitization of production systems, we
propose a state of the art dedicated to the flexibility and modularity concepts. In section 4 we review the
shop floor control system and the ICT structure, along with discussing the vertical integration in favor
of 4.0 functionalities, and analyses recent 4.0 architectures. The Data dimension is introduced in section
5.
Fig. 9. Framework of Industry 4.0 Integration Challenges into The Manufacturing Systems
In a saturated and globalized market, producing in quantity ‘Mass Production’ is no longer an adequate
response. It becomes necessary to adapt the manufactured products to the increasingly individualized
146
demand ‘Mass Customization’. Faced with this problem, the responsiveness and adaptability of
production systems is a major asset to respond to fluctuations in demand (Erol and Schumacher, 2016).
Several studies were conducted in the literature in the context of product customization focusing on
incorporating the configuration principle into process planning aiming to promote the process plan
generation. (Schierholt, 2001) identified the process configuration concept which is meant to simplify
the process planning generation for new product variant using principles known from the product
configuration concept. Two main concepts for process configuration systems were presented by
(Schierholt, 2001): interactive process configuration and automation-based process configuration.
(Zheng et al., 2008) proposed a systematic knowledge model for multiple variants of products and
applied a process configuration method for rapid process planning by applying configuration rules. A
tree unification approach is proposed by (Zhang and Rodrigues, 2009) to develop generic processes
from existing production data of product families helping companies to fulfil a diversity of customized
products and therefore reducing cost and lead time.
While, in the context of Industry 4.0, an enhanced process configuration can be enabled through the
flexibility and adaptability of the production systems. Traditional industries are currently undergoing a
digital transformation and industrial processes are increasingly merging with modern information
technologies. This new industrial revolution is mainly characterized on the one hand by connected and
interoperable production machines which are driven intelligent. And on the other hand, by new
functionalities such as flexibility, prediction, modularity, and mass product customization. In other
words, industry 4.0 contributes to enhance the process configuration in order to produce increasingly
individualized products with a short lead-time to market and conform to the required quality. A lot could
be saved in the reconfiguration reuse of the production equipment or indeed dynamic changes in the
product at the top end of high agility (Schlapfer et al., 2015).
Although, several research and studies are undergoing in the industrial firm aiming to introduce fast,
effective and cost-effective solutions in order to enhance the flexible manufacturing concepts in the
existing production lines. Based on the promised facilities of industry 4.0 technologies and its ability to
integrate the business processes and activities, including Big Data and Analytics, Autonomous Robotic
Systems, Cloud Computing, Industrial Internet of Things, Simulation and Prototyping, Additive
Manufacturing, Augmented Reality, Horizontal and Vertical Integration, and Cybersecurity (Rüßmann
et al., 2015a), the manufacturing system can increase the flexibility of the manufacturing processes
(Kamble et al., 2018).
From another side, (Mes and Gerrits, 2019) addressed the fact that hierarchical control systems are not
perfect for the need for a more diverse product mix. Therefore, more flexible and reconfigurable
manufacturing concepts have been introduced consisting of autonomous and intelligent controller
147
modules. These modules dynamically interact with each other to achieve local and global objectives.
Some of the industrials such as ABB (Bloch et al., 2018b) started developing what is called the Modular
Type Package MTP, an initiative to change the control process infrastructure in the shop floor, driving
the way of a small automation concept as defined by (Strategy&, 2015). According to (Mourtzis et al.,
2016), Computer-Aided Process Planning CAPP must consider highly customized product variants in a
dynamic manufacturing environment, rather than operating in isolation from shop floor production data
and suffering from the lack in interfacing with the IT systems. Therefore, they proposed an Internet-
based service-oriented system for adaptive process planning based on machine availability monitoring
assuring the real-time collaboration between process planning and execution.
In this context, the flexibility of production systems appears as a necessity to maintain the
competitiveness of industrial enterprises. This evolution is part of the transition to industry 4.0. The
purpose of this research is to present a new control architecture that runs in a modular mode to define
and reconfigure the production plan in a short time compared with the traditional approach. This novel
modular approach is meant to overcome limitations without changing the whole existing control process
but by moving the decision-making process to the Manufacturing Execution System (MES) layer to
enhance flexible production operations.
According to PwC Stratgy& Global Digital Operations Study (Geissbauer et al., 2018), four digital
ecosystem layers are the foundations of an industry 4.0 system, Fig. 10:
- Customer Solutions Ecosystem: in our case, the capacity to capture the demand, its variability and
the needs for customization will feed the modular production system.
- Operations Ecosystem: in our case, the heart of the system is a Flexible Manufacturing System with
a modular approach.
- Technology Ecosystem: in our case, MES, API, machines and sensors with connectivity are the key
technological elements.
- People Ecosystem: in our case, we take into account the ordinary skills of production people to
manage production avoiding complex programming of API.
The flexibility of the production machines and the modularity of the manufacturing operations are two
indissociable concepts to achieving new functionalities in terms of flexible production. This means that
machines will operate independently or in coordination with humans to manufacture customer-oriented
product (Sung, 2018). In order to achieve these goals in the industry, special attention should be paid to
several level of manufacturing, particularly machines (shop floor tools & equipment), control, and
manufacturing execution systems. Therefore, we present in the next section the concept of FMS and
modular process. The FMS is a requirement prior to the implementation of a modular process of
manufacturing operations.
148
Fig. 10 The four digital ecosystem layers (Geissbauer et al., 2018)
A Flexible Manufacturing System FMS is differentiated from a traditional production system by its
ability to perform multiple operations from a limited number of resources. The FMS thus becomes an
149
interesting concept to adapt the production to a constantly changing environment, since it makes it
possible to avoid long periods of downtime related to the reorganization of the processes. According to
(Kaschel C. and Y Bernal, 2006), Flexibility refers to the ability of a manufacturing system to respond
cost-effectively and rapidly to changing production needs and requirements. This capability is an
essential key to the design and operation of manufacturing systems to overcome the unpredictable
market movement. The scheduling flexibility has to be integrated, offering a reliable decision-making
responding to the clients’ order and the market evolution. By reducing the number of operations
necessary for its reconfiguration, an FMS improves performance in terms of costs and deadlines, the
FMS allows quality benefits. (Fragapane et al., 2020) claims that manufacturing flexibility enhances
productivity with a fast adaptation to the ever-changing client demands, thus, reducing costs and
resources. (Fragapane et al., 2020) investigated the decentralization of material flow which can provide
more flexibility for production systems using Autonomous Mobile Robots AMR in the Process
Industries PI for high mix production and addressed that new planning and control models are needed
for production networks and decision-support systems in order to control material flows in the era of
Industry 4.0.
According to (Qin et al., 2016), The reconfigurable manufacturing system is the closest to a smart
factory system, followed by the flexible manufacturing system. (Mabkhot et al., 2018) identified the
suggested requirements in terms of design principles of the smart factory, mentioning the ‘modular
machine tools’ referring to the flexibility of machines and workstations to be reconfigured in terms of
changing the shop floor layout and adjusting the process function, and ‘modular material handling’
equipment refers to the possibility of reconfiguring material handling equipment (i.e., conveyors,
Automated Guided Vehicles AGVs) on the shop floor or changing equipment capability to transfer the
required product. (Mourtzis et al., 2012) studied the flexibility regarding product and volume changes
in the design phase, considering alternative solutions such as conventional press, CNC and laser
machines for a system’s punching department producing commercial refrigerators. Based on investment
cost and return performance, (Mourtzis et al., 2012) proposed a methodology using Penalty of Change
POC method to evaluate the implementation of critical capital investment decisions in the manufacturing
technology. (Mourtzis et al., 2020) presented the design and development of a flexible manufacturing
cell as learning factory in the goal of minimizing the human intervention in the production processes
using a robotic arm mounted with an HD camera to perform the quality control task. (Mazars and
OpinionWay, 2018) summarized the advantages and disadvantages that arise from the operation of an
FMS. On the benefits side, we mention the rapid adaptation to market evolution both in quantity and
customization, reduction of stocks, better control of production by providing real-time information,
better use of equipment, reduction of direct labour (contribution of automation), and reduction of the
general expenses (less expectations, losses, ...). On the disadvantages side, it is relevant to take into
consideration the high cost of realization, the complexity / sophistication of the implementation, the
150
heavy incidence of uncertainties in certain cases (grouping of operations on the same means), and the
need for qualified personnel and specific skills in modern technologies.
Modularity is a principle of Industry 4.0 and one of its essential functionalities. Answering the client
demands, the modular production systems give the possibility to adapt and adjust modules of the
production process in a more comfortable and useful way depending on the production and the product
evolution. Modularity is concerned with shifting from linear manufacturing and planning, toward an
agile system that can adapt to ever-changing circumstances and requirements, without the need for a
huge and sophisticated redevelopment and reprogramming work. In other words, the modularity is a
potential solution enabling what is called product customization (Kumar, 2007), the key principle of
Industry 4.0. Modularity involves the entire production and manufacturing levels, and builds on agile
supply chain, flexible material flow systems, modular decision-making procedures, and flexible
processes (Ghobakhloo, 2018). According to (Weyer et al., 2015), the high product variability with
shortened product-life cycles for new products demands requisites an agile and flexible production
structure. They addressed Cyber Physical Systems CPS and IoT to be implemented along with the need
of coordinated standardization between technologies and automation for interoperability purposes in
order to assure modular factory structures and enable dynamic re-engineering processes.
In terms of investment decisions, modularity in FMS moves the trade-off between the other traditional
production systems (Kaushal et al., 2016). By switching easily in a high range of volumes or variety,
modular production systems can replace shop floors with stand-alone machines or dedicated lines and
improve the capacity of the production system for dealing with uncertainty, Fig. 11. More specifically,
A production system can be defined as a set of modules connected by flows, with the function of
transforming raw materials into a product. The overall flexibility of the production system stems
logically from the flexibility of the modules and flows that compose it.
Modularity is not only being able to change the layout of the shop floor simply but also have a flexible
structure that allows the extension of the module to increase production capacity or integrate new
functionalities (Mabkhot et al., 2018). A module is loosely coupled and can be moved, added, or
removed from the system in a plug-and-play manner. (Gorecky et al., 2016) described a module as a
puzzle block that can work alone or in combination with other modules to form a production system.
151
The planning phase of production operations has a crucial technical aspect which requires on the one
hand defining the production capacity of the machines, this parameter is classic and not much variable.
On the other hand, to know the way in which the production operations are structured, this parameter is
recent and has been introduced recently by various works published in 2017 for the automation of
modular process plants. (Bloch et al., 2017) introduced the concept of modularity of production
operations within the framework of a joint project between TDU and the ABB research centre on the
theme 'Information Models and Architecture for Modular Concepts in the Process Industries'.
Several research and development contributions has been published in the context of modular production
systems. (Weyer et al., 2015) presented the SmartFactoryKL, a modular production system and a novel
project for cross-vendor solutions, addressing standardization as the crucial challenge for highly
modular, multi-vendor production systems. (Bloch et al., 2018a) of the IAT institute published works in
2018 where he considers the problematic of modular operations as an approach that meets the increasing
demands for flexibility in the manufacturing industry. He also addresses the problem of conventional
control systems that do not properly support flexible production systems. According to (Perzylo et al.,
2019), instead of producing high quantities of similar products over a long period of time, companies
have to satisfy the market demand for customized or even individualized products. As a result, their
production lines may have a multitude of different variants, which may only be produced in small lot
sizes. (Bloch et al., 2018b) suggested the modularization of the process control system, using different
process modules. These process modules will provide encapsulated process functions as services to the
superior control system SCS. Within the SCS, services are orchestrated by the plant operator to achieve
the desired production process. (Fan et al., 2020) proposed Function Block FB based Closed-Loop
Adaptive Machining CLAM for adaptive process planning and execution. IEC 61499 FBs enabled High-
152
Level Controller HLC are designed to plan and execute the finishing process planning referring to initial
process planning and instantiation. Low-Level Controller LLC are implemented to manage low-level
tasks, such as, motion control, and data acquisition.
This section discusses the challenges facing the evolution of manufacturing systems in terms of ICT
infrastructure, starting from the limitations of the traditional manufacturing IT system structure,
addressing technical and control issues, and thereafter presenting several proposed 4.0 architectures
from the literature, followed by a brief discussion.
Information and communications technology ICT is an extensional term for information technology
IT that stresses the role of unified communications and the integration of telecommunications
(telephone lines and wireless signals) and computers, as well as necessary enterprise software,
middleware, storage, and audio-visual systems, that enable users to access, store, transmit, and
manipulate information (“Information and communications technology,” 2021).
Figure 12 shows the ICT roadmap proposed by many researchers (Chen, 2017; Kaynak, 2007;
Wilamowski, 2005), as well as its evolution through different industrial revolution stages reaching the
industrial information integration for the emerging era of IoT and CPS (Xu et al., 2018a).
The advances in manufacturing technologies in the middle of twentieth century were behind the rise
of the ICT, especially after introducing automation and microelectronic technology into
manufacturing. In the Third Industrial Revolution, the advancement of ICT was at the core of every
major shift of the manufacturing paradigm. The widespread adoption of computer numerical control
CNC and industrial robots made flexible manufacturing systems FMS possible; the technologies for
computer-aided design CAD, computer-aided manufacturing CAM and computer-aided processing
planning CAPP made computer integrated manufacturing CIM possible (Feng et al., 2001).
153
Fig. 12 The ICT evolution from Industry 1.0 to Industry 4.0 (Xu et al., 2018a)
Industry 4.0 is mainly represented by CPS, IoT and cloud computing (Henning et al., 2013; Hermann
et al., 2016; Jasperneite, 2012; Lasi et al., 2014; Moeuf, 2018). IoT is the key enabler of the advanced
manufacturing systems such as the cloud manufacturing similar to cloud computing, and a main
support to the smart factories development. Cloud-based manufacturing can contribute significantly
to the realization of Industry (Thames and Schaefer, 2016). CPS is the core foundation of Industry 4.0.
CPS enables a higher level of integration and coordination between physical and computational
elements (Gürdür et al., 2016). Research indicates that with the introduction of CPS, machines will
be able to communicate with each other with decentralized control systems and therefore optimize the
production. Industry 4.0 generally comprises many complex components, and has broad applications
in numerous industrial sectors (Xu et al., 2018a).
Traditional ICT structure in conventional production systems is presented by the Computer Integrated
Manufacturing CIM pyramid Fig. 13. At the enterprise management level, orders are treated along with
resources and long-term planning in the Enterprise Resource Planning ERP system, supporting business
planning, supply chain management, sales and distribution, accounting, and human resource
management. In traditional ERP tools, the decision process is centralized on the highest level of the
automation pyramid. Most of the available ERP solutions do not support fast adaptation in production
planning due to unplanned events (Rojko, 2017).
154
Fig. 13 CIM Pyramid
At the second level of the traditional automation pyramid, the Manufacturing Execution System receives
the orders as the executive manager and in his turn forwards the commands to the API / Controllers
layer to run the respective production program of a certain product thus operating production lines and
machines. MES has several functionalities including production reporting, scheduling, dispatching,
product tracking, historization, maintenance operations, performance analysis, and it covers aspects such
as management of the shop floor and communication with the enterprise (business) systems (Santos et
al., 2017). This traditional hierarchical ICT structure suffers from many limitations when it comes to the
flexibility, changing production environment, dynamics of customers’ order flow and mass-
customization purpose of Industry 4.0, since many existing MES solutions are not distributed to the shop
floor elements and have a centralized architecture (Rojko, 2017).
The next operative level of the CIM pyramid is process level control, including Supervisory Control and
Data Acquisition SCADA control system architecture and controllers such as Programmable Logic
Controllers PLCs, robot controllers and micro-controllers. As a matter of fact, in the case of fully
automated production, the flexibility is one of the biggest challenges at controller level. Indeed, once
the process is frozen in the Application Programming Interface API layer, only reprogramming the
controllers allows the workshop to be used for another task than the active / configured one. For each
new product, it will be necessary to create and redevelop a new program, allowing to produce another
product rather than the old one. This situation represents the conventional industrial architecture type
3.0. (Karnouskos et al., 2010) claimed that communication and networking for data exchange between
devices and services including sensors, devices, PLC controllers, MES, and ERP need to be adapted for
a dynamic production within Industry 4.0. According to (Theorin, 2012), manufacturing companies
must reconfigure production plants in a short time frame, and therefore flexible and adaptable
automation devices - able to be easy to integrate, configure, extend, and reuse - are required. Where
control tasks are becoming more complex, control systems suffer from the lack of the well-defined
155
information flow between process planning and implementation due to the low implementation level
and vendor-specific programming methods (Ollinger et al., 2014) and thus necessitate hard
commissioning and reengineering of control programs (Theorin, 2012). Therefore, (Lin et al., 2019)
addressed the importance of improving manufacturing control systems since the latter face external
unpredictable events and internal complexity, and also stated the limitations of the traditional
hierarchical control system being centralized, rigid, and unable to act quickly responding to changes and
failures.
The last level of the automation pyramid is the machine/device level. On the contrary to the top two
layers, this level has a naturally independent control level (Rojko, 2017). From another side and
responding to customization and flexible production initiatives, machines must operate independently
or in coordination with humans to produce customer-oriented manufacturing that constantly works to
maintain itself. Thereby, the machine’s configuration needs to be adapted in order to become an
independent entity that can collect data, analyse it, and interpret it. (Sung, 2018) proposed self-
optimization, self-cognition, and self-customization as configuration aspects capable of enhancing
machines and enabling manufacturers to communicate with computers rather than just operating
machines.
Continuing on, Industry 4.0 is mainly represented by CPS, IoT and cloud computing (Henning et al.,
2013; Hermann et al., 2016; Jasperneite, 2012; Lasi et al., 2014; Moeuf, 2018). The IoT is the main
catalyst for advanced manufacturing systems such as cloud computing, likewise it is a main medium for
the development of smart factories (Thames and Schaefer, 2016). CPS enables a higher level of
integration and coordination between physical and IT elements (Gürdür et al., 2016). Research indicates
that with the introduction of CPS, machines will be able to communicate with each other using
decentralized control systems and therefore help optimize production. Industry 4.0 generally comprises
many complex components and has wide applications in many industrial sectors (Xu et al., 2018b).
(Frank et al., 2019) subdivided the related technologies to the Smart Manufacturing dimension into six
main purposes: vertical integration, virtualization, automation, traceability, flexibility and energy
management.
Factory’s vertical integration comprises advanced ICT systems that integrate all hierarchical levels of
the company – from shop floor to middle and top-management levels – helping decision-making actions
to be less dependent of human intervention. To reach vertical integration, the first step at shop floor is
the digitalization of all physical objects and parameters with sensors, actuators and Programmable Logic
Controllers PLC. The data is then gathered with Supervisory Control and Data Acquisition SCADA, for
156
production control and diagnosis at the shop floor. At the managerial information layers, Manufacturing
Execution Systems MES obtain data from SCADA, providing production status to the Enterprise
Resource Planning ERP system. When all systems are properly integrated, the information of production
orders also flows in the inverse way (downstream), from ERP to MES and then to SCADA, helping to
deploy the enterprise resources into manufacturing orders. Therefore, vertical integration provides more
transparency and control of the production process and helps to improve the shop floor decision-making
process. To enhance adaptability for different types of products, Smart Manufacturing comprises
networked machines at shop floor, through machine-to-machine communication M2M. M2M consists
in a communication system with interoperability, which makes machines capable to understand each
other, facilitating their adaptation in manufacture lines.
This capability is supported by virtual commissioning, which emulates the different PLC-codes of
machines and validates virtually setup procedures, avoiding extended downtime due to the long setup
of equipment. This simulation is more advanced with digital manufacturing, which besides PLC-codes
also considers data from all virtualized objects of the shop floor and then simulates operations’
processes, considering several parameters that can affect production.
Smart manufacturing also promotes an enhanced automation. Robots can perform tasks with more
precision than in the past, increasing productivity while being much less prone to fatigue. In our work,
we differentiate robots and automation from collaborative robots. The former is designed to automatize
operational processes and, therefore, we included it as a part of the Smart Manufacturing dimension,
while the latter is designed to work with humans, supporting tasks that help to enhance human’s
flexibility and productivity. Therefore, were included collaborative robots as a technology of the Smart
Working dimension, as we explain after.
Moreover, artificial intelligence gives support for Smart Manufacturing in many ways. In machines,
advanced analytical tools can analyse data gathered from sensors to monitor and forecast machinery
failures, overloads or any other problems. This enables predictive maintenance which helps to avoid
downtimes due to unexpected failures during the production process. Machines with artificial
intelligence can also automatically identify product nonconformities in earlier stages of the production
process, increasing quality control and reducing production costs. Furthermore, artificial intelligence
also complements systems like ERP, predicting long-term production demands and transforming them
into daily production orders, considering last-minute orders and operations’ restrictions.
For internal traceability, sensors are applied in raw materials and finished products in the factory’s
warehouse. This optimized inventory control gives support for recall actions, through identification of
specific components in batches of finished products. Internal traceability can also give support to
adaptable systems with flexible lines, in which machines read products requirements in the sensors
157
embedded in them, and perform the necessary actions to manufacture them. Flexible lines can also
comprise modular machines that are easily plugged into a manufacturing line with minimum setup. This
enables the production of different types of products at small batches, with minimum loss of
productivity. In addition, to customize products, additive manufacturing is a promising technology of
the Industry 4.0 concept. Additive manufacturing uses 3D printing of digital models that can be altered
for customization, using the same resources to manufacture different goods. Additive manufacturing
also promotes a sustainable production, as it only requires one process that generates less waste than
traditional manufacturing. However, for large-scale productions, the use of additive manufacturing is
still limited due to its low throughput speed. Lastly, to enhance factory’s efficiency, Smart
manufacturing also comprises energy management (monitoring and improving energy efficiency).
Efficiency monitoring relies on data collection of energy consumption in electrical power grids, while
its improvement is achieved through intelligent systems for energy management that schedule intensive
stages of production in times with favourable electricity rates.
To sum up, (Lin et al., 2019) discussed the importance of improving manufacturing systems to deal with
unpredictable external events and internal complexity, especially when the traditional hierarchical
control system is centralized, rigid and unable to act quickly in responding to changes and failures. Four
key elements are identified by (Lin et al., 2019) for manufacturing systems evolution as represented in
Fig. 14: Agility for the management of unexpected events, Adaptability for product diversity, Efficiency
for process optimization and Reorganization to create new features.
Fig. 14 Key elements for manufacturing systems evolution (Lin et al., 2019)
158
5.4.3 Existing 4.0 Architectures
In the bibliographic analysis of this chapter, we evoke some innovative architectures from the literature.
The objective of the analysis of existing 4.0 architectures is to define an original positioning in relation
to the state of the art which leads us consequently to identify our research hypothesis.
In this context, several innovative architectures were proposed in the literature in order to surpass these
limitations and to meet industry 4.0 expectations. (Lin et al., 2019) Proposed an architecture for Cyber-
Physical Manufacturing Control System Fig. 15, reviewing the weaknesses of the traditional CIM
Pyramid in terms of limitations and complexity when it comes to the reconfiguration of the production
line through the MES, and the lack of real time update between the ERP system and the MES regarding
the production status. Therefore, another architecture was proposed based on transforming machines to
CPS entities with embedded sensors, connected with each other and communicating directly to a cloud
/ fog computing platform playing the role of operation management. Industrial Micro Controller Units
MCU were used to control machines and interact in between with the cloud platform. This system has
the ability to self-aware, self-predict, and self-diagnose data being treated based on machine learning
techniques through the cloud.
Similar to the cloud platform of the abovementioned model, (Bonnard et al., 2019) proposed a big
data/analytics platform for Industry 4.0 implementation in SMEs Fig. 16. Whereas this platform requires
a ‘flexible system’, it consists of a Representational State Transfer REST and Application Programming
Interface API collecting data from connected objects, PLC’s, SCADA, MES, and ERP designed in a
service-oriented architecture, published via a web interface or a mobile application and connected to a
MySQL database. The platform supports several protocols allowing communication with different
entities from the shop floor such as OPC-UA, Modbus or Profibus, wireless communication, or Ethernet.
Collected data will be stored on a local machine in order to be sent then via API REST to the cloud
server. The platform provides data processing by the analytics AI and Genetic which allows decision-
making support and recommendations where action planning can be developed.
159
Fig. 16 Platform for Big Data Analytics (Bonnard et al., 2019)
A domain framework for implementation of IoT ecosystem and expanded on production planning and
scheduling in shop floor is proposed by (Zdravković et al., 2018). (Zdravković et al., 2018) addressed
mainly the system dynamics and the environment dynamics by using smart automation where
production resources are able to make decisions in order to optimize the process execution. Hence, the
shop floor IoT system could provide higher agile, and reconfiguration means to the production system
when machines and entities communicate with each other improving the decision-making process based
on utility-based-agents as work centres in a holonic multi-agent system.
(Block et al., 2018) introduced a production planning and control PPC approach Fig. 17, based on the
concept of using CPSs connected to machines, collecting data when no central database is required,
sharing information with other CPSs and containing a model of the considered physical object. Core
Manufacturing Simulation Data CMSD is used for data exchange and linked to AutomationMLin order
to ensure communication with higher-level processes and MES. The decentralized CPSs are considered
as edge computing level since all data are processed and simulated on the CPS machine. Finally, an
event-driven PPC is deployed in a central station to encapsulate units in a Service Oriented Architecture
SOA.
160
Fig. 17 Decentralization of the automation pyramid (Block et al., 2018)
(Theorin et al., 2017a) suggested the Line Information System Architecture LISA, an event-driven
manufacturing information system based on SOA Fig. 18, which offers flexibility and scalability both
for control of low-level applications and aggregation of higher-level information, such as KPIs. LISA is
introduced as an Event-Driven Architecture EDA based on flexible message structure mainly the
message bus using Enterprise Service Bus ESB, a special simple message JSON format and service
endpoints characterized as loose coupling applications. This system enables firstly the integration of
new services in an easy way, and secondly the optimization of the production processes due to its
deployment in the shop floor layer. Regarding control services, LISA supports the interaction between
production machines, coordination software, and field devices.
An application has been occurred to demonstrate the control service using LISA consisting of a PLC
connected to a CNC machine, and using JGrachart programming language which supports service
orchestration with Devices Profile for Web Services technology DPWS and OPC UA.
161
Fig. 18 Overview of the LISA communication architecture (Theorin et al., 2017a)
In the context of modularity, (Kim et al., 2020) have developed a modular factory testbed for the rapid
reconfiguration of manufacturing systems Fig. 19. The testbed is designed in a multi-agent architecture
for distributed shop-floor control, consisting of three main layers:
- Coordinator, responsible for factory configuration and production planning, developed using C#.
- Workstation agent, assure the communication with other agents using wireless networks,
responsible for controlling and monitoring the assembly operations in the workstation from
scheduling to internal operation management, developed using ANSI C. The workstation agents are
considered as clients in the system reporting to the coordinator (server).
- Workstation executor, an embedded control software developed using LabView, responsible for
controlling field devices in a workstation and managing internal communications using PC-based
control systems instead of PLCs for rapid reconfiguration. A message-based parametric
reprogramming has been defined as a command format to ease the reconfiguration of production
tasks and reduce its required time.
162
Fig. 19 Communication and control structure of the testbed (Kim et al., 2020)
A modular manufacturing system is presented by (Gorecky et al., 2016) Fig. 20, consisting on five
conceptual components and applied on an industry 4.0 pilot plant:
- Product. The product layer considers a dynamic production environment where products should be
tracked. RFID tags were used to assure identification of each part and the respective digital
information specifications.
- Production layer. the production layer is built from several production modules to achieve different
production processes such as handling, machining, storing, … and equipped with a modular
transport system (revolving conveyor belts). the digital product memory provides the machine /
module with the production parameters, and the latter status gets updated when processing is done.
Each module has a standardized operating concept which gives information about operating states
and behaviour states of the respective module.
- Supply layer. This layer describes operational supply infrastructure of the modules such as current
supply, pressured air, security concept and communication.
- Integration layer. The integration layer has the function of connecting the production modules and
the IT system, by collecting and aggregating data from the production modules in order to feed the
IT system layer which follows a service-oriented architecture.
163
- IT system layer. at the top, the IT system layer represents the dynamic control, process planning and
simulation of the production plant, from ERP, to MES, PLM and data processing.
The study conducted in the abovementioned discussed basics and presented architectures, focus mainly
on developing models to resolve limitations in traditional manufacturing systems and find solutions to
make better values from data collection and reconfiguration use of production lines for full deployment.
It has been clarified that vertical and horizontal integration of Industry 4.0 relies on the evolution of
manufacturing systems assuring flexibility, modularity, reconfiguration, interoperability, and
intelligence. The architecture of advanced and complex manufacturing systems is heterogeneous,
unbounded, dynamic and undefined. Modelling and designing these systems need to focus on services
over components, interoperability and cross-platform functionality, flexibility and distribution, and
abstraction to deal with complexity. Manufacturing models and architecture are designed to enhance
one or more functionalities proposed by the 4.0, taking into consideration the flexibility and the
modularity attended. For instance, the Service Oriented Architecture SOA focus on the service
orientation principle, the IoT Architecture reinforces real-time capabilities and interoperability, and CPS
and Modular Architecture augment modularity and decentralization. An advanced manufacturing
system can be based on a combination of more than one architecture. Some are considered as modelling
approaches such as SOA and Manufacturing as a Service MaaS, and others are an architecture design
164
like Industrial Internet of Things IIoT and CPS. Modelling approaches and architecture can be combined
since the modelling paradigm is chosen for the control system depending on the type and objectives of
the production application, and then the manufacturing system shall be designed through a dedicated
architecture (Borangiu et al., 2019). The dynamic issue of the traditional vertical structure of the
automation pyramid is our main concern. Based on what has been discussed, the evolution of the
production systems towards Industry 4.0 is strongly dependent on vertical integration. Two key factors
are to be considered in order to achieve the transition goal:
- The first is the flexibility factor identified by (Lin et al., 2019). A production system capable of
adapting to new orders/products (under the factory capacity) and able to respond quickly to
fluctuations in demand and workflow assuring the mass customization aspect.
- The second factor identified by (Sung, 2018) concerns data management. A production system
capable of exploiting data collected from machines, sensors, failures, and historic data, in order to
learn and self-optimize its production processes.
Concerning the data dimension in the context of Industry 4.0, manufacturing systems are updated to an
intelligent level. Intelligent or smart manufacturing aims to convert data acquired across the product
lifecycle into manufacturing intelligence in order to yield positive impacts on all aspects of
manufacturing (Tao et al., 2018). It enables all physical processes and information flows to be available
when and where they are needed across holistic manufacturing supply chains, multiple industries, Small
and Medium-sized Enterprises SMEs, and large companies. Intelligent manufacturing requires certain
underpinning technologies in order to enable devices or machines to vary their behaviours in response
to different situations and requirements based on past experiences and learning capacities. These
technologies enable direct communication with manufacturing systems, thereby allowing problems to
be solved and adaptive decisions to be made in a timely fashion. Some technologies also have artificial
intelligence AI, which allows manufacturing systems to learn from experiences in order to ultimately
realize a connected, intelligent, and ubiquitous industrial practice (Zhong et al., 2017). According to
(Lee et al., 2018), AI is a cognitive science with rich research activities in the areas of image processing,
natural language processing, robotics, and machine learning. For example, Robotic systems are playing
an interesting role in enabling Industry 4.0 implementation. Modern robots are characterized as systems
offering autonomy, flexibility, cooperation, and reconfigurability (Kamble et al., 2018). A layer of
flexibility and intelligence is offered through the integration of machine learning techniques to the image
processing stage, and the interconnection between robots and machines which extend the
decentralization aspect, and assure several benefits such as lower defect rate, higher quality and
reliability (Ghobakhloo, 2018; Gupta, 2018), and improved security.
In order to enhance the production performance, assure the security means of workers and machines,
improve product quality, and develop smart products, industrials must deal first with a big amount of
165
data generated through the product lifecycle including visual inspection images and quality control
processing (Bajic et al., 2018). Secondly, Manufacturing systems must be able to integrate such flexible
and AI-Enabled modules in as less as possible intrusive way.
166
6. PhD Hypothesis
After this deep study covering the aspects and the principles of Industry 4.0 and the detailed
bibliographic research, we propose one hypothesis, based on the variety of main challenges and
problematics facing the production system evolution toward the digital transformation. Having declined
the principles of Industry 4.0 supported by a detailed literature study, we propose in this chapter a
hypothesis which considers that the transition to the digitization of production systems depends on major
principles which must be contextualized according to the objectives targeted by companies.
Before proposing the hypothesis, a bibliographic analysis of presented recent architectures and models
proposed in the literature are discussed in section 6.1. Section 6.2 introduces our hypothesis. Finally,
section 6.3 presents the structure followed to validate the proposed hypothesis.
We notice in the literature that the subject addressed by this doctoral thesis represents an originality and
importance in terms of topic, which explains the scarcity of approaches that focus on the evolution of
existing production systems towards the digital transition especially when considering the entire
production system including different levels. The architectures presented in the context of the integration
of Industry 4.0 into production systems are innovative but on the other hand are mainly based on the
development of new data transfer initiatives that are not recognized until today in companies and which
require radical changes in existing architectures which can cause potential disruptions in production
sites. In addition, a second critical aspect to mention during this analysis concerns the limitation of such
architecture proposals like the one developed by (Block et al., 2018) considering specifically the
application in “process manufacturing” which mark a big difference with the field of industrial
manufacturing.
In addition, several research studies consider the management dimension of generic digital transition
roadmaps in companies. We cite (Kamble et al., 2018) who proposed a sustainable framework of
Industry 4.0 built with three components, the technologies of Industry 4.0, the integration of processes
and the sustainable results of this integration. Based on an in-depth review, this framework was proposed
given the promised facilities of Industry 4.0 technologies and its ability to integrate business processes
and activities, making the manufacturing system more flexible, economical and environmentally
friendly of the environment.
A three-step model for the transformation of Industry 4.0 “Envision, Enable, Enact” is proposed by (Erol
and Schumacher, 2016). The main objective of the model is to guide companies in the development of
their specific Industry 4.0 objectives as well as a set of measures to achieve their company-specific
vision and roadmap. (Ghobakhloo, 2018) proposes a strategic roadmap for Industry 4.0 transition, taking
167
into account the entire supply and value chain of the company, in the form of recommendations aimed
at facilitating the transition to Industry 4.0.
However, our approach, which will be contextualized through the proposed hypothesis, takes into
account the current state of the production system operating in 3.0 mode and tries to achieve an
innovative approach which appears in the proposal of a model followed by a framework of the digital
transition of manufacturing systems while respecting the existing architecture of the production system.
The analysis carried out above leads us to take up and justify the positioning of our research problem
defended in Chapter 4, Which model for a successful digital transition of 4.0 manufacturing systems?
which lies between the three defined axes, ‘digitization’, ‘manufacturing system’, and ‘production
processes’.
The agility offered by Industry 4.0 allows plants to handle a more diverse product range with shorten
lead time, and quickly customize products to specific requirements, assuring higher responsiveness to
customers’ needs. While Industry 4.0 offers enormous, new opportunities for businesses; the scale of its
impact also brings new challenges.
A detailed state of the art has been provided to define the principles and concepts of the digitization of
manufacturing systems, and to identify the respective main challenges to this digital integration. The
digitization of production systems requires a specific approach that considers the entire production
system with its different layers as well as its current architecture. This approach aims to evolve the
production system while ensuring the functionalities expected by companies.
The dynamic issue of the traditional vertical structure of the production is to be investigated, since
production systems need to be capable of adapting to new products and responding quickly to
fluctuations in demand first. Second, production systems must allow deploying data collected from
machines, and sensors in order to learn and self-optimize its processes. The main challenges and issues
defined according to our research direction and interest are the implementation of Flexible and Modular
Manufacturing System in the context of Vertical Integration which requires critical changes considering
the different level of company’s structure.
Consequently, and following this oriented bibliographic study, we propose the following hypothesis
which comes to answer our research problem identified as following:
Evolving production systems towards digitization in the context of Industry 4.0 requires acting on
three main axes Production System including Shop Floor Control System SFCS, Machines &
Advanced Techs, and Data Processing & Deployment.
168
Fig. 21 presents the different axes of interest / dimensions identified by the hypothesis. The first
dimension Production System & SFCS concerns mainly the production information system, its ability
and requirements to adapt to the flexibility and modularity expected. The production information system
includes the agility of the IT network as well as the digital production platforms such as MES, ERP and
other software. This dimension is concerned also by the Shop Floor Control System and its configuration
which must respond to fluctuations in demand and allow for flexible production processing and control.
The second dimension Data Processing and deployment is about enabling data flows in both descendant
and ascendent direction. After the implementation of sensors and IoT, data should be collected from
machines and feedbacked to it with mined data, enhanced decision making and optimized processes.
The third dimension Machines and Advanced Technologies considers mainly the implementation of
connected machines, the integration of Cyber Physical Systems CPS and advanced technologies
introduced by Industry 4.0. This dimension takes into consideration the machines configuration and
requirements in the context of flexible manufacturing and modular production processes.
The objective of this proposal is to develop a scientific approach of progressive transition to the
digitization of production, thus offering new functionalities that are absent in existing production
models. This is a generic approach intended to be adapted to the trajectory that the company sets itself
under constraints of the desired performance, the expected technological adoption and its aim of
evolution. Taking into account these three axes identified by the hypothesis, we consider that the model
that responds to the desired digital transition followed by the integration framework will have a tangible
effect on the evolution and adaptation of planning and management of production, as well as the control
and processing of production Fig. 21.
From new functionalities identified and targeted by the company, in terms of operational flexibility,
integrated quality, decision autonomy, predictive maintenance, optimization of energy consumption and
the desired level of portability, our approach presented in Fig. 22 consists in proposing a new conceptual
169
model which defines and integrates a distributed system architecture, cyber-physical systems CPS,
connected 4.0 technologies, an integrated system for storing and processing data. Obviously, the
technological aspect is not the only factor of success, other factors such as the project management, and
required competencies contribute to the outcome of such a project which marks a break in industrial
practices. However, only the technological aspect will be discussed in this thesis.
Subsequently, and in order to apply the developed conceptual model, we propose an innovative
framework for the digital transition of manufacturing systems that takes into consideration the different
conceptual dimensions of our model and transforms them into a detailed support structure to deploy the
digital transition into existing production systems.
Finally, the digital transition framework will be applied to the industrial platform 4.0 introduced in
section 2.3.2. The experimental integration work will be well detailed in chapter 9. From the deployment
of the new system architecture, we subsequently propose a design methodology for the configuration of
modular processes followed by the application of this modular approach in the manufacturing program.
A quantitative application will finally be carried out on the flexibility of the system deployed in order
to justify our scientific approach.
In summary, the objective of the model and the proposed framework is therefore to develop a new
modular process to deploy the following functionalities of the production system:
170
- Manufacturing flexibility
- Personalization
- Portability
The aim of the proposed model is therefore developing a new modular process to enhance
manufacturing systems. Our modular approach is to transform a rigid production system into an agile
production system by applying 4.0 functionalities on the different levels / layers of a production system,
enhancing the production planning and monitoring performance and evolving the production control
and processing flexibility in the context of Industry 4.0.
In the remainder of this thesis, chapter 7 presents the model. chapter 8 discusses the integration
framework proposed assuring mainly the vertical integration and allowing the horizontal one into the
company’s structure. In chapter 9, a novel integration architecture toward the digital manufacturing
transition is developed and explained based on the literature review findings in order to help companies
toward the digital transformation. chapter 10 will discuss results and findings, and chapter 11 concludes
the PhD thesis.
171
7. Modelling of 4.0 production systems
The production system is a complex system that transforms materials, energy and knowledge into added
value products and services. Manufacturing has evolved from artisanal production to mass production
and even mass customization. With the introduction of technologies, manufacturing systems have
evolved into flexible, reconfigurable and intelligent production systems (Zhang et al., 2017).
Based on the concepts defined and cited above, (Benfriha, 2020) proposed an initiative to model the
concept of 4.0 production systems. First of all, the model is presented in its originality adapted to a
general modelling of a 4.0 production system. Consequently, the three axes of the proposed hypothesis
are being associated to the different concepts evoked by the model. Second, the model will be adapted
to the situation of an existing production system that needs to be upgraded to a 4.0 production system.
The goal here is to exploit the concepts developed in the literature and to build a new model of 4.0
production system. We have identified three representative concepts of production systems that could
be encountered in industry, addressed by (Bonnard et al., 2019), by (Lina et al., 2019), and (Gorecky et
al., 2016). In general, all the concepts are concurrent, however their specificity originates in the context
of use. Before embarking on a process of implementing Industry 4.0 concepts in a company, several
experts recommended an optimized organization of production as a prerequisite, in particular through
the use of methods such as lean manufacturing. It is worth to mention at this stage that any
implementation of a 4.0 production system must consider the Reference Architecture Model for Industry
4.0 RAMI4.0 as a digitalization reference framework for standardization.
From an analysis of the concepts defined in the literature within the perimeter of the three axes well
defined by the proposed hypothesis, we present here a convergence initiative which consists in
proposing a new simplified general concept that reorganizes different technologies and architectures
involved in a complete and coherent vision of a future 4.0 production system model. This general
conceptual model considers different technological bricks and essential basic concepts, represented by
the 5 axes identified by (Benfriha, 2020), see Figure 23, which constitutes the basic principles of the
modelling a digital production system 4.0. The three axes identified by the hypothesis “Production
System & SFCS”, “Data processing and deployment”, “Advanced Machines and Technologies” with
the objective of effecting the digital transition of production systems rely, as shown in Fig. 23, on the 5
essential conceptual axes of the modelling process proposed by (Benfriha, 2020): applications, data,
CPMS systems, IIoT systems, and IT systems.
172
Fig. 23 General modelling concept (Benfriha, 2020)
The 1st axis concerns the dimension of advanced machines and technologies introduced by Industry 4.0
and considers that the manufacturing system must be configurable in Cyber Physical Manufacturing
Systems CPMS. This axis assumes that machines available natively integrate the capacity to work in a
connected environment or to make them evolve to be compatible with a 4.0 production system. This
allows a feedback of data that reflects their operating state and at the same time can execute generated
data for monitoring via the local network. Added to this, the possibility of embedding intelligent
applications (MAS) on certain machines allowing them to analyse their own data and activate
interactions with their nearby environment.
The goal of CPS is to build an engineering system that is controllable, reliable, scalable and capable of
interacting in real time. CPS are based on the latest developments in computing, information and
communication technologies, and manufacturing science and technology with the perspective of
increasing production performance through intelligent control systems based on the data collected and
transformed (Joost et al., 2012). To meet new market demands, new technological paradigms must be
applied to manufacturing systems with the aim of improving agility, responsiveness and ensuring
product quality. CPS can meet the challenges of future manufacturing systems as they bridge the gap
between physical elements and remote-control systems. The combination of several interconnected CPS
and enhanced by cloud technologies has led to increase the processes of new 4.0 functionalities.
173
There is an abundance of definitions of cyber-physical systems in recent literature. The concept of cyber-
physical systems is to integrate computational processes with physical processes through networks,
which are emerging as a new generation of intelligent engineering systems. Computing processes
supervise physical processes via information networks, conversely, physical processes affect computing
processes. A CPS interacts with its environment in which it retrieves data, processes it and through a
feedback loop controls or influences the process with which it is associated. An intelligent production
line can include many reconfigurable cyber-physical systems capable of interacting and exchanging data
in real time with other remote systems. The heterogeneity and complexity of CPS applications represent
a real challenge in their development where aspects of reliability, security and certification are
prioritized.
The 2nd axis concerns the IIoT devices which superimpose the CPMS systems by an extended coverage
of a variety of sensors and cameras, considering the dimension "Data processing and deployment". The
IIoT layer paves the way for virtual modelling applications of manufacturing processes (DTS and DDT)
and allows data processing via Cloud computing or Edge computing. The Internet of Things IoT is one
of the advanced disruptive technologies and considered by McKinsey Global Institute as a technological
advancement that "will transform life, business and the global economy."
IIoT system goes much more beyond collecting data from sensors, it is considered as a holistic system
that deals with data from collecting to analysing a large amount of data that can be used to improve the
overall performance of industrial systems. The Industrial Internet of Things framework includes
infrastructure, technologies and applications that bridge the gap between the real industrial world and
the virtual world (Islam and Kim, 2019). (Khan et al., 2020) give the most adequate definition of IIoT
system as “the network of intelligent and highly connected industrial components that are deployed to
achieve high production rate with reduced operational costs through real-time monitoring, efficient
management and controlling of industrial processes, assets and operational time”.
The relevance of IIoT systems is also considered for the dynamic production planning and scheduling
in the shop floor. IoT system can integrate the shop-floor system, location sensor data are used to make
the most optimal internal logistics decisions; accelerometers can be deployed for the purpose of
predictive maintenance (vibration monitoring); ultrasonic sensors can be used to look for cavities in
castings in a production line, etc. IoT system considers reliability of critical services, such as those
related to work safety regulations (by prioritising data traffic from air quality sensors and smoke
indicators) and ensuring safest internal logistics routes (proximity sensors).
The 3rd axis is an essential matter considered by the “Production System & SFCS” dimension. This
axis addresses the software applications necessary for the execution and control of a production system
with a specific care on the software interfacing issue. Indeed, various applications are involved in
174
different phases of manufacturing on an ad hoc or continuous basis. The problem of software
architecture and interfacing is a crucial point in the success of an integrated and autonomous
manufacturing process.
The traditional approach of process planning in various small and medium enterprises is done based on
knowledge and experience of the personnel involved in the manufacturing activities. The design and
development phase of a smart product must consider including the data of the following three stages:
process planning, operation scheduling and operation sequencing. The major aim must be to collect as
much data as possible so that digitalized and virtual network of manufacturing machines and the product
can be established directly. The configuration of the process planning, and operation sequencing tools
must provide an enhanced efficiency for process optimisation and an assured capability for mass
customization of products.
According to (Kumar et al., 2019), there are three important criteria associated with product planning
software: process planning, operation sequencing and finally machine selection. The software for
product planning collects information from different parts of the supply chain linked to the important
manufacturing activities. Mass of the product, its material, shape, surface finish, geometrical features
and the quality forms the sub-criteria for the process planning criteria. The second criteria under product
planning software is operation sequencing. The product planning software must consider aspects which
influence the manufacturing processes, such as finishing operation, number of operations required, the
fixation method and the priority of operation. The third major criteria for product planning software is
machine selection and scheduling criteria. Related important sub-criteria include time of operation, tools
required, quality required, cost of machine, availability of machine and number of products ordered.
Selection of machine forms an important part of the product planning software, which is done through
the information on the available machines in the manufacturing system. Taking into consideration these
criteria enable the manufacturing system to perform the operation with the overall objective of cost
minimization and time of manufacturing.
The 4th axis raises the question of the data generated or imported from several point of view, storage,
security, processing, filtering, sharing, and visualization. It is certain that Industry 4.0 will generate more
data, and more than enough to know how to collect, analyse and use in order to derive added value. The
analysis of the data makes it possible to generate information which can increase the performance of the
production system.
This axis is undoubtedly at the heart of this industrial revolution and represents a major issue in terms
of management, optimization and cybersecurity. Companies need to be able to identify and locate data
in its original formats at a data lake. The challenges of controlling data are significant. Indeed, according
to the Data Literacy Index (Jung, 2002), companies that succeed in improving their ability to use data
175
could increase their value by 3% to 5%. In addition, it appears to have a strong correspondence between
improving the ability of the staff to use data for decision-making on the one hand and business
performance measures, such as gross production margin, return on assets, return on equity and
profitability on sales. A data lake designates a global storage space that absorbs the flow of data present
within a 4.0 production system. It involves doing so with enough flexibility to interact with the data in
their original format, whether raw, structured, and unstructured data. Above the storage, one of the
challenges of the data lake is to be able to easily process and transform data in order to extract useful
information to enhance the production processes performance but also to enrich the design phase of the
product. The choice of processing data locally EdC or in a cloud CdC will depend on the purpose of use.
The 5th axis is undoubtedly the most complex one concerned with the vertical integration and the
communication between different layers of the production system.
Indeed, the traditional pyramidal organization stiff restrictions on communication, where all layers can
only change information with adjacent neighbours. This restriction should be surpassed in industry 4.0,
as basically, all data should be easily shared through the distributed system, regardless of its physical
places. To prevent this, industry 4.0 technology and paradigms must be conceived to provide adequate
support to transport and process a large amount of information at the end. In advanced automation,
production lines and production processes are more compact and modular, but they can also establish
direct connections effectively (either physical or logical) to share information. And since the market
increases the demand for modular and flexible systems, the availability of powerful built-in
infrastructure and production on the basis of advanced communication architecture is increasingly
important.
According to (Zunino et al., 2020), smart factories based on cyber physical production systems CPPSs
require a resilient network supported by: the availability of large communication bandwidth to support
the exchange of huge amounts of data in real-time, the security of communications and protection
against attacks and malicious behaviour, the augmented resilience of networks to grant normal operation
with constant levels of quality of service QoS, the scalability of future networks to accommodate the
exponential growth of connections and dynamically changing demand for QoS, the predictability of
links and connections to enable easy adaptation to quickly changing needs and working situations, and
176
finally the real-time communication support to enable a strong enterprise control systems integration,
where factory control and supervisions are tailored to small production batches.
(Theorin et al., 2017b) addressed the relevance of integrating SOA into the manufacturing operations
management level and the control level. Most SOA tools are tailored for business processes, which do
not have strict timing or resource requirements, while devices which execute at the monitoring and
control level often have strictly limited memory and computational power. SOA is a distributed software
architecture where self-contained applications expose themselves as services, which other applications
can connect to and use. To reach its full potential, SOA applications should be self-describing,
discoverable, and platform- and language -independent. This leads to loose-coupling and high
flexibility. The adoption of SOA in a company typically starts as an IT initiative to improve
infrastructure efficiency and can then mature into optimised use for business purposes.
In the case of an existing production, the proposed model must take into account machines and networks
already installed. The approach we are proposing consists of analysing what already exists in order to
enhance it with a technological layer adapted to the desired functionalities and to the potential of the
company to make investments in the production lines in respond to the desired business objectives.
Before starting the transition process, we consider that the company wishing to evolve towards a 4.0
production system already benefits from an existing CIM architecture Figure 24 which operates on a
model close to 3.0 and its manufacturing process is already subject to continuous improvement, in
particular using lean manufacturing. This starting point is crucial for the successful transition to evolved
systems.
Based on the conceptual model introduced above and the axes identified by our hypothesis, the
transformation of the workshop involves 5 aspects. The material aspect considering the acquisition of
equipment for transfers between machines, integrated quality control or additive manufacturing
machines. Added to this, the reconfiguration of existing machines in terms of controls in order to make
177
them compatible with a connected production system. The second aspect concerns overlaying an IoT
layer such as sensors, cameras with dedicated networks interoperable with the existing network.
The third aspect concerns the acquisition of software, in particular to manage the workshop. In addition,
the development of a specific digital management and industrialization environment is required, to
which the control functions will be transferred, without forgetting the software interfacing issues called
upon to dialogue during the execution of a production plan. The 4th aspect concerns the computational
capacities, deported or local, necessary to support a 4.0 production system equipped with advanced
functionalities, in particular when using cameras with expectations of real-time responsiveness.
The last aspect concerns programming of machines. Indeed, in 3.0, machine programs are designed to
have a configuration dedicated to mass production. Machines become a single entity that ensures the
smooth flow of a well-organized production plan. On the other hand, in 4.0 each machine is a separate
entity and can be considered as a CPS, or included in a CPS, connected, intelligent and autonomous.
This modularity at the workshop level leads us to design specific manufacturing programs for a modular,
autonomous, adaptable and reconfigurable production plan.
178
In conclusion, the modelling of 4.0 production systems explores different concepts combined and
deployed to build an integrated and intelligent system. The concepts interfere with different fields and
expertise and their interoperability constitutes a major technical challenge. A large majority of the
articles consulted evoke a multi-scale complexity, including but not limited to interfacing, security and
competencies. The industrial strategies for evolution face as well consultancy issues delivered by the
vendors of digital platform solutions and manufacturers of 4.0 technology systems. This competition,
under a well-developed marketing envelope, often leads to difficulties in our choice of solutions and in
their implementation. Therefore, a transition framework toward industry 4.0 production systems is
proposed in the next chapter in order to apply the proposed model concretely and guide companies
aiming to evolve their production systems.
179
8. Proposed transition framework toward industry 4.0 production
systems
The conceptual model developed above is followed by a novel proposed framework in order to exploit
it and generate a supporting structure aiming to guide the integration of 4.0 in the existing production
system. As identified in the last section, the project management of this integration is an essential key
to be considered. Since we are focusing in this thesis on the technical and development dimension,
existing industry 4.0 frameworks form a project management point of view are identified based on the
literature in the next section. Following, a design principles framework assuring mainly the vertical
integration and allows the horizontal one into the company’s structure is introduced and explained.
(Kamble et al., 2018) proposed a sustainable industry 4.0 framework model built with three
components, industry 4.0 technologies, process integration and sustainable outcomes Figure 25. This
deep review proposed this framework based on the promised facilities of industry 4.0 technologies
and its ability to integrate the business processes and activities, making the manufacturing system
more flexible, economical, and environmentally friendly.
The first layer of this framework includes big data technologies, virtual reality, robotic systems, cloud
computing, internet of things, simulation and prototyping, and 3D printing as the emerging
technologies of Industry 4.0 which have a great added value to the manufacturing process. This layer
has the main role in supporting the second layer, process integration, both maintained by the cyber
security as a critical aspect to well improve the system accessibility and protection. The successful
deployment of industry 4.0 technologies boosts the creation of smart products and processes. The
process integration has to ensure the implementation of Industry 4.0 principles as in (Kamble et al.,
2018), interoperability, decentralization, virtualization, modularity, and service orientation. The last
layer of this innovative framework, Sustainable outcome, has three dimensions. Starting first by the
‘Environmental protection’ - Industry 4.0 technologies have the potential to make the existing
manufacturing processes environmentally sustainable, since it allows for efficient resource allocation
based on production processes and other stakeholder’s real-time data resulting in sustainable green
practices (de Sousa Jabbour et al., 2018; Stock and Seliger, 2016). The 2nd layer is Process automation
and safety – an optimized management of the operations emerging from the stabilized manufacturing
processes shall improve the working conditions and environment, thereby many safety-related tasks
could be performed at the organizational level to prevent safety hazards and operational risks.
Economic sustainability is the last dimension – due to the digital transformation of the supply chain,
companies are expecting many economic benefits, reducing manufacturing costs and time by
improving predictive maintenance, reducing losses, improving purchasing decisions, and lean
180
manufacturing.
A three-stage model for Industry 4.0 transformation Envision, Enable, Enact, is proposed by (Erol
and Schumacher, 2016). The main goal of the model is to guide companies in developing their specific
Industry 4.0 objectives along with a set of measures to reach their company-specific vision and
roadmap. The Envision stage consists of developing an industry 4.0 vision taking into account the
company particularities. Different management levels are involved in the configuration of this vision,
along with different stakeholders and main customers and business partners. External experts are
invited to the Envision stage, in order to shed the light on different Industry 4.0 technologies and
theories. Followed up by concluding a vision in a form of strategic roadmap for the long-term future
of the company. The “Enable” stage is dedicated to transform the long-term vision into a concrete
business model. The aim of this phase is to identify the short, mid and long-term transforming level
considering different production or manufacturing processes, mainly, the network, the process, the
production and the market. Third stage is Enact, where the strategies are well separated into
managerial, technical, and organizational projects to be implemented. This approach ensures the well
planning of the business model transformation in a collaborative way through the development of the
company vision and strategy Figure 26.
181
Figure 26 Three-stage model for Industry 4.0 transformation (Erol and Schumacher, 2016)
Introducing the digital transition as an enabler for Industry 4.0, this transition requires changes in the
company physical infrastructure, manufacturing operations and technologies, human resources, and
organizational models. (Ghobakhloo, 2018) suggests a strategic roadmap for Industry 4.0 transition,
taking into consideration the whole company supply and value chain, in the form of recommendations
aiming to facilitate the transition toward Industry 4.0. At first a strategic management should be
defined, and time-based plans and strategies are to be framed. At this point the goal is to define
functional needs and to plan technological implementation required by Industry 4.0 transition, provide
the cost-benefit analysis, and facilitate the internal collaboration and changes across the company.
Through the data driven marketing and the establishment of market sensing and learning strategy, the
marketing strategy is to enable the execution of smart products as services business model. The Human
resources strategy is to provide the company with the most important success factor in implementing
digitization. The assessment of human resources competencies is essential for the digital transition
strategy. The IT maturity strategy is a key potential to make different process integration possible. IT
efforts are required to propose new IT infrastructure architecture. Thereby, in order to provide the
required connectivity for data collection and interoperability an IT strategy is to be followed to
implement the new network. Integrated simulation, digital twin model, and Intelligent ERP are to be
developed through a smart manufacturing strategy, allowing better connectivity, transparency, agility,
flexible manufacturing and self-adaptation. Many steps are recommended by the roadmap such as
production process supervision via distributed control system, manufacturing execution system and
intelligent ERP system implementation. Last point focused on the smart supply chain management
strategy. As the mass customization is at the top of Industry 4.0 functionality, it is critical to achieve
real-time communication across the supply chain, and interoperability and security across different
platforms within the entire supply chain along with protecting each partner property.
182
8.2 Advanced Manufacturing System, A Design principles Framework
Industry 4.0 is a dynamic and integrated system for employing control over the entire value chain of the
products lifecycle. Vertical and horizontal integration and fusion of the physical and the virtual
components are critical for Industry 4.0, and technology trends implementation such as Cyber Physical
Systems and Internet of Things. In other words, the core concept of Industry 4.0 is to build an intelligent
and self-optimized manufacturing system based on the processed data and rely on the data mining
gathered and supervised in real time. Well, this data mining process should be well implemented in the
manufacturing system and capable of optimizing the configuration of the ongoing production process
(flexible manufacturing system) along with enriching the business level of the company based on the
exploited data. This implementation imposes consequently a list of requirements and radical changes all
over the production system, targeting each layer of the manufacturing system.
As the smart and connected manufacturing processes introduced by Industry 4.0 are increasingly
attracting companies trying hard to adapt to customer demands, a design principles framework is
designed and applied with a novel architecture that promotes the digital manufacturing transition. Taking
into consideration the negotiated requirements of the digital manufacturing, the proposed framework as
shown in Fig. 27 is based on design principles and functionalities of Industry 4.0 especially the
modularity dimension, which is composed of five layers explained in the following sections, assuring
mainly the vertical integration, and allowing the horizontal one into the company’s structure. The
vertical integration involves flexibility at the shop floor level but followed by flexibility at the shop floor
control level. The operational flexibility assured by the first two dimensions should be reinforced by the
fact of taking into consideration the data management dimension since the overall objective of the digital
transition is improving the operational performance through data mining. The full deployment of this
integration is assured by the modular service integration dimension at the production information system
level. Lastly, the openness for business-level communication and distributed industries dimension is to
guarantee the ability of the production system to enable horizontal integration across the industrial value
chain.
183
Fig. 27 Design Principles Framework
At the machines/devices level, technical aspects of the transition towards a 4.0 production system are
addressed by the application of the genetic concepts of Cyber-Physical Systems CPS, Industrial Internet
of Things IoT, and the integration of new functionalities to the industrial machines. The understanding
of industry 4.0 is the deployment of the internet of things to the industrial processes, extending it by
information and communication technology (ICT) as a key factor, alongside the development of CPSs
as the core elements of this digitization strategy (Bloch et al., 2018b). Consequently, the resulting smart
and connected objects drive to build a system that allows for more flexible, efficient, and transparent
planning control, and execution of production and logistics (Santos et al., 2017).
More concretely, technically addressing the flexible shop floor concept, several aspects regarding the
machines should be considered. First of all, a Flexible Manufacturing System FMS in the context of
Industry 4.0 is required. An FMS is differentiated from a traditional production system by its ability to
perform multiple operations from a limited number of resources. It is based on flexible machines such
184
as robots or Robotic Process Automation RPA. Machines flexibility requires at least solving issues such
as interoperability, remote access to machines, and enabling flexible configuration of the machining
process. In addition to the abovementioned requirements, integrating machines with a decision-making
level would effectively help in turning this machine into a CPS module, ideally when it is capable of
performing algorithms and communicate data to other modules within the network. Second, the
modularity of the manufacturing system is another aspect to be considered. Modularity in FMS is an
optimal trade-off between other traditional production systems (Kaushal et al., 2016). By switching
easily in a high range of volumes or variety, modular production systems can replace shop floors with
stand-alone machines or dedicated lines and improve the capacity of the production system for dealing
with uncertainty.
More specifically, A production system can be defined as a set of modules connected by flows, with the
function of transforming raw materials into a product. The overall flexibility of the production system
stems logically from the flexibility of the modules and flows that compose it. The modularity in this
context relies strongly on the flexible shop floor control system and is basically dependent on its
configuration.
To the end of implementing I4.0 and exploiting its innovative functionalities, it is hence necessary to
assure the flexibility of manufacturing control systems and production processes. As discussed in the 1st
section, the flexibility of the machines at the shop floor level once configured, needs to be supported by
the control system in order to be deployed to the production system. Flexible machines and modular
CPS mean flexible operations and processes, where their execution and sequencing require a flexible
shop floor control system. Therefore, a different configuration of our production systems is required to
allow first dynamic workflow and process planning, and second the real-time autonomous optimization
of the production processes, which is not supported by the conventional control systems due to its
inflexible configuration. In what follows several studies and solutions are proposed to develop and
deploy flexible SFCS.
(Bloch et al., 2018b) of the IAT institute considers the problematic of modular operations as an approach
that meets the increasing demands for flexibility and customization in the manufacturing industry, and
the need for a more diverse product mix. According to (Perzylo et al., 2019), instead of producing high
quantities of similar products over a long period of time, companies have to satisfy the market demand
for customized or even individualized products. As a result, production lines may have a multitude of
different variants, which may only be produced in small lot sizes. (Bloch et al., 2018a) suggested the
modularization of the process control system, using different process modules. These process modules
185
will provide encapsulated process functions as services to the Superior Control System SCS. Within the
SCS, services are orchestrated by the plant operator to achieve the desired production process.
(Karnouskos et al., 2010) addressed the gaps in the existing Supervisory Data Acquisition and Control
SCADA, and Distributed Control Systems DCS by the incompatibilities among the systems, hard code
data, different view on how systems should be configured, co-existence of very old technologies, and
the use of reactive process automation systems instead of proactive configuration. They considered that
control systems should be based on process control algorithms, and configured in scalable and modular
architectures and platforms (plug and play). Therefore, they proposed the ArchitecturE for Service-
Oriented Process-Monitoring and Control AESOP based on service-oriented process control algorithms,
scalable and modular SOA-based SCADA and DCS platforms.
(Theorin, 2012) considered as well, the integration of service-oriented architectures within automation
SOA-AT aiming to link the planning phase to the control system, using service-oriented process control
by decomposing processes steps as functions and assigning them to encapsulated services. Grafchart is
used as the process modelling and service orchestration language and Devices Profile for Web Services
DPWS as the service technology, which allows consequently a hardware-independent development of
the process logic and a simpler programming and reuse of control programs. The encapsulation of low-
level functions to services enables a higher degree of abstraction for the implementation of the control
logic (Theorin, 2012).
SOA used at the business process ERP and planning MES domains must be realized on the control level
by enhancing the functionality of the PLC and improving the software engineering for manufacturing
control (Ollinger et al., 2014). According to (Ollinger et al., 2014), the concept of a SOA-PLC
overcomes the current limitations of existing PLC systems, for a flexible, reusable, and dynamic
manufacturing environment. Encapsulating function blocks into web services field device make
functions accessible outside the PLC on higher level of the automation pyramid. In addition, the SOA-
PLC enables a dynamic orchestration of services realized by any DPWS on any conventional PC rather
than an inflexible sequential code. (Beckhoff Automation GmbH & Co KG, 2021) introduced the call
up method which ensured performance and data consistency between the PLC and the upper layers. The
call up method using an SOA-PLC shortened the communication roundtrip between PLC and MES via
OPC UA making the Remote Procedure Call (RPC) available right into the programmed PLC function
block, leading to higher production throughput, and reducing engineering costs for the data link between
shop floor and top floor quite dramatically.
The wide implementation of Internet of Things, cloud computing, big data analysis and Artificial
Intelligence in the industrial firm is enhancing the effectiveness of smart manufacturing and helping
186
companies to be more competitive by adapting data-driven strategies. Big data refers to large amounts
of multi-source, heterogeneous data generated through-out the product lifecycle, which is characterized
by high volume, variety (the data itself comes in different forms and is generated by diver resources),
velocity, veracity, and value. Generally speaking, big data generated by manufacturing processes can be
classified according to the following categories as identified by (Tao et al., 2018):
- Management data collected from manufacturing information systems including a variety of data such
historical data and alarms, data related to product planning, order dispatch, material management,
production planning, maintenance, sales and marketing, distribution, customer service.
- Equipment data collected from smart factories by Industrial IoT technologies, concerning mainly data
related to real-time performance and operating conditions.
- User data collected from internet sources such as ecommerce platforms and social networking platforms
which may enhance the integration of customer needs and recommendations into the production
preferences.
- Public data collected from governments through open databases which can be used to guarantee the
manufacturing processes and products.
In conventional production system, the data flows are not perfectly conducted in all directions. In other
words, the data flows from the production system control as order signals and commands, another flow
of data is coming back from machines and sensors and collected by the production system. These data
should be processed in real time in order to optimise the ongoing manufacturing processes and adapt the
production system itself depending on the customised production planning first and the mined data base
for optimisation and enhanced performance concerns.
The integrated data management must consider first the local data processing for low level information
including machines, devices and CPS’s especially security measures, machines availability, and the
overall shop floor management. Local and low-level data must be supported by processing modules and
edge computing units if needed. Secondly, the aggregation of higher-level information such as
production planning information and historical data, trends and KPI’s is essential to enhance processes.
Additionally, the wide use of sensors and digitalization by data acquisition provide the analysts and the
decision-making process with information. Using the predictive applications, management can
investigate the expectable events and optimize the business decisions and processes based on it.
Therefore, the integrated data management should allow the storage of this data in the cloud which will
enable its accessibility and processing anytime with low cost and effectiveness.
To sum up, the data management dimension is critical to Industry 4.0 integration. Data must be classified
and well aggregated, transferred and stored, in order to enhance the performance of the modular and
flexible manufacturing systems. And so IT architecture’s evolution must consider the data flows and
187
storage (data lake) assuring a suitable structure for different data types first, and second considering
cyber security measures at all levels.
Referring to modular production systems, modularity is a design principle of Industry 4.0 and one of its
essential functionalities. Answering the client demands, modular production systems give the possibility
to adapt and adjust the individual modules of the production process in a more comfortable and useful
way depending on the production and the product evolution. Modularity is concerned with shifting from
linear manufacturing and planning, rigid systems and inflexible production models toward an agile
system that can adapt to ever-changing circumstances and requirements, without the need for a huge and
sophisticated redevelopment and reprogramming work. Modularity involves the entire production and
manufacturing levels, and builds on agile supply chain, flexible material flow systems, modular
decision-making procedures, and flexible processes (Ghobakhloo, 2018).
The modularity of a production system encloses two main topics: the operational processing modularity
which concerns deeply the shop floor machines and devices along with its control system, and the
manufacturing services modularity which addresses the manufacturing information system. The
operational modularity means that a production system can be defined as a set of modules with an
integrated decision-making level connected by flows, with the function of transforming raw materials
into a product. Modules perform operations manually or automated to transform raw materials, and
flows correspond to the transfer of material from one post to another. The overall flexibility of the
production system stems logically from the flexibility of the modules and flows that compose it.
Operational modularity is not only being able to change the layout of the shop floor simply but also have
a flexible structure that allows the extension of the module to increase production capacity or integrate
new functionalities (Mabkhot et al., 2018). A module is loosely coupled and can be moved, added, or
removed from the system in a plug-and-play manner.
Considering the modular manufacturing information system, since manufacturing systems face external
unpredictable events and internal complexity, the modularity of manufacturing services is imperative.
Traditional centralized and hierarchical control becomes powerless in processing different orders,
unable to respond quickly to unpredictable changes and mined data. In this context, new decentralized
paradigms have emerged, such as multi-agent systems (MAS), which can solve problems that are
difficult or impossible for an individual agent or a monolithic system to solve, driving the industrial firm
to what is called distributed control and information system. Several aspects must be taken into
consideration while addressing the distributed information structure. Embedding intelligent applications
on certain machines allows them to analyse their own data, activate interactions with their nearby
environments, and take local decisions. In addition, the mass of data generated becomes compatible with
188
the use of off-site artificial intelligence applications on the Cloud in order to enhance the production
system. Well, this aspect raises the question of the data generated or imported from any point of view,
storage, cyber security, processing, sharing, and visualization. The last aspect is undoubtedly the most
complex one since it purposes to deploy an adapted network architecture. An adapted network
architecture such as Event-Driven Architecture EDA or Service-Oriented Architecture SOA deployed
to the different entities/manufacturing applications that intervene including MES and ERP, according to
different protocols, in real or deferred time, to the objective of developing a well implemented modular
and distributed system.
The defined goal of Industry 4.0 is the implementation and utilization of an individual interdependency
matrix fitted to every single business process. In this respect, horizontal and vertical integration of IT
systems are required.
Vertical integration is concerned with all hierarchy levels in the enterprise to the management and
business level which comprises the enterprise planning (Bodrow, 2017) and is the main concern of the
189
proposed framework. While the horizontal integration covers the product life cycle from the supplier
until it reaches the customer, this integration should be allowed through the network and taken into
consideration. It involves the design and development of products, production planning, production
rump-up and management, logistics, sales, and distributions (Man and Strandhagen, 2017).
In general, Industry 4.0 is centred on cumulative, real-time, real-world data across an array of
dimensions such as smart warehouse, smart factory, smart product and smart business partners, meaning
that real-time capability is deeply supported by the cross-network. In distributed manufacturing facilities
and across multiple manufacturing sites, Industry 4.0 promotes horizontal integration between
Manufacturing Execution Systems at the factory level. Data from production facilities are transparently
shared between sites. Production operations could be automatically transferred between facilities to
respond quickly and efficiently to production variables. Taking into consideration the security measures,
different actors of the supply chain could be connected to the enterprise, providing a high level of
collaboration in logistics and distribution operations.
190
9. Experimentation, Application on Industrial Platform 4.0
Our approach has been Experimented on a production platform 4.0 installed in the research laboratory
Fig. 28. This platform carries a global research project led by M. Benfriha (Benfriha, 2020) and which
revolves around 2 main axes. our research study falls under the 1st axis, which aims to develop a 4.0
production system model.
Basically, this platform is built from a fixed robot - responsible of managing raw parts between initial
and transition store, as well as transferring the finished parts to the conveyor in their way to the
accumulation table, a mobile robot - responsible of transferring parts and feeding machines, a fixed
camera system for raw material identification and classification, a lathe, a milling machine, and a laser
cutting machine.
At the very beginning, the platform can manufacture, assemble, and transfer components with a closed
and non-flexible automatic control process, resulting in the production of a single product of
polyoxymethylene (POM) shock absorber type composed of 3 elements (piston, spring and body). This
operating mode corresponds to the classic mass production mode subjected to uncontrolled disturbances
close to 3.0 where it produces identical products in small series. After specifying the desired number of
191
shock absorbers, the PLC executes the production plan in a loop in the form of a grafcet deployed with
various machining, transfer and assembly operations. This classic operating mode is far from being
dynamic, given that the PLC does not react to any anomaly except for security incidents configured
through logical functions that can cause the interruption of the cycle. In this old configuration, this
production platform was fairly representative of the way companies operate.
The aim of the experimentation at the platform 4.0 is to deploy an intelligent and flexible manufacturing
process in order to validate our hypothesis “Evolving production systems towards digitization in the
context of Industry 4.0 requires acting on three main axes Production System including Shop
Floor Control System SFCS, Machines & Advanced Techs, and Data Processing & Deployment”,
While integrating several technologies and implementing a new flexible production system.
Following the proposed design framework as shown by the experimental protocol presented in Fig. 29,
the evolution of the production system architecture will be explained below starting from implementing
flexibility to the shop floor and the shop floor control system, along with deploying a modular service
integration at the information system layer. An integrated data management is also assured, as well as
introducing the openness for business level communication and distributed industries. Several
technologies were added to the industrial platform 4.0 such as the FARO 3D Scanner and embedded
cameras on the robots. Once different levels of integration are deployed, the rationale and demonstration
192
of our conceptual model and the transition Framework will be established through the implementation
of a new modular production program. The innovative integrated methodologies will allow the
emergence of new functionalities, such as operational flexibility, customization, dynamic production,
lead-time efficiency, decentralization, and portability. Finally, a demonstrative application will be
carried out in order to validate the dimension of flexibility and lead-time yield of the advanced
manufacturing system.
The design framework proposed above is applied on the industrial platform and enriched with concrete
details by deploying a new architecture. In what follows, the deployment of the framework is explained
layer by layer.
Concerning the machines layer, two level of modularity were developed, machines modularity and
operational modularity. Machines modularity deal with transforming existing machines to CPS entities
and/or flexible machines. Starting from turning the robots into smart and connected CPS by
193
implementing innovative configurations considering an internal data base developed and integrated into
the production system. This data base is connected to trajectories programs which enable the transfer
functions between different part stores and machines. Loop points per area are created to eliminate any
collision between robots and machines / obstacles. These trajectories are well optimized consequently
by adding several parameters: ID part, part diameter, height of the robot’s gripper from the bottom of
the part, speed of motion, and Tool number.
Regarding CNC machines, the fact is that automatic production machines are not well developed to be
connected, intelligent, and open to exploit. To solve these issues, an IoT layer was added first overlaying
the manufacturing machines in order to exploit data and useful information to enhance and supervise
machines working states. Therefore, vibration sensors are added to machines along with temperature
and IoT position sensors depending on the functional requirements of each machine in order to gather
the maximum useful machining information. Secondly, machines control was extended by additional
I/O modules to allow the selection of the desired manufacturing program with parametrization ability,
activated through a main flexible G-code. Remote accessing machine’s data through a USB Networking
Adapter with the CNC enabled the file transfer of the machining programs to CNC machines from the
MES already connected to a CAD/CAM tool. The CNC machining operation is parametrized first by
the ID corresponding to the subroutine program reference, the cutting depth, and the speed of cutting.
Third, a 3D scanner is installed for advanced quality control function. The MES activates the 3D scanner
via OPC UA. The 3D scanner generates a cloud point and compare it with the native design of the
selected product and send back the results of conformity to the MES in a form of report. Adding to this,
another camera system is embedded to the robots and has two main functionalities, a supplement
conformity control in-situ which means parts will be controlled on machines before getting recovered
by the robots and readjusted if needed, and an extra security check to eliminate any collision between
robots vs machines and robots vs parts.
Machines modularity presented below contribute to the operational modularity. The operational
modularity / flexibility implicates a flexible production with a new configuration perspective including
all useful machining scenario enabling the customisation of production processes which will be
explained in details in section 9.4. The production is consequently organised through functional modules
that can be launched in a different order. This will allow us to take advantage of the full potential of the
platform and gain flexibility. Each of these modules is composed form standardized operations and will
be then integrated into the automation control process. To this end, we created several configured
parametric operations that represent all the production operations adapted to the potential of our 4.0
industrial platform:
194
- Machining operations including CNC machining (G-codes for machining processes), and laser cutting
operations.
- Quality control operations including 3D scanner operations for quality control, and Dimensional
quality of parts being manufactured using robot embedded camera.
- Inspection operations using fixed and embedded camera including security checks, storage and
anomaly controls.
To conclude, having a flexible and adaptive production system is essential for our scientific experiments
in terms of flexibility. To achieve this goal, several changes would affect the conventional organisation
of the shop floor, mainly its configuration perspective at both the operational and devices level. Several
advanced technologies and options introduced by I4.0 Fig. 30 have been deployed such as connected
machines, 3D scanner, intelligent embedded camera systems and IoT sensors, along with deploying
modular standardized parametric operations, a new modular planning approach compatible with the
desired flexibility in the context of “Industry 4.0”.
In order to deploy the operational flexibility created at the devices / machines layer, the shop floor
control system must support the introduced flexibility and modularity. In order to overcome the control
limitation for an improved manufacturing system, the approach we propose is similar to the call up
method by deporting the control of automation logic functions of the API layer to the upper layer of the
195
CIM pyramid, so that they are accessible to users of the workshop. For this purpose, logic functions
were adapted to be used without requiring the intervention of an automation redevelopment work. We
are no longer using master grafcets to manage the commands of the operations, it is the Manufacturing
Execution System MES (supervision & execution) that will perform this function. Concretely we
generate small parametric production sequences that the engineer can use separately as standard
operations or assembled to form functional modules in order to build a complete production plan. These
standardized operations and modules activate the logical functions of the controller, keeping this way a
clean path by using the standardized automation protocols to controlling shop floor machines.
Well, this solution concerning the control system is not unique. First of all, a suitable workshop control
system must take into account the standard automation requirements concerning the control and
monitoring of industrial systems: ISA-88, 95,106 standards, providing direct control via small
microcontrollers or standard PLCs. The use of standard PLCs or microcontrollers will be provided by
the deployment of logic functions or functional blocks. Subsequently, and with the goal of establishing
operational flexibility, a remote scheduling function to an orchestrator will help free up and expand the
decision-making process to configure a flexible production plan for a product. Consequently, the
production plan execution process can be dynamic and optimizable thanks to the flexibility implemented
at the control level. Finally, the possibility of integrating AI for small tasks through intelligent control
modules is now possible thanks to new low-level intelligent automatic modules recently present in
industrial markets.
Focusing on the modular manufacturing information system, the main objective is to deploy a service-
oriented manufacturing system capable of providing manufacturing functionalities such as production
planning, production management, production optimisation and supervision. This modular information
system must host manufacturing service applications and support the CPS modules integration by
solving interoperability issues. Accordingly, this structure will enable a decentralized decision-making
level distributed through the different modules of the system and allowing the integration of a distributed
architecture at the site level.
The development has focused on adding a flexible layer into this production line by enabling it to
produce several products and following different manufacturing cycle in a rapid configuration. An
advanced distributed and service oriented MES is deployed to the platform, where alarms, trends and
all data from machines and sensors are collected. The production system is therefore composed of
different applications responsible each of a precise function including historization and trends,
communication integration server, production planning, AI data processing and AI image processing.
These applications are not necessarily hosted by the MES but also internally developed applications
196
especially the AI processing modules which are connected to the production system and can provide it
with the interpreted information. An example of the integrated modules is the smart visual inspection
system. An inspection camera was added to the production platform and controlled via a machine
learning algorithm for human detection (El Zant et al., 2021). When a human is detected approaching
machines, the camera system sends a notification to the MES to pause the production and ensure security
of operators. This module is loosely coupled to the production system and can be used for several
applications and generate opportunities to develop different parallel modules for data and image
processing and integrate them into the manufacturing system in order to enhance and well optimize the
whole production process. Two additional modules integrated to the production system will be presented
in section 9.6 along with the respective architectures. The interoperability issue was solved thanks to the
communication integration server of the MES which allow a wide variety of communication protocols
through communication drivers. The PLC for example is communicating with MES via Sidirect
operations integration server, and the 3D scanner is communicating through OPC UA.
Finally, the control in real time of the operations and the scheduling are done through a web application
HMI developed through the distributed MES. The objective is to provide an interface which
communicates with robot’s programs and machines via their identifier, and to orchestrate control and
monitoring services in order to organize the production processes in the desired order as well as
supervising and monitoring the production facility.
As discussed before, data management including data aggregation and data mining is an essential
objective in the digital transformation of the production systems. To do so, the shop floor layer has been
superposed with an IoT sensors layer driving to exploit potential data and information from machines
such as vibration sensors for predictive maintenance, power consumption for energetic optimisation of
the production processes, temperature sensors for safety assurance and security matters. Some of the
collected data are crucial for the production planning and management thus addressed to the
management level to enrich the production system. Other low data are processed locally on a designated
CPS assuring machine to machine communication and allowing for a local decision-making process
regarding flow management. As an example, advanced robots at the platform are playing the role of
managing transition stores and on-going product transfer at the shop floor locally using an integrated
data base, which is shared as well with the MES layer for products tracking. Data from shop floor control
system such as availability of machines are well considered and shared through the network with the
management level since the production system will be in charge of managing the process flow and the
overall production planning. Control data are critical equally to enhance the security of the production
processes.
197
Technical wise, two TCP/IP networks were implemented. The first one is considering the management
and business level, connecting the MES with other application modules including Historian, the HMI,
and a portable HMI via 4G, as well as a bypass to the cloud. The second network represents the devices
field bus where two flow of data can be defined, control and monitoring data for processing through
controllers’ routes, and data transfer flow using different communication protocols collected from
machines, sensors and actuators. Management level will provide the production system with valuable
data especially historical data, trends and KPI’s, but also designed to gather raw data and exploited data
for optimisation objectives. Alongside, data mining service modules are connected directly to the
production system providing it with treated data, such as smart inspection camera modules. In addition,
data management include the development and the implementation of cloud platforms where data can
be stored and processed in parallel to the local machines. Cloud platform can provide cloud computing
for raw data and learning new patterns to enrich the local AI algorithms. Finally, the production system
is well connected to the business level through the ERP platform once deployed, where data can be
shared in two directions in a well-defined perimeter taking into consideration cyber-security matters.
Taking into consideration the integrated data management dimension allowed to make use of data in
organising and monitoring the production system and will contribute to the performance and the
optimisation of the system since the data flows are assured in both ascendant and descendent directions.
Additionally, the integrated data management helps defining the data lake of the platform and
aggregating different types of data including low level data (raw data from machines including machines
availability), sensors and IoT data, unstructured data such as images for image processing, point cloud
data for quality control, and processing data including machining programs. The final design of data
flows will be presented in detail in section 9.7.
The integration of MES systems with ERP systems or other business applications has become a critical
mission for manufacturing operations. In particular, this provides flexibility in production planning. It
also improves supply chain visibility and decision support by updating business applications with real-
time production information.
The production system digital platform integrated to the production system, provides the ability to
exchange messages and enable automatic data synchronization between multiple workshop applications
and corporate workshops. This enables manufacturing and business systems to communicate in an easier
and more secure manner. This includes integration and exchange capabilities, including XML messages,
flat files, SQL queries that allow the integration between multiple workshop applications and ERP or
other business systems. Extensions allow through custom plug-ins to extend the connectivity of the ERP
with any system or to standardize the use of the product in a multi-site scenario. The integration features
198
of such distributed ERP include high integrity of data exchanges, flexible message exchanges and native
integration with MES. Users can also benefit from a complete history of message exchanges, Store &
Forward functionality, extensions for personalized management of data transformation as well as web
interfaces.
Based on the vertical integration implemented through the proposed transition framework, the system
architecture has been upgraded to an advanced level in terms of flexibility and autonomy. The upgraded
version of the system architecture is established and supported by the IT infrastructure of the platform
assured by the new IT architecture deployed. The new IT architecture conducted to assure the modularity
of the system and to solve the interoperability constraints. Consequently, thanks to the modular
information system and the flexibility assured at the shop floor level, a new configuration of the
decision-making level of different machines and modules is subsequently enabled assuring the
decentralisation functionality of the production system.
The production system architecture as well as the IT architecture of the platform will be represented in
the next sections. Finally, and before discussing the emergence of the production system configuration
in section 9.3, the decision-making level model-based machines is presented in 9.2.3, introducing the
decentralisation assured through the new system architecture.
As shown in Fig. 31, the system architecture has clearly evolved. Machines layer composed from lathe,
milling, and cutting machines in addition to the robots have a direct control communication with the
PLC / API layer. This communication is established through I/O, Profinet and other industrial
communication protocols for low level data transfer and automatic steering controls. Data shared
between machines and the PLC are directed to the MES through a communication established between
the MES and the PLC in Sidirect driver for Siemens Profinet, thanks to the integration server of the
MES. From another side, the IoT layer superposed to the machines including accelerometer sensors
from Intellinova on the lathe, power consumption sensors from Socomec installed on the machines
power source, and Sick IoT box connected to position and temperature sensors on the laser cutting
machines are all communicating directly to the MES through Modbus communication protocol. The 3D
Scanner from FARO installed to the processing machine is communicating through OPC UA
communication protocol to the MES on the same network. Embedded cameras Sensorpart are connected
to the local network and the robots via Profinet, while the fixed camera system from Cognex is integrated
to the robots system. Another data transfer communication is established directly though the network
between the MES and the Machines for Gcode programs sharing. Gcode programs generated by the
CAD/CAM tool from SprutCam are accessed by the MES and shared through the network in order to
199
update the active programs on the CNC machines. An inspection camera from IDS is connected to the
system platform on the network. The inspection system is installed on the processing machine and
communicating with the MES through OPC UA.
The MES presented in this architecture represents several service-oriented applications including the
Integrated Development Environment IDE of Aveva System Platform, the integration server, the HMI
developed through Intouch OMI, and the historian. A DNS server is installed to the main server in order
to activate the Intouch Access Anywhere application playing the role of a deported HMI application for
portable use through another 4G network separated from the local one. The simulated digital twin
application is integrated to the production system and will be presented in section 9.6 along with the
smart visual inspection system. Finally, the cloud computing module is not already deployed but enabled
first through Azure and Aveva Insight application connected to Aveva System Platform through the 4G
network and enabled as well on the processing machine on the same network if needed. The
configuration of the cloud computing platform will be conducted by the PhD thesis of my colleague EL
HELOU Marwan, working on the integration of Cloud Computing in the manufacturing systems.
In order to clarify well the system architecture, the IT architecture implemented to the platform is
presented in Fig. 32. The IT architecture is organised through two main networks, the low-level network
and the high-level network. The low-level network is a TCP/IP one, connecting machines layer, control
layer, and the management layer presented by the main server as a field bus network, where different
200
communication protocols are deployed as required for each entity. The second network is connecting
the management layer composed from different machines/PCs adapted for different use depending on
the implemented application role, and enabling the connection to the cloud and the portable HMI
application. This network is connected to the business layer as well represented by the ERP system
which is not deployed for the moments since we focus on the production system dimension.
The system architecture and the information system configuration as well as the modularity assured on
the machines layer, enables an enhanced level of decentralisation in terms of decision-making level
concerning production processing and information modules. Fig. 33 presents through a model-based
machines diagram, the decision-making level of each entity of the system. In our configuration, the MES
is playing the role of operation manager, responsible of process configuration, process planning and
system monitoring. While robots are responsible of managing different stores regarding placing
availability and transfer. PLC system is responsible for steering controls, but embedded cameras are
responsible of confirming the well execution of the orders such as doors opening of machines.
201
Processing modules implemented to the processing machines such as smart inspection system and
quality control are performing actions at the processing machines level and only supervised by the MES
which collects in his turn the processed information.
9.3.4 Synthesis
The framework based on the proposed conceptual model helps to consider the different layers of a
company's production system and recommends new advanced modifications to ensure a robust evolution
towards the digitization of the manufacturing system. This framework, which has been applied to the
existing traditional architecture of the platform, has contributed to evolving the architecture of the
202
manufacturing system, in particular by ensuring the functionalities set at the beginning of this chapter,
flexibility, customization, lead-time performance, modularity, decentralization, and portability.
The flexibility of operations and production processes has been well mentioned in sections 9.2.1 and
9.2.2. The new flexible shop floor configuration consisting of turning machines into CPS will have the
main effect on operational flexibility, followed by the equally flexible new configuration of the SFCS
of the control layer. The flexibility of the system will be well developed and demonstrated in the
following two sections by deploying a new configuration of the production program. Finally, the
demonstrative application presented will be carried out mainly to justify the flexibility dimension of the
evolved system. In addition, the application is meant to validate the lead-time yield, particularly in the
industrialization phase which concerns the development of a new production plan for a product. The
simplicity of defining a new production plan and the remarkable time savings in carrying out this task
are also due to the evolution of the system architecture as well as to the new configuration of the
production program implemented subsequently. The modularity of the information system explained in
section 9.2.4 is due to the integration of modular services following the deployment of the digital SOA
platform and the integration of image processing modules. The portability of the HMI is one of those
information modules implemented in the system in the form of an Intouch Access Anywhere modular
application which deport the HMI via the 4G network. The data dimension is configured by integrating
the vibration sensors installed on the lathe, collecting generated data in order to be processed by
algorithmic modules to be developed, taking into account the level of integrated data management.
Finally, the decentralization appearing in the distribution of decision-making levels between the
different entities of the production system was well presented in the previous section.
The deployment of the Framework within the vertical integration dimension and the evolution of the
architecture of the production system has led us to deploy a new modular configuration of the production
program. The contribution of this section is the implementation of a new advanced modular SFCS
approach where the control process is upgraded to the MES software. This new approach impacts the
classic Computer Integrated Manufacturing (CIM) pyramid, including shifting decision making and
scheduling control to the MES layer rather than being stacked in the Programmable Logic Controller
(PLC) at the control layer. This novel architecture is enabled thanks to the use of new advanced industrial
and production process SOA software Wonderware System Platform WSP as the MES (“Wonderware
System Platform - Fondation de l’Industrie du Futur,” 2017), as well as the integration of new
technologies such as advanced robots, a 3D scanner for quality control, and the integration of Iot sensors
and embedded camera systems. This modular approach aims to unleash the potential of the new system
architecture without changing the entire existing control process but moving the decision-making
process to the MES layer through the implementation of an advanced and interoperable digital platform.
203
This modular approach is meant to overcome these limitations without changing the whole existing
control process but by moving the decision-making process to the MES layer thanks to the new,
advanced and interoperable MES.
The potential of MES is in the capability of easily creating production programs using allowed standard
programming languages to develop complex scripts, which will be very difficult to develop on logic
controllers or even impossible sometimes. The functionalities targeted by the deployment of the MES
are the production monitoring including the execution of manufacturing operations, and the production
management including the management of modular production plans, by responding to customer quality
and quantity requirements, and the functional needs of the system in terms of maintenance and flow
management. This makes it possible to go further in the optimization of production processes, thus
facilitating the development for operators and production engineers without having to make continuous
changes at the level of logic controllers and the device layer of the system workshop. We can identify
two main use cases with this new modular process:
- For the industrialization of a new product: the capacity to reduce time to market for the production
of a new product by reusing existing modules.
- For the production scheduling: the capacity to launch a new production with no lost time and to
switch easily from one routing to another.
In this section, the modular process methodology is presented in 4.1, the evolution of our SFCS is
clarified in 4.2, followed by 4.3, where the modular process design patterns are discussed.
204
Fig. 34 Modular Process Methodology
The "Standardized Parametrized Operations" phase is reserved for the definition of basic production
operations for a specific product, encapsulated to the MES and parametrized. The implementation of
basic operations in MES is organized in two stages. The first takes place on machine-specific systems,
such as CAM tools dedicated to the generation of tool paths for CNC machines. Indeed, all the operations
necessary for a product production plan are validated upstream. Then comes the second step, which
consists of preparing their integration into MES through an encapsulation action in order to make them
compatible with its operational environment. This process is based on advanced technical knowledge of
the production line and obeys specific codification in order to be identifiable in different interfaces.
Production operations of the “standardized robot trajectory” type are interesting illustrative examples to
mention. In our platform 4.0, we have analysed all the possible trajectories of the robots and we have
succeeded in defining 33 trajectory that can be used in a combined way and thus cover all the needs in
terms of transfer in the workspaces of the platform 4.0. This approach helps to increase the flexibility
of operations and thus facilitates the design of a production plan by assembling standardized operations.
The "Functional Modules" or "industrialization" refers to the preparation of modules, where each
module can group together, in a precise schedule, several operations from the previous phase. These
modules can constitute a sequence of logical production operations dedicated for a specific product. For
various reasons, any module must be able to be simulated and executed. For example, the engineer can
develop a specific quality control module M1 which consists of using the Mobile Robot RM to transfer
the component and drop it to the rotating plate of the 3D scanner through RM_Drop_Scanner operation,
and activate the scanner using Scan_ProgX which recovers the point cloud and compares it to the native
3D model of the part X. Another module M2 will be created to pick the controlled part and transfer it to
the Transition Store (TS) using respectively RM_Pick_Scanner and RM_Drop_TS operations Fig. 35.
This phase allows production engineers to prepare modules by overcoming the issues of interoperability,
robot programming, scanner control, etc. which provides a real advantage in productivity.
205
Fig. 35 Functional Modules Examples
The 3rd phase called "Orchestration of a Product Production Plan" is dedicated to finalizing a production
plan for a given product. This phase consists of orchestrating modules to generate the production flow
of a product, defining anteriority constraints between modules, and setting up quality control rate for the
product. The production engineer collects all the modules necessary for the execution of a given
production order and imposes the constraints of inter-module anticipation. Some modules are
configured, in particular the quality control modules, which must be activated in proportion to the
number of components manufactured. In this phase, the MES should offer several options, the first
consists of taking into account the constraints imposed between modules and proposing all possible
scenarios for scheduling production modules for a defined product. Finally, the 4th phase "Production
Process Planning" or "piloting" is a logical continuation of the previous phase. The objective here is to
prepare launching the production and to ensure it is managed in real time.
Actually, in the case of fully automated production, one of the flexibility locks is located at the controller
level. Indeed, once the process is frozen in the Application Programming Interface (API) layer, only
reprogramming the controllers allows the workshop to be used for another task than the active /
configured one. For each new product, it will be necessary to create and redevelop a new program,
allowing to produce another product rather than the old one. This situation represents the conventional
industrial architecture type 3.0. In order to overcome this limitation for an improved manufacturing
system, the approach we propose is to deport the control of automation logic functions of the API layer
to the upper layer of the CIM pyramid, so that they are accessible to users of the workshop using the
call-up technique. For this purpose, logic functions should be adapted to be used without requiring the
intervention of an automation redevelopment action. Master grafcets are no longer used to manage the
operations, it is the Manufacturing Execution System MES (supervision & execution) that will perform
this function.
Concretely we generate small parametric production sequences that the engineer can use separately as
standard operations or assembled to form functional modules in order to build a complete production
plan. These standardized operations and modules activate the logical functions of the controller, keeping
206
this way a clean control path by using the standardized automation protocols to controlling machines.
Fig. 36 presents the proposed command architecture:
(Weyer et al., 2015) identified the issues of the design of a heterogeneous production line by new control
architectures, new engineering and programming paradigms, communication standards and IT security
challenging the digital evolution of the production systems. Three paradigms were identified by (Weyer
et al., 2015) to specify central aspect of Industry 4.0 concerning the production systems Fig. 37. Starting
from the smart product in the meaning of product-oriented production configuration, where the product
embedded its own operational data and requirements to be processed in a modular production system.
Secondly, machines getting smart when being transformed into Cyber Physical Production Systems
(CPPS) enabling its self-organization, decentralization of the decision making at the shop floor level
and allowing an easy plug-and-play integration. Lastly, the augmented operator considering the smart
working environment as a requirement for technicians as for production engineers technological support
when dealing with highly flexible and modular production systems.
207
Fig. 37 Modular Process Design Patterns (Weyer et al., 2015)
The first dimension ‘Smart Product’ is considered in our configuration in a holistic way since we are
using polyethylene plastic material without an RFID tag nor an embedded memory but categorized by
ID series lot, identified by the fixed camera system and an ID number given by the MES for the part.
The MES itself manage the production processing modules in our case rather than embedding the
operational information into the product. Hence, this identification is used first to recognize the raw part
for a specific product based on its ID series and its dimension (diameter), second to trace and track the
part in the production processes and its place ID in different stores using its ID number, third to build a
processing information data base of the processed parts on the robots bay shared between robots and the
MES and updated instantly using internal robots functions taking into consideration loading / unloading
/ transferring parts as well as machining data. This data base serves in addition to assure the self-
organization and collaboration between fixed and mobile robots in order to manage the transition stores
as it serves as well in securing the robots vs parts loaded in the stores and the machines.
The smart machines as the second dimension is our essential concern since the interoperability and the
compatibility issues were not at all considered when designing the line with its basic conventional
configuration ‘automatic mass production’. Starting from turning the robots into smart and connected
CPPS by implementing innovative configurations considering the internal data base developed and
integrated in the production system, and its connection to the trajectories programs enabling the transfer
functions between different part stores and machines by creating loop points per area to eliminate any
collision between robots and machines / obstacles. These trajectories are well optimized consequently
by adding the speed of the robots’ motion to the transfer operations in order to recompensate the time
loss when following the loop points as starting and ending points in each trajectory.
208
Regarding CNC machines, other initiatives were implemented concerning machining flexibility. An
access issue was crucial avoiding remote controlling the machine in updating machining programs,
activating the program, and launching the activated one. First, the G-code transfer problem was resolved
using a USB Networking Adapter with the CNC controller since the controller supports network access
for file transfer – a fast and convenient method to transfer G-code programs directly to the controller
from the MES where the activated machining programs list is set. The MES in his turn is connected to
a CAD/CAM software where the machining programs are generated. Second, the activation issue is
resolved by creating a main conditional G-code program to be activated when initializing the line, which
in his turn call for each machining program of different parts as a subroutine stored in the subroutines
file shared with the MES. To do so, machines controller was extended by additional I/O modules to
allow remote access to machines and activating several machining sub-programs with parametrization
ability such as cutting depth and cutting speed. Additionally, the fact is that automatic production
machines are not well developed to be connected, intelligent, and open to exploit. To solve these issues,
an IoT layer was added overlaying the manufacturing machines in order to exploit data and useful
information to enhance and supervise machines working states. Therefore, vibration sensors are added
to the machines along with temperature and position IoT sensors depending on the functional
requirements of each machine in order to discover and gather the maximum useful machining
information. Concerning the quality control task, a 3D scanner is installed for advanced quality control
function. Adding to this, another camera system is embedded to the robots and has two main
functionalities, a supplement conformity control in-situ, which means parts will be controlled on
machines before getting recovered by the robots and readjusted if needed, and an extra security check
of robots vs machines and robots vs parts in order to avoid any accident due to a wrong signal or a
manual manipulation.
Furthermore, this flexibility implemented to the machines at the shop floor is reinforced and assured by
the new configuration of the control system layer as presented in the section 9.2.2 discussing the shop
floor control system configuration which enables not necessarily the physical integration but simply the
plug and play modules integration in a form of standard programmed operations.
Lastly, the augmented operator dimension is considered clearly at the MES stage where an ergonomic
production-oriented HMI is developed using Intouch OMI, enabling the production engineer to develop
the product production plan by assembling core functional modules connected by flows and supporting
the modular process configuration. The production engineer when receiving the order for a new product,
starts from creating the correspondent machining programs using the CAD/CAM developer in case of
not having the samples in the library, and imports them directly to the MES. He builds a list of all
machining and transfer parametrized programs at the operations stage, before creating functional
modules composed of one or more basic operation in order to transform this group of operations into a
209
meaningful process such as recuperate a part from the transition store by the mobile robot, load it to the
lathe machine and run a specified machining program. The step forward is to arrange these modules and
connect them by flows with the ability of adding anteriority constraint between two consecutive modules
if needed along with setting the quality control rate corresponding to the percentage of controlling this
product’s lot when finished using the 3D scanner. Finally, the production engineer creates the list of
products for the daily production planning. This organisation helps the production engineer and the
operators to deal with the new modular configuration, saves a plenty time of the industrialization phase
integrating a new product and a variety of products, and also simplifies the industrialization phase
avoiding the production engineer from re-engineering processes, reprogramming machines, and
reconfiguring the control system any time a new product processing is required.
The platform is controlled through Wonderware System Platform WSP - Schneider product
(“Wonderware System Platform - Fondation de l’Industrie du Futur,” 2017), where alarms, trends and
all data from machines and sensors are collected Fig. 38. The production line, thanks to developing this
flexible layer, is able to produce different products and to drive different manufacturing plan.
210
9.5.1 Standardized Parametric Operations
Due to the shop floor flexible machines configuration, the development of the standardized parametrized
operations is enabled. Several configured parametric operations were created representing all the
production operations adapted to the potential of our 4.0 industrial platform. The parametric dimension
expresses a variability that allows the operation to adapt to the context, in particular to the variability of
the components manufactured. The operations can be defined in different ways, it all depends on the
standardization sought. Corresponding to our platform, we have chosen to group the operations into 4
categories, summarized in the first column of Table. 4. The interest here is to identify all the operations
necessary for manufacturing, controls, transfers, assembly, etc. which allow the production of a
component to be ensured, starting from the initial stock until the delivery of the finished product and the
controlled one where it could be recovered from the accumulation table. In other words, we are
organizing production processes into multi-task production operations as follow:
- Machining operations including CNC machining operations (G-codes for machining processes), and
laser cutting operations.
- Quality control operations including 3D scanner operations for quality control, and Dimensional
quality of parts being manufactured using robot embedded camera. The main dimensional and geometric
quality control operation is conducted through the 3D scanner which compares the point cloud of the
part being controlled to its native model. A dimensional quality control operation is performed by the
embedded camera on the mobile robot in situ for parts being manufactured before recovering. This
operation helps detecting deviations in order to readjust the manufactured part if possible before
recovering.
- Monitoring operations using fixed and embedded camera including security checks, storage and
anomaly controls. Fixed camera system above the initial and the transition store assists in the storage
control and management of raw and manufactured parts. The supervision camera helps monitoring the
production line by detecting any human presence nearby robots to alert the operator and pause the
machines. Finally, supplement security checks are performed in addition through the embedded cameras
to the robots to assure the availability of a place to drop a part as well as the opening of machines door
to avoid any collision.
Transfer Robotic transfer between the transition store and the machines Kuka Agilus 1100
211
Robotic transfer between machines and the 3D Scanner Kuka Agilus 1100
Robotic transfer between initial and transition store Kuka Agilus 900
Milling Tormach
Universal Laser
Engraving - ULS
System
Supervision
Monitoring Anomaly control
Camera
Certain operations may require specific configuration, in particular with regard to the flexibility of the
platform. For example, the robot’s gripper in terms of stroke and shape, has to handle different
workpiece dimension. Indeed, the typology of the different components that will be manufactured
suggests that we maintain the "Gripper" aspect as an input parameter to transfer programs. The robot
trajectories for transfer and machines loading are created as trajectory entities that can be assembled and
used in any production process. For this, we have defined areas within the platform, which will be the
starting and ending points of the robot trajectories. These points are called Loop Points and will be used
at the beginning and the end of each trajectory and will ensure that each trajectory can be performed
before or after any other trajectory. Indeed, the paths between each of these loop point have been tested
to avoid any risk of collision with the equipment of the cell. Some additional trajectories dedicated to
specific functions or security matters have also been added including basic security control sequences,
such as controlling the opening of the doors of the machines to which they are heading. The Chosen
trajectories for programming correspond to every possible path connecting different zones Fig. 39 and
each machine.
212
Fig. 39 Industrial Platform 4.0 Zones
Table. 5 illustrates in the second column an example of standardized production operations that can be
requested in different production programs. The last column in the table presents some parameters
dedicated for the correspondent operations. A robot trajectory is parametrized by the part ID, the
diameter of the part, the height of the robot’s gripper from the bottom of the part, and the robot’s speed.
The CNC machining operation is parametrized first by the ID corresponding to the subroutine program
reference, the cutting depth, and the speed of cutting.
Robot K900 W_RF_Pick_InitStore Pick the part from the Initial Store ID, Ø, H, S
Robot K900 W_RF_Drop_TransStrore Drop the part in the Transition Store ID, Ø, H, S
Robot K1100 W_RM_Pick_TransStore Pick the Part from the Transition Store ID, Ø, H, S
Embedded ID, Ø, H, S
Camera W_RM_Control_Tour Presence check of the part in the Lathe
Robot K1100 W_RM_Pick_Tour Pick the part from the Lathe ID, Ø, H, S
213
Robot K1100 W_RM_Drop_Scanner Drop the part to the 3D Scanner ID, Ø, H, S
Frequency
°
Scanner FARO Prog_Frao_1 Start Scanning (referenced part program) /°
Robot K1100 W_RM_Drop_TransStore Drop the part to the Transition Store ID, Ø, H, S
Robot K900 W_RF_Pick_TransStore Pick the part from the Transition Store ID, Ø, H, S
Robot K900 W_RF_Drop_Conv Drop the part to the final conveyor ID, Ø, H, S
9.5.2 Modules
In fact, the MES transform basic production operations from the interconnected machines, robots,
scanners, PLC’s, actuators and sensors, into modules which encapsulate machine programs but also
information and commands linked to its operational environment. Several strategies can intervene in the
definition of modules from the operations defined in the previous step. The first strategy is to define
modules that group together a minimum of operations under an identification that can be understood by
the different operator involved in development and production. The advantages are multiple, in real
time, this strategy increases the flexibility of production, in particular by its ability to reorganize the
production plan in order to avoid the occurrence of undesirable events on the platform. On the other
hand, in deferred time, this strategy offers the possibility of enhanced optimization in the scheduling
phase of production modules. The second strategy is to define modules grouping together a sequence of
operations such as transfer operation in order to facilitate visibility of the production plan, but may result
in increasing the number of modules to respond to different tasks. The interoperation prior constraints
make it easier to group them into a module; therefore, the parameterization of certain operations is
transferred to the module which returns the value of the parameter to the operation concerned in the
production phase.
It can be seen in Table 6 that certain modules group together a single operation, while other modules
group more than one operation for various reasons. We distinguish three main categories of the modules.
First category considers transfer modules, which can regroup one or more transfer operation concerned
by robots. Transfer modules such as the first and the 4th modules listed in table 6, are concerned with
transferring parts between machines and different stores as well as assembly functions. Second type
considers processing modules concerned with machining operations, quality, and monitoring control.
These modules are meant to be single and specific to the machining of a precise component.
Furthermore, processing modules such as the "Lathe machining - Pawn" module, are meant to be
independent of transfer modules in terms of functionality in order to enhance the system flexibility and
214
liberate robots to perform other tasks. Other processing modules have percentage settings set to define
the frequency of the module request for a product family such as presence control module resented in
table 6. Last category considers the personalised modules. This category stands for enabling the operator
to develop a personalised module by regrouping several transfer and processing operations for specific
functional requirements.
The possibility to create such separated modules, thanks to the developed modular process, provides the
manufacturing system with a considerable level of flexibility since multiple modules could be launched
in parallel depending on the machines availability first and the queuing management generated by the
manufacturing system based on the optimisation criterion (time, power consumption, machines priority,
…). The development work of these modules requires multidisciplinary skills in automation, robotics,
and machining, but it has a great added value for the following production engineering work because
when it is done, no more important development will be required.
Machining lathe -
Prog_Tour_1 Start the machining program Lathe
Pawn
The 3rd phase of this section is the scheduling of production modules to generate a digital production
plan and therefore orchestrate the created functional production modules. We distinguish in our
approach 3 types of production plans. The first called "initial" which takes into account the precedent
constraints in-between modules. This initial plan is not well optimised as it stands, it only gives a global
overview of the initial production plan since it activates the defined modules sequentially. The second
version of the production plan, called "dynamic", which will be subject to the capacity constraints of the
215
platform and the machines availability. Consequently, the MES will be able to manage production by
executing several production plans in parallel while managing flow management. The MES system will
be able to plan its execution taking into account the current production load by managing the production
flows (Queuing Management). The 3rd version of the general production plan called “optimized”
consists of optimizing the machining and transfer operations in terms of time, cost or energy, based on
the processing of data collected from the various sensors and IoT superimposed on the machines. The
criteria which can be used are cost, time, or energy. The optimization can be mono-objective, for
example producing a production plan which minimizes energy consumption, or multi-objective, when
we seek for example to identify a production plan which offers the best time of production at the lowest
cost.
Fig. 40 illustrates a production plan for a Pawn organized into several possible flow scenarios. The first
flow transfers the machined part from the lathe chuck to the 3d scanner, activates the scanner to measure
dimensional deviations, then transfers it from the scanner to the conveyor. The second flow transfers the
machined part directly from the lathe chuck to the conveyor without going through the 3d scanner. These
alternative flows are conditioned by the frequency of dimensional control defined for this specific
product. This parameter, controlled by the MES in the "production control" phase, can be fixed or varied
depending on the evolution of dimensional deviations observed over time. The historization of
dimensional deviations of machined parts from their CAD model is a real source of optimization to
maintain production at an optimal rate. Indeed, the dynamic quality control makes it possible on the one
hand to limit interruptions for adjustment and on the other hand to limit the number of parts rejected for
dimensional and geometric non-conformities.
216
Fig. 40 Production Plan Example for A Chess Pawn
Finally, the control in real time of the operations and the scheduling are done through an HMI developed
on Wonderware System Platform WSP. WSP software applications offer enhanced integration and
provide a common and strategic industrial application services platform based ArchestrA real-time
service-oriented architecture SOA technology. This integration provides a development environment for
MES, SCADA, HMI, Historization, and other manufacturing service applications using a single, unified
SOA software platform. These new unified software solutions are designed to help manufacturers reduce
inventory costs and improve production order lead time for rapid responses to changes in demand, as
well as increase the capacity of existing assets through improvements in asset utilization (“Wonderware
enhances Operations and Performance Software,” 2008). Fig. 41 presents the experimental HMI version
to test the reliability of the production system. The tab presented in figure 41 is dedicated to the
orchestration of the modules composing the production plan of the product.
217
Fig. 41 Production HMI for the Industrial Platform 4.0
The production plans, selected and validated in the previous phase, will be loaded and positioned in
priority execution order. The MES manages the synchronization of production plans with the capacity
load of the platform. In our approach, synchronization is carried out at the modules level, bringing the
operational flexibility to modular flexibility. In its preparation phase, the MES takes into account the
load plan of each machine, scheduled shutdowns, stocks, initial scheduling, the availability of reference
documents, the constitution of batches, etc. In its active phase, it takes into account the real-time
balancing of flows, failures, measured deviations, contingencies, traceability and batch release, etc. To
this end, the MES developed here will be endowed with a level of autonomy and must therefore be able
to make decisions based on ascending data almost in real time. Thus, the execution of modules becomes
dynamic and its updating depends on the one hand on the evolution of the situation in real time and on
the other hand on the hazards encountered.
As mentioned in the section 9.2.4 Modular Service Integration, two modules were integrated to the
system in terms of demonstration of the deployed modular information system. We are focusing in this
section on the integration process of two developed modules, along with explaining their functionalities
218
briefly since the development work will be conducted and continued by the PhD thesis of my colleague
CHARRIER Quentin, working on the digitalization of the production processes.
The system given below in Figure 42, presents the contribution of an integrated module which figures
in the successful implementation of a Machine Learning ML algorithm into an industrial production
system, and the effectiveness of using ML to assure in real time visual inspection and rapid decision-
making process. Thanks to the interoperability and modularity assured by the MES and the system
architecture, the intelligence and the self-adoption of the manufacturing system is possible by adding
smart CPSs and smart modules using machine learning and deep learning algorithms and techniques.
The ML system runs on the processing machine and reports to the MES the processed data through the
PLC using snap7 communication protocol or directly to the MES using an OPC UA communication
server. The MES takes into consideration the received signals and information and acts properly by
modifying the initial program and upgrading it as it should.
219
This smart system could play a role in being an edge computing module, which reports unexpected and
untrained data to the cloud in order to train the model for continuous data acquisition and analysis. This
model can be used and developed in our ongoing project for several data processing tasks, providing
valuable information regarding the predictive maintenance task, the quality control process, and the
power consumption optimization.
To test our model, we chose a security application, since we are using two robots in our line Figure 43,
we should not have any human presence inside the production line when it’s functioning to avoid any
accidents. So, to ensure security, the platform has a physical barrier and an access door that must be
closed, preventing anyone from entering the interior when it is operating. However, if the door is closed
and a person was present in the production facility or comes in through another unsecured access, no
error is reported, and the person can be then in danger. To overcome this problem, we decided to conduct
a smart visual inspection system which will detect the presence of the human within the production line
using video streams.
If a person is detected, the information will then be forwarded to the MES which will pause or reduce
the speed of the robots and which will alert the operator who will be able to raise the alert after
confirmation that no one is present in the line. The video streams are then fed back to this algorithm
which will carry out the detection, and in case of detection of someone present will send the information
back to the MES which will pass the line in safety mode Figure 44 and will warn an operator who will
have to raise this alert once the risk has passed.
220
Fig. 44 Program Paused, Person Detection Inside the Platform
A digital twin application module has been developed as well using Unity3D and integrated to the
system. Thanks to the potential of the advanced simulation technologies and tools, 3D modelling of
robots and machines were developed through KUKASim and exported to the Unity3D project. The
digital twin application was configured in two modes: Monitoring mode and Simulation mode.
The monitoring mode presented in Fig. 45 below, is a loosely coupled application module integrated to
the system. The monitoring application collected real data from the WSP via its OPC UA server and
visualize machines status and virtual functioning as well as products tracing and evolution following the
production processes. The data flow via the WSP is bidirectional, orders can be sent back to the
production system using the digital twin monitoring application and the production program can be
launched from it. Other direct data exchange flows were superposed between robots and the digital twin
application using WCF communication server for real time visualization of the robots movement
evolution (Axis coordinates), facilitating as well the sharing of internal data base developed on the
robots. This module helped us first experiencing different communication protocols in terms of real time
data exchange. Second, the digital twin application is a well demonstrative module of creating a real
time connected virtual replica of the production system. Finally, several functionalities are assured
through this application from monitoring, real time inspection, and security enhancement since the
application can follow the robots evolution and predict any collision before being occurred.
221
Fig. 45 Digital Twin Application, Monitoring Mode
The digital twin simulation mode is an offline working mode of the digital twin module which is
connected directly to the robots virtual system at 0 velocity through Office Lite as shown in Fig. 46.
The application is connected also to the WSP and it is able to simulate new robots trajectories before
deploying them into the system operations. Moreover, it enables the production engineer to simulate the
production program before getting deployed to the WSP, creating a potential added value concerning
the industrialization phase by reducing development time and decreasing the deployment phase since
the production program will be tested and commissioned virtually without causing a long period of
disruption of the production.
Before conducting a demonstrative application to validate the integration process in section 9.8, we aim
to represent in the section the whole system using an ontology model figuring the implemented system
as shown in Fig. 47. This figure represents the new holistic approach of the system architecture,
considering each machine and application integrated to the system as a self-standing module having its
own functionality and performing services across the whole system. This service can be a data
processing service, or a production processing operation configured to be adapted in the overall system.
First aspect highlighted through this model is the different functionalities of Industry 4.0 assured through
the system development, starting from an enhanced flexibility concerning the processing dimension, an
enhanced quality control well integrated to the system, the adaptability assured to different
circumstances and unintended changes in the production, the intelligence implemented through the use
of AI, and finally the customization enabled thanks to the system architecture and configuration. The
second aspect highlighted is the diversity development axis and dimensions enabled through the
deployed production system and the emergence of technological activities and research fields: robotics
systems, computer vision, data mining, cloud computing, predictive maintenance, modular and flexible
production information system, and digital integration. This diversity of fields drives us far to point out
the mass of competencies required and needed to be developed in the context of production systems
digitalization.
As discussed earlier, two key factors were identified in order to achieve the transition goal:
- A production system capable of adapting to new orders/products (under the factory capacity) and
able to respond quickly to fluctuations in demand and workflow assuring the mass customization
aspect.
223
- A production system capable of exploiting data collected from machines, sensors, failures, and
historic data, in order to learn and self-optimize its production processes.
The first goal was presented well in the different sections of the experimentation chapter. At the end, it
is imperative to present the contribution of this work regarding the data aggregation performed though
the implementation process. For each data level Fig. 48, data type is identified, the source of data is
presented, the relevance of data is pointed out by its signification, data lake and local migration is well
identified, and finally the bridge to the cloud is identified along with a relevance recommendation
proposed.
As shown in Fig. 48, data collected from the platform is aggregated into different data type. Low level
data considers the raw data from machines type I/O, collected from transfer and machining, serving as
alarms, machines availability, machines status, and processing data. The structured data type includes
both continuous data and IoT data collected from embedded sensors on machines, measuring power
consumption, vibration on the lathe, temperature and plate positioning on the laser cutting machine,
driven to the MES and historized for data processing. The quality control data includes structured
scanning data and generated reports sent to the MES, evaluating dimensional and geometrical aspects
of the produced parts and controlling measurement drift values. The processing of quality control is
done on the processing machine where reports are generated and shared with the production program
on the MES. The unstructured data including camera images collected from embedded cameras on the
robots and the inspection camera are shared to the network, while images from fixed camera system are
processed inside the robots system. The last data type taken into consideration is the operational data
224
including unstructured data, machining data such as Gcodes, and production plan data saved to the MES
library as products data base along with external data from business level. This data type is considered
as very relevant information and data base for production management. Finally, Fig. 49 offers a separate
representation of data flow for each data type.
Fig. 49 Industrial Platform 4.0, Data Lake & Data Flow Architecture
From the conceptual modelling of digital production systems, an Integration Framework has been
proposed to augment the 4.0 platform with new functionalities such as flexibility, modularity,
customization, decentralization, and portability. To this end, a new system architecture has been
deployed at the platform. As part of this experimentation, a qualitative application is presented in this
section which aims to validate the new functionalities of the production system. Among the various
functionalities defined above, we mainly want to validate the flexibility of the production system as well
as the lead-time yield for the industrialization phase and configuration of the production plan of a new
product as well as the execution phase. The validation of the flexibility of the system will be based on a
functional comparison of the current production system, close to 4.0, with the old 3.0 operating mode,
by proposing a reference for quantitative assessment of the new performance.
225
Following the development of the experimental interface, a commissioning phase was held to verify the
well-functioning of the flexible operations. Concerning the transfer functions, all possible combinations
were tested in order to validate the secure and successful passage of the robots between all machines
and production stages / zones. Regarding CNC machines, the flexible G-code is tested as well on both
the lathe and the milling machine. The monitoring is confirmed through the ability to change the sub-
program selection (from a list of 8 sub-G-code programs active on the machine) corresponding to a
specific product.
In order to validate the proposed architecture and comparing it with the traditional approach, several
tests were conducted by a group of interns and production engineering students. The aim of the
conducted experimentations is to validate the simplicity of creating a new production plan for a new
product along with the significant short-time consumed at the industrialisation phase to configure a new
production plan. The 2nd objective is to validate the flexibility of the new architecture of our production
system in comparison with the old architecture. For this, the experiments are conducted in two stages
considering two scenarios. The 1st scenario concerns the definition of a new production plan for a pawn
"A". The 2nd scenario concerns receiving a new order for a pawn "B" with a dimensional variant
compared to the pawn "A", knowing that the grippers mounted on two (fixed and mobile) robots are
identical and accept a diameter margin between 20 and 35 mm. Pawn "A" and Pawn "B" have the same
base diameter of 25mm.
9.8.1 Phase 1
The experimentation is supported by a library module created on the HMI to guide the operator and
assist him with basic information concerning different types of operations fulfilled with descriptions and
annotations. A chess piece Pawn “A” has been identified to be produced by the platform in its new
configuration. The production engineer uses the HMI to create a new product called Pawn A. The first
step is to assign different transfer operations necessary for the robot to manipulate the part, as well as to
select the appropriate parameters corresponding to the dimension of the raw part, knowing that all
possible transfer operations are already defined, listed and well indexed on the HMI. All parameters
configuration of transfer operations identified for Pawn A are similar for all trajectories, as follow:
- ID = 1001, the ID 1001 has been assigned to this Pawn A which designates product number 1 of series
1.
- H = 5, the height of the part gripping between the base of the clamp and the base of the part is 5 mm.
This value also corresponds to the height of the grip when recovering the part from the lathe chuck,
knowing that the chuck carries the part at 3 mm from its base.
226
- S = 15%, corresponding to 15% of the maximum speed of robots including linear and Point to Point
(PtP) movements.
Regarding machining operations, the operator generates machining programs in the form of G-code
corresponding to the product based on its 3D model through a CAM application. Created and indexed
machining programs are imported to the HMI module of the MES and subsequently shared with CNC
machines by updating the list of 8 active subroutines. Once the transfer functions and machining
operations have been configured and assigned to the product, the operator generates functional modules
grouping configured operations as explained previously. Next, the operator creates the production plan
as shown in Figure 45, adding a set of modules presented by a list box, and orchestrating the generated
modules based on the selected product. The last phase is to select the configured product in the
production plan interface where the operator can add the required quantity of the product and execute
the production.
9.8.2 Phase 2
The 2nd step is to create a new production plan for a Pawn “B”. Once the Pawn “A” production plan is
properly configured, we now assume the arrival of a new order Pawn "B". The two requested pawns
have the same geometric shape; however, Pawn B marks a height difference of +10 mm compared to
Pawn A, see Figure. 50. The objective of choosing the same parts with different heights figures in the
need of changing the associated production plan that we wish to investigate. This assessment
demonstrates the ability of the system to respond to personalized customer requests. The height of Pawn
A, X1 is 50 mm, while the height of Pawn B, X2 is 40 mm. Therefore, the height Y1 of the cylindrical
base (its thickness) of the Pawn A is 8mm and that of the Pawn B is 6.5mm.
A new task of creating the production plan for Pawn B is required through the developed HMI. Transfer
operations required for machining Pawn B are the same as those for Pawn A:
- RF_Pick_InitStore: corresponding to picking the raw part from the initial store by the fixed robot
- RF_Drop_TransStore: corresponding to placing the raw part in the transition store always by the fixed
robot
- RM_Pick_TransStore: corresponding to picking the raw part from the transition store by the mobile
robot
- RM_Drop_Lathe: corresponding to feeding the raw part into the chuck of the lathe
- RM_Pick_Lathe: corresponding to recover the machined part from the lathe chuck
227
- RF_Pick_TransStore: corresponding to picking the machined part by the fixed robot
- RF_Drop_Conv: corresponding to drop / place the machined / finished part on the finished products
conveyor in his way to the accumulation table by the fixed robot
However, the required changes to transfer operations can be found in the settings. Indeed, the diameter
and the speed of robots remain identical to the Pawn A, while a new identifier of the Pawn B will be
generated by the MES (ID2 = 2001, part number 1 of the series number 2), and associated to this new
product. However, a change in the grip height will be required specifically for the RM_Pick_Tour
trajectory (ID2, Ø, H2, S) which recovers the Pawn B after being machined in the lathe. Since the chuck
holds the part to a depth of 5 mm, the remaining dimension of the cylindrical base of the Pawn B will
be 1.5 mm, which causes difficulty in gripping the part and a risk of the part falling into the machine.
Consequently, the height of the grip for the trajectory RM_Pick_Tour will be readjusted to the value of
15 (H2 = 25 mm), in order to recover the Pawn B by holding its body from the spherical profile at the
top of the part. Subsequently, a second g-code will be generated by the CAM tool, imported to the MES
and sent directly to the lathe updating the list of active machining programs on the machine. The Pawn
B production plan will be created by assembling transfer and machining functional modules following
the same logic of grouping basic operations as the Pawn A.
The last phase consists in selecting both configured products, Pawn A and Pawn B successively, in the
production piloting interface. The production volume for this experiment is limited to one piece by
product. The production was executed successfully according to the "initial" production plan
corresponding to an automatic sequential execution.
228
9.8.3 Proposed measurement frame of reference for the quantitative
assessment of the new performance
Indeed, the quantification for this type of project is difficult to achieve, especially when evaluating the
integration work and the evolution of different layers of the production system. In order to quantify our
contribution, we have proposed an estimated performance frame of reference to validate the
functionalities of the production system. This estimated comparison reference mainly considers
flexibility and lead-time performance for the industrialization and configuration phase of the production
plan of a new product as well as the production execution phase. Table 7 represents a measurement
assessment proposed to be able to estimate the new performance of the production system of the 4.0
platform in terms of flexibility and lead-time efficiency by comparing the 4.0 architecture with the
traditional 3.0 architecture.
The creation of a production plan for a product includes elements of measures that concern several tasks:
- The configuration of machining operations consists in generating g-codes associated with the product,
which is required in both approaches. The estimated time required to perform this task corresponds to
the CAM tool used in terms of simplicity and skills of the operator (estimated to one hour); while the
possibility of parametrizing the machining program in the 4.0 approach has an advantage over the
traditional approach. However, setting up machining programs in the 4.0 approach requires an additional
task of encapsulating programs generated by the CAM tool according to a well-defined protocol. The
additional time required to perform this configuration is estimated to 30 minutes.
- The configuration of transfer operations consists of generating robot trajectories associated with the
manufacture of a new product, taking into account its dimensional specifications. This task is necessary
to be carried out in the traditional approach, given the lack of possible parameterization of the
trajectories, which requires additional costs considering the intervention of a roboticist and loss of time
estimated to one working day. In the case of a flexible 4.0 workshop, trajectories are designed to serve
a variety of products and are available with a flexible and parametric configuration which marks a
significant advantage over the traditional approach in terms of time and cost.
- The configuration of functional modules as well as the orchestration of the production plan of a product
and the overall production planning are grouped together in the 3rd task concerning the creation of the
production plan. The configuration of this task is required in the traditional approach by creating logic
functions and grafcets associated to a product and carried out with passage conditions in terms of
229
settings. This task requires the need for the intervention of an automation engineer marking an additional
cost and time estimated to one working day for the creation of new automatic production programs. In
the new 4.0 architecture, it is adequate to perform these configurations through the dedicated production
HMI without the need for additional development at the automatic layer and without any modification
required for the PLC program.
Table. 7 Proposed measurement frame of reference for the evaluation of the new performance
The piloting process includes other elements of measures that concern the production execution:
- Remote access to CNC machines and robots is limited to the automatic execution of machines
estimated at a maximum accessibility rate of 50% in the traditional approach, while this rate rises to
80% - 90% for machines and robots respectively in the new interoperable approach. This estimation was
based first of all on the possibility of parameterizing robot transfer functions and machining programs
thanks to the new configuration of these machines, the possibility of selecting and executing a list of
230
programs exchanged remotely with machines, and the creation of an on-board robot database that
manages product flows at the shop floor level of the platform and exchanges this information with the
production system. From a functional point of view, the fact of evolving machine / robot configurations
at the shop floor level towards a CPS configuration while solving the interoperability constraints as
explained in section 9.2.1 has contributed to improving the functioning and the access level of machines.
- The commissioning task is pre-requisite for the execution of production in the case of the traditional
approach to validate new automatic programs developed and verify the assignment of input / output
variables. This task is estimated at half a day of additional work or even one full day versus one hour of
general validation of machines operation in the new 4.0 approach.
- Automatic execution is possible for both approaches, but automatic successive execution of several
products is not allowed in the traditional approach. Actually, in the traditional approach a manual
deployment of PLC programs product by product is needed. Furthermore, the possibility of global
optimization of production processes marks a difference as well for the 4.0 approach estimated at a
minimum of 80% thanks to the generation of data and the possibility of processing it.
- Production supervision and product traceability are estimated to 99% in the 4.0 mode against 50% in
the traditional approach, thanks to the integration of the MES and the exchange of data generated from
various workstations during production.
The traditional approach of configuring a production plan for a new product consists of creating all
correspondent machining and transfer operations (and sometimes from scratch) required to manufacture
the product. Afterward, a technical or a control engineer is needed to assign first logical functions to
configured operations and consequently create a new control program and deploy it to the PLC
controller, followed by a commissioning phase.
Considering the phase 1 of the application carried out, the simplicity of various tasks and sub-tasks,
measured in the repository, of the creation and configuration of a new production plan has been well
validated. Indeed, the realization of a new production plan based on the traditional architecture of the
platform requires first the development of a new program at the PLC level in the form of a Grafcet which
deploys various machining and robotic transfer functions, associated with a well-defined product. A
commissioning phase will always be required in order to verify the assignments of logical variables
corresponding to different operations, and to verify the operating of new configured processes.
The 2nd phase of the application validates once again the simplicity and speed of creating a new
production plan. Henceforth, it highlighted the flexibility of the new configuration of the production
system at the Planning / monitoring level as well as at the shop floor level. By carrying out the same
231
scenario on two products Pawn A and Pawn B with the old traditional approach, three technical
dimensions were added. The 1st dimension concerns the constraint resulting from the need to recover
the machined part at a specific gripping height for the Pawn B, which is different from the configured
one. This constraint required two technical tasks in the traditional configuration of the platform. The 1st
technical task required is the creation of a new specific trajectory to recover this product from the lathe
which takes the same movements of the basic trajectory but changing the point of arrival on the machine
chuck in Z axis. The second technical task concerns the assignment of a new cycle code to this trajectory
and its deployment by creating a new logical function to be implemented in the grafcet corresponding
to this product. This difficulty was dealt in the new platform architecture with configuring the parameters
of robotic transfer functions. The second technical dimension concerns the need to activate a new g-
code machining program on the lathe. Loading the new g-code into the machine was manual, as was its
selection. This constraint was resolved thanks to the interoperability deployed at the platform level
which facilitated the exchange between the MES and various machines. The 3rd dimension considers
the advanced role of MES in the planning and management of the production of a variety of products.
Always comparing the application carried out with the old approach, it was necessary to launch the 1st
production program of Pawn A which figures in the PLC program, then deploy the 2nd production
program of Pawn B at the PLC level and simultaneously launch the production cycle of the 2nd product.
The fact that marks an increased flexibility in the new architecture and a considerable level of monitoring
in terms of autonomy.
Transfer functions are well configured, and the flexible design of this production system has performed
well, giving the operator the ability to easily drive the production plan in various configurations and
combinations through a simple HMI without any additional coding or technical effort required as shown
in the measurement reference. Three main functional dimensions are worth mentioning in terms of
comparing the new approach with the traditional approach. Mainly technical skills, including the
"complexity" dimension and the "time saving in the industrialization phase" dimension. Considering the
1st dimension, the operator or the production engineer does not need to have specific skills in control,
robotics and other technical fields for the generation of a new production plan, thus reducing the costs
of the intervention of external integrators. The time saving is ensured by reducing the configuration time
required for the industrialization phase and production execution with an estimation of 14 hours less.
Thus, the reduction in time achieved is also linked to minimizing the need for a commissioning phase
since the new approach is flexible and designed upstream to manage a variety of products taking into
account machine capacities. The third dimension focuses on the flexibility of the evolved production
system. This flexibility deployed in the system is due to the integration of the Framework and the
evolution of different layers of our production system including the shop floor layer but also the control,
management, and planning layers.
232
The experimentation presented in this section shows that the MES plays an important role in intelligent
manufacturing processes, and also demonstrates that the notion of operational flexibility must be taken
into account in the design phase of the advanced production system, especially when working in a more
complex manufacturing system equipped by more machines, devices and sensors. The framework
proposed has shown the relevance of our methodological choices and our research orientation. The
perspective in terms of development consists in developing the autonomy of the “management” function
as well as deploying multi-criteria production optimization algorithms.
233
10. Results & General Discussion
The vertical integration taken into consideration at the modelling level along with essential modelling
concepts drive us far at the experimentation phase by applying the proposed framework at a wide range
including different levels of the production system. The integration done based on the framework has
led us to study and develop a novel approach at the configuration level of the new production system to
be deployed. This configuration based on the flexibility and the modularity assured through the new
system architecture contributed to different opened opportunities considering the production system
organisation, process planning and enhanced modules integration, along with an advanced deployment
of the implanted technologies.
The work carried out during this PhD thesis is a new approach allowing increased flexibility of a
production system. This allows us to easily control the manufacturing of various products through an
HMI rather than reprogramming the industrial controllers which is normally required to enable a new
product manufacturing process. The contribution of this thesis lies first in the successful integration of
the production system based on the proposed model and the developed Framework. Second, the
development of a flexible and modular approach is achieved, based on several advanced technologies
such as advanced robotics, embedded camera systems and flexible machine tools, as well as the
implementation of a new control architecture which consists of upgrading the decision-making process
to an advanced and interoperable MES system. The operation of our architecture has been successfully
tested. Despite interoperability challenges, we have succeeded in developing an automated modular and
flexible production system capable of manufacturing a wide range of products from the same raw
materials.
Following the development of the experimental interface, all possible combinations were tested, in order
to validate the secure and successful passage between all machines and production stages. The paths are
well configured, and the modular design of this production system worked very well, giving the operator
the opportunity to drive the production plan in various configuration and combination easily through a
simple HMI without the need for any additive coding or technical effort. This validates an interesting
revenue concerning the industrialization phase with a remarkable reduced time set to integrate a new
product to the production system.
The proposed novel model for a smart visual system well integrated into the manufacturing execution
system as well as the digital twin module contribute to improve the production system. This integration
is a concrete contribution in terms of the modular information system that can be used for several
applications and cases and assure the optimization of the production process and control. These modules
give us wide opportunities to develop different parallel modules for data and image processing and
integrate them into the manufacturing system in order to enhance and well optimize the whole
234
production process. Finally, the data dimension is well considered by the development phase. The data
management integration in parallel with the new flexible configuration of the platform enable the
potential of a continuous self-optimisation of production processes through the data processing and
mining to be implemented.
Furthermore, and based on the developed methodology, the proposed architecture can be applied to other
manufacturing system with specific requirements involving controllable and programmable robots
playing transfer functions. The specific requirements abovementioned concerns first the degree of
flexibility of machines at the shop floor and its ability to perform multiple operations from a limited
number of resources. Machines flexibility requires at least solving issues such as interoperability,
accessibility or remote access to machines, and enabling flexible configuration of the machining process
in terms of dimensional capability to carry out a margin of parts, and in terms of parametrization.
Second, a special attention should be paid to the configuration of logical functions at the API layer
taking into consideration the operational flexibility of machines. Third, the digital MES platform
developed or adapted to play the role of Production Planning and Control PPC system must handle new
challenges created by Industry 4.0. MES needs to become logically decentralized and composed of
decoupled objects or service applications with an external service responsible for the coupling of
connected machines and CPS. Lastly, it is worth mentioning that applying the proposed design
methodology should be driven by a new configuration insight of the shop floor taking into consideration
the flexibility deployed and the expected patterns from the production system. We believe that our work
is a foundation that can be deployed at the levels of manufacturing companies, especially for small and
medium enterprises, where operational flexibility plays an important role in an ever-changing market.
In terms of limitations, we distinguish two axes: scientific research limitations and technical limitations
considering the development and experimental work. Regarding research limitations, specific attention
to the principle of multi-agent systems could enrich this research thesis. On the other hand, the definition
of the required skills for the digitization of manufacturing systems as well as identifying jobs evolution
in the context of Industry 4.0 could improve this work. Regarding the experimental limitations, it should
be mentioned at the very beginning that the realization of the integration work and the deployment of
the new architecture is a main challenge in terms of integration project management. Generating a
roadmap from this experimental approach could be appreciated.
235
Conclusion & perspective
Future manufacturing systems need to be more flexible and intelligent in order to embrace customer and
market demands. Industry 4.0 is a promised industrial revolution, taking the industrial world to an
advanced level of product mass customization, and enabling a high level of digital manufacturing
facilities. To well manage the targeted transition, several stages of development are required, innovative
technologies are to be integrated, and the company has to deal with a wide area of challenges affecting
different component of the company’s structure. The main challenge is how to guide companies among
the evolution from Industry 3.0 to 4.0. The aim of this PhD thesis is to propose a transition framework
based on an appropriate model to help companies achieve a successful digital transformation and
integration Fig. 51. This Framework takes into consideration three main axes Production System
including Shop Floor Control System SFCS, Machines & Advanced Techs, and Data Processing
& Deployment.
In this thesis, the state of the art of Industry 4.0 is well discovered. Industry 4.0 definition and concept
is presented, as well as its changes and benefits affecting the manufacturing sector. The axes of research
are well stated, and a summary of Industry 4.0 technology trends, principles and functionalities is
236
delivered. In addition, the existing integration models were presented, and the challenges facing
companies in the digitalization of their manufacturing systems are well discussed in order to answer our
central research problematic: which model for a successful digital transition of 4.0 manufacturing
systems?
In result, the PhD hypothesis is identified as Evolving production systems towards Industry 4.0
requires acting on three main axes Production System including Shop Floor Control System
SFCS, Machines & Advanced Techs, and Data Processing & Deployment.
This proposal is based on a deep literature study and related challenges and problematics regarding our
research direction and topic. Therefore, a novel conceptual model is introduced and explained. The
proposed model is then deployed to an existing conventional production system in order to identify the
evolutions required and to define consequently the transition framework regarding the actions needed
to be taken for the digital integration goal. A design principles framework along with a novel integration
architecture toward the digital manufacturing transition is introduced. The proposed framework
contributes to considering the different layer of the production system of an enterprise and recommends
novel and advanced modification in order to assure a robust evolution towards the digitalization of the
manufacturing system. This framework is well applied to an existing traditional architecture
implemented to the industrial platform 4.0.
Finally, a demonstrative application was carried out at the industrial platform 4.0 of the Arts et Métiers,
in order to validate the new architecture of the production system assuring the digital transition to
Industry 4.0. This architecture has allowed the emergence of new advanced functionalities such as
flexibility, and lead-time performance.
Several further research and development trends are identified as perspectives. Based on the above-
mentioned development, we believe the production system configuration regarding the process planning
must be taken into consideration with the evolved flexible system. The production planning
configuration is a separate field of research and development, and a special attention should be paid on
it in the context of Industry 4.0. Process planning in the digitalized production system is a crucial trend
and has a strong connection with Lean 4.0 especially when considering the smart logistics aspects.
Additionally, further development will focus on the optimisation opportunities enabled by the system
considering time, cost and power consumption criterion through embedded sensors already mounted on
the machines. The data mining dimension and the possibility of having a continuous self-optimisation
of the production processes through the data processing enabled thanks to the flexible production system
is critical to benefit from the digitalization of the manufacturing processes and to value the aimed
performance of the 4.0 integration.
237
Finally, it is necessary to express the need for a multidisciplinary approach in projects of this type. The
transition to Industry 4.0 is the paradigm of the intersection of multiple areas of engineering expertise
and competencies. Capitalizing on the knowledge and the required competencies of each of these areas
to transform the industry is, in our opinion, one of the major issues.
238
References
Adjoul, O., Benfriha, K., Aoussat, A., 2019. Algorithmic strategy for optimizing product design
considering the production costs. Int J Interact Des Manuf 13, 1313–1329.
https://doi.org/10.1007/s12008-019-00571-w
Bajic, B., Cosic, I., Lazarevic, M., Sremcev, N., Rikalovic, A., 2018. Machine Learning Techniques for
Smart Manufacturing: Applications and Challenges in Industry 4.0 11.
Beckhoff Automation GmbH & Co KG, 2021. TwinCAT | Automation software | Beckhoff France
[WWW Document]. Beckhoff Automation. URL https://www.beckhoff.com/fr-
fr/products/automation/twincat/ (accessed 1.24.21).
Belvedere, V., Grando, A., Bielli, P., 2013. A quantitative investigation of the role of information and
communication technologies in the implementation of a product-service system. International
Journal of Production Research 51, 410–426. https://doi.org/10.1080/00207543.2011.648278
Benfriha, K., 2020. Conception, Modélisation, Architecture, Technologie et Pédagogique (Mémoire
HDR). ENSAM Paris.
Bloch, H., Fay, A., Knohl, T., Hoernicke, M., Bernshausen, J., Hensel, S., Hahn, A., Urbas, L., 2017. A
microservice-based architecture approach for the automation of modular process plants, in:
2017 22nd IEEE International Conference on Emerging Technologies and Factory Automation
(ETFA). Presented at the 2017 22nd IEEE International Conference on Emerging Technologies
and Factory Automation (ETFA), IEEE, Limassol, pp. 1–8.
https://doi.org/10.1109/ETFA.2017.8247573
Bloch, H., Grebner, T., Fay, A., Hensel, S., Menschner, A., Urbas, L., Hoernicke, M., Knohl, T.,
Bernshausen, J., Ag, B., 2018a. Orchestration of Services in Modular Process Plants, in: IECON
2018 - 44th Annual Conference of the IEEE Industrial Electronics Society. Presented at the
IECON 2018 - 44th Annual Conference of the IEEE Industrial Electronics Society, IEEE, D.C.,
DC, USA, pp. 2935–2940. https://doi.org/10.1109/IECON.2018.8591300
Bloch, H., Hensel, S., Hoernicke, M., Stark, K., Menschner (neé Hahn), A., Fay, A., Urbas, L., Knohl,
T., Bernshausen, J., 2018b. State-based control of process services within modular process
plants. Procedia CIRP 72, 1088–1093. https://doi.org/10.1016/j.procir.2018.03.037
Block, C., Lins, D., Kuhlenkötter, B., 2018. Approach for a simulation-based and event-driven
production planning and control in decentralized manufacturing execution systems. Procedia
CIRP 72, 1351–1356. https://doi.org/10.1016/j.procir.2018.03.204
Bodrow, W., 2017. Impact of Industry 4.0 in service oriented firm. Adv. Manuf. 5, 394–400.
https://doi.org/10.1007/s40436-017-0196-3
239
Bonnard, R., Vieira, K.M.M., Arantes, M.D.S., Lorbieski, R., Nunes, C., Mattei, A.P., 2019. A BIG
DATA / ANALYTICS PLATFORM FOR INDUSTRY 4.0 IMPLEMENTATION IN SMEs
10.
Borangiu, T., Trentesaux, D., Thomas, A., Leitão, P., Barata, J., 2019. Digital transformation of
manufacturing through cloud services and resource virtualization. Computers in Industry 108,
150–162. https://doi.org/10.1016/j.compind.2019.01.006
Carvalho, N., Chaim, O., Cazarini, E., Gerolamo, M., 2018. Manufacturing in the fourth industrial
revolution: A positive prospect in Sustainable Manufacturing. Procedia Manufacturing 21, 671–
678. https://doi.org/10.1016/j.promfg.2018.02.170
Chen, H., 2017. Theoretical foundations for cyber-physical systems: a literature review. Journal of
Industrial Integration and Management, 2(03), 1750013.
Connectivity IML Magazine, 2019. Preparing Manufacturing Systems for Industry 4.0 [WWW
Document]. URL http://www.connectivity4ir.co.uk/article/167720/Preparing-Manufacturing-
Systems-for-Industry-4-0.aspx (accessed 1.31.20).
Dalenogare, L.S., Benitez, G.B., Ayala, N.F., Frank, A.G., 2018. The expected contribution of Industry
4.0 technologies for industrial performance. International Journal of Production Economics 204,
383–394. https://doi.org/10.1016/j.ijpe.2018.08.019
de Man, J.C., Strandhagen, J.O., 2017. An Industry 4.0 Research Agenda for Sustainable Business
Models. Procedia CIRP 6.
De Pace, F., Manuri, F., Sanna, A., 2018. Augmented Reality in Industry 4.0 6.
de Sousa Jabbour, A.B.L., Jabbour, C.J.C., Foropon, C., Godinho Filho, M., 2018. When titans meet,
can industry 4.0 revolutionise the environmentally-sustainable manufacturing wave? the role of
critical success factors. Technological Forecasting and Social Change, 132:18{25.
Derbel, N., GHOMMAM, J., ZHU, Q. (ed. ), 2019. New Developments and Advances in Robot Control.
Springer Singapore.
Dilberoglu, U.M., 2017. The Role of Additive Manufacturing in the Era of Industry 4.0. Procedia
Manufacturing 10.
Duckworth, N., 2019. Industry 4.0: Why Manufacturers Need to Keep Their Eye on the Long Game
[WWW Document]. Manufacturing Business Technology. URL
https://www.mbtmag.com/business-intelligence/article/13251875/industry-40-why-
manufacturers-need-to-keep-their-eye-on-the-long-game (accessed 2.25.21).
El Zant, C., Charrier, Q., Benfriha, K., Le Men, P., 2021. Enhanced Manufacturing Execution System
“MES” Through a Smart Vision System, in: Roucoules, L., Paredes, M., Eynard, B., Morer
Camo, P., Rizzi, C. (Eds.), Advances on Mechanics, Design Engineering and Manufacturing
III, Lecture Notes in Mechanical Engineering. Springer International Publishing, Cham, pp.
329–334. https://doi.org/10.1007/978-3-030-70566-4_52
240
ElMaraghy, H.A., 2005. Flexible and reconfigurable manufacturing systems paradigms. Int J Flex
Manuf Syst 17, 261–276. https://doi.org/10.1007/s10696-006-9028-7
Erol, S., Schumacher, A., 2016. Strategic guidance towards Industry 4.0 – a three stage process model
8.
Esmaeilian, B., Behdad, S., Wang, B., 2016. The evolution and future of manufacturing: A review.
Journal of Manufacturing Systems 39, 79–100. https://doi.org/10.1016/j.jmsy.2016.03.001
Fan, W., Zheng, L., Ji, W., Xu, X., Wang, L., Lu, Y., Zhao, X., 2020. Function block-based closed-loop
adaptive machining for assembly interfaces of large- scale aircraft components. Robotics and
Computer-Integrated Manufacturing, 66: 101994.
Feng, S., Li, L.X., Cen, L., 2001. An object-oriented intelligent design tool to aid the design of
manufacturing systems. Knowledge-based systems, 14(5-6), 225-232.
Foresight, U., 2013. The future of manufacturing: a new era of opportunity and challenge for the UK -
summary report 54.
Fragapane, G., Ivanov, D., Peron, M., Sgarbossa, F., Strandhagen, J.O., 2020. Increasing flexibility and
productivity in Industry 4.0 production networks with autonomous mobile robots and smart
intralogistics. Ann Oper Res. https://doi.org/10.1007/s10479-020-03526-7
Frank, A.G., Dalenogare, L.S., Ayala, N.F., 2019. Industry 4.0 technologies: Implementation patterns
in manufacturing companies. International Journal of Production Economics 210, 15–26.
https://doi.org/10.1016/j.ijpe.2019.01.004
Geissbauer, R., Lübben, E., Schrauf, S., Pillsbury, S., 2018. Global Digital Operations Study 2018.
Digital Champions. How industry leaders build integrated operations ecosystems to deliver end-
to-end customer solutions.
Ghobakhloo, M., 2018. The future of manufacturing industry: a strategic roadmap toward Industry 4.0.
Jnl of Manu Tech Mnagmnt 29, 910–936. https://doi.org/10.1108/JMTM-02-2018-0057
Goerzig, D., Bauernhansl, T., 2018. Enterprise Architectures for the Digital Transformation in Small
and Medium-sized Enterprises. Procedia CIRP 67, 540–545.
https://doi.org/10.1016/j.procir.2017.12.257
Gorecky, D., Weyer, S., Hennecke, A., Zühlke, D., 2016. Design and Instantiation of a Modular System
Architecture for Smart Factories. IFAC-PapersOnLine 49, 79–84.
https://doi.org/10.1016/j.ifacol.2016.12.165
Gupta, S., 2018. Artificial Intelligence for Enterprise Networks, in: Artificial Intelligence for
Autonomous Networks. Chapman and Hall/CRC, pp. 263–284.
https://doi.org/10.1201/9781351130165-11
Gürdür, D., El-Khoury, J., Seceleanu, T., Lednicki, L., 2016. Making interoperability visible: Data
visualization of cyber-physical systems development tool chains. Journal of industrial
information integration, 4, 26-34.
241
Hartwell, R.M., 2017. The Causes of the Industrial Revolution: An Essay in Methodology 20.
Henning, K., Wolfgang, W., Johannes, H., 2013. Recommendations for implementing the strategic
initiative INDUSTRIE 4.0. Final report of the Industrie, 4, 82.
Hermann, M., Pentek, T., Otto, B., 2016. Design Principles for Industrie 4.0 Scenarios, in: 2016 49th
Hawaii International Conference on System Sciences (HICSS). Presented at the 2016 49th
Hawaii International Conference on System Sciences (HICSS), IEEE, Koloa, HI, USA, pp.
3928–3937. https://doi.org/10.1109/HICSS.2016.488
Horst, D.J., Duvoisin, C.A., de Almeida Vieira, R., 2018. Additive manufacturing at Industry 4.0: a
review. International journal of engineering and technical research, 8(8).
Ibarra, D., Ganzarain, J., Igartua, J.I., 2018. Business model innovation through Industry 4.0: A review.
Procedia Manufacturing 22, 4–10. https://doi.org/10.1016/j.promfg.2018.03.002
Information and communications technology, 2021. . Wikipedia.
Islam, M.M., Kim, J.M., 2019. Automated bearing fault diagnosis scheme using 2D representation of
wavelet packet transform and deep convolutional neural network. Computers in Industry, 106,
142-153.
Jakl, A., Schoffer, L., Husinsky, M., Wagner, M., 2018. Augmented Reality for Industry 4.0:
Architecture and User Experience 5.
Jasperneite, J., 2012. Was hinter Begriffen wie Industrie 4.0 steckt. Computer & Automation, 19.
Joost, D., Sutherland, J.W., Dornfeld, D.A., Herrmann, C., Jeswiet, J., Kara, S., Hauschild, M.Z.,
Kellens, K., 2012. Towards energy and resource efficient manufacturing: A processes and
systems approach. CIRP Annals, vol. 61(2), pp. 587-609.
Jung, J.Y., 2002. Manufacturing cost estimation for machined parts based on manufacturing features.
Journal of Intelligent Manufacturing, vol. 13(4), pp. 227-238.
Kagermann, H., Wahlster, W., Helbig, J., 2013. Recommendations for implementing the strategic
initiative INDUSTRIE 4.0 (No. 1).
Kamble, S.S., Gunasekaran, A., Gawankar, S.A., 2018. Sustainable Industry 4.0 framework: A
systematic literature review identifying the current trends and future perspectives. Process
Safety and Environmental Protection 117, 408–425. https://doi.org/10.1016/j.psep.2018.05.009
Kang, H.S., Lee, J.Y., Choi, S., Kim, H., Park, J.H., Son, J.Y., Kim, B.H., Noh, S.D., 2016. Smart
manufacturing: Past research, present findings, and future directions. Int. J. of Precis. Eng. and
Manuf.-Green Tech. 3, 111–128. https://doi.org/10.1007/s40684-016-0015-5
Karnouskos, S., Colombo, A.W., Jammes, F., Delsing, J., Bangemann, T., 2010. Towards an architecture
for service-oriented process monitoring and control, in: IECON 2010 - 36th Annual Conference
on IEEE Industrial Electronics Society. Presented at the IECON 2010 - 36th Annual Conference
of IEEE Industrial Electronics, IEEE, Glendale, AZ, USA, pp. 1385–1391.
https://doi.org/10.1109/IECON.2010.5675482
242
Kaschel C., H., Y Bernal, L.M.S., 2006. Importance of Flexibility in Manufacturing Systems. INT J
COMPUT COMMUN 1, 53. https://doi.org/10.15837/ijccc.2006.2.2285
Kaushal, A., Vardhan, A., Rajput, R.S., 2016. Flexible Manufacturing System A Modern Approach To
Manufacturing Technology 8.
Kaynak, O., 2007. The exhilarating journey from industrial electronics to industrial informatics.
Presented at the 2007 2nd IEEE Conference on Industrial Electronics and Applications (pp. xli-
xlii). IEEE.
Kennedy, S., 2015. Made in China 2025. Centre for Strategic and International Studies.
Khan, W.Z., Rehman, M.H., Zangoti, H.M., Afzal, M.K., Armi, N., Salah, K., 2020. Industrial internet
of things: Recent advances, enabling technologies and open challenges. Computers & Electrical
Engineering, 81, 106522.
Kim, D.-Y., Park, J.-W., Baek, S., Park, K.-B., Kim, H.-R., Park, J.-I., Kim, H.-S., Kim, B.-B., Oh, H.-
Y., Namgung, K., Baek, W., 2020. A modular factory testbed for the rapid reconfiguration of
manufacturing systems. J Intell Manuf 31, 661–680. https://doi.org/10.1007/s10845-019-
01471-2
Kumar, A., 2007. From mass customization to mass personalization: a strategic transformation. Int J
Flex Manuf Syst 19, 533–547. https://doi.org/10.1007/s10696-008-9048-6
Kumar, K., Zindani, D., Davim, J.P., 2019. Process Planning in Era 4.0, in: Industry 4.0, SpringerBriefs
in Applied Sciences and Technology. Springer Singapore, Singapore, pp. 19–26.
https://doi.org/10.1007/978-981-13-8165-2_2
Lasi, H., Fettke, P., Kemper, H.G., Feld, T., Hoffmann, M., 2014. Industry 4.0. Business & information
systems engineering, 6(4), 239-242.
Lee, J., Davari, H., Singh, J., Pandhare, V., 2018. Industrial Artificial Intelligence for industry 4.0-based
manufacturing systems. Manufacturing Letters 18, 20–23.
https://doi.org/10.1016/j.mfglet.2018.09.002
Li, Z., Wang, Y., Wang, K.-S., 2017. Intelligent predictive maintenance for fault diagnosis and
prognosis in machine centers: Industry 4.0 scenario. Adv. Manuf. 5, 377–387.
https://doi.org/10.1007/s40436-017-0203-8
Liao, Y., Deschamps, F., Loures, E. de F.R., Ramos, L.F.P., 2017. Past, present and future of Industry
4.0 - a systematic literature review and research agenda proposal. International Journal of
Production Research 55, 3609–3629. https://doi.org/10.1080/00207543.2017.1308576
Lin, Y., Lan, C.-B., Huang, C.-Y., 2019. A Realization of Cyber-Physical Manufacturing Control
System Through Industrial Internet of Things. Procedia Manufacturing 39, 287–293.
https://doi.org/10.1016/j.promfg.2020.01.449
Lina, Y.J., Lana, C.B., Huanga, C.Y., 2019. A Realization of Cyber-Physical Manufacturing Control
System Through Industrial Internet of Things Through Industrial Internet of Things. Presented
243
at the 25th International Conference on Production Research Manufacturing Innovation: Cyber
Physical Manufacturing Cyber Physical Manufacturing, Chicago, Illinois (USA).
Mabkhot, M., Al-Ahmari, A., Salah, B., Alkhalefah, H., 2018. Requirements of the Smart Factory
System: A Survey and Perspective. Machines 6, 23. https://doi.org/10.3390/machines6020023
Man, J.C. de, Strandhagen, J.O., 2017. An Industry 4.0 Research Agenda for Sustainable Business
Models. Procedia CIRP 63, 721–726. https://doi.org/10.1016/j.procir.2017.03.315
Mazars and OpinionWay, 2018. Les dirigeants face à l’Industrie 4.0 [WWW Document]. Mazars.fr.
URL https://online.mazars.fr/industrie-40 (accessed 3.5.20).
Mekni, M., Lemieux, A., 2014. Augmented Reality: Applications, Challenges and Future Trends 10.
Mes, M., Gerrits, B., 2019. Multi-agent Systems, in: Zijm, H., Klumpp, M., Regattieri, A., Heragu, S.
(Eds.), Operations, Logistics and Supply Chain Management, Lecture Notes in Logistics.
Springer International Publishing, Cham, pp. 611–636. https://doi.org/10.1007/978-3-319-
92447-2_27
Moeuf, A., 2018. The industrial management of SMEs in the era of Industry 4.0. International Journal
of Production Research 20.
Mourtzis, D., Alexopoulos, K., Chryssolouris, G., 2012. Flexibility consideration in the design of
manufacturing systems: An industrial case study. CIRP Journal of Manufacturing Science and
Technology 5, 276–283. https://doi.org/10.1016/j.cirpj.2012.10.001
Mourtzis, D., Angelopoulos, J., Dimitrakopoulos, G., 2020. Design and development of a flexible
manufacturing cell in the concept of learning factory paradigm for the education of generation
4.0 engineers. Procedia Manufacturing, Learning Factories across the value chain – from
innovation to service – The 10th Conference on Learning Factories 2020 45, 361–366.
https://doi.org/10.1016/j.promfg.2020.04.035
Mourtzis, D., Vlachou, E., Xanthopoulos, N., Givehchi, M., Wang, L., 2016. Cloud-based adaptive
process planning considering availability and capabilities of machine tools. Journal of
Manufacturing Systems 39, 1–8. https://doi.org/10.1016/j.jmsy.2016.01.003
Ollinger, L., Abdo, A., Zühlke, D., Heutger, H., 2014. SOA-PLC – Dynamic Generation and
Deployment of Web Services on a Programmable Logic Controller. IFAC Proceedings Volumes
47, 2622–2627. https://doi.org/10.3182/20140824-6-ZA-1003.02189
Perzylo, A., Grothoff, J., Lucio, L., Weser, M., Malakuti, S., Venet, P., Aravantinos, V., Deppe, T.,
2019. Capability-based semantic interoperability of manufacturing resources: A BaSys 4.0
perspective. IFAC-PapersOnLine 52, 1590–1596. https://doi.org/10.1016/j.ifacol.2019.11.427
Qin, J., Liu, Y., Grosvenor, R., 2016. A Categorical Framework of Manufacturing for Industry 4.0 and
Beyond. Procedia CIRP 52, 173–178. https://doi.org/10.1016/j.procir.2016.08.005
244
Radziwon, A., Bilberg, A., Bogers, M., Madsen, E.S., 2014. The Smart Factory: Exploring Adaptive
and Flexible Manufacturing Solutions. Procedia Engineering 69, 1184–1190.
https://doi.org/10.1016/j.proeng.2014.03.108
Rojko, A., 2017. Industry 4.0 Concept: Background and Overview. Int. J. Interact. Mob. Technol. 11,
77. https://doi.org/10.3991/ijim.v11i5.7072
Rüßmann, M., Lorenz, M., Gerbert, P., Waldner, M., Justus, J., Engel, P., Harnisch, M., 2015a. Industry
4.0: The Future of Productivity and Growth in Manufacturing 20.
Rüßmann, M., Lorenz, M., Gerbert, P., Waldner, M., Justus, J., Harnisch, M., 2015b. Industry 4.0: The
Future of Productivity and Growth in Manufacturing Industries 14.
Santos, C., Mehrsai, A., Barros, A.C., Araújo, M., Ares, E., 2017. Towards Industry 4.0: an overview
of European strategic roadmaps. Procedia Manufacturing 13, 972–979.
https://doi.org/10.1016/j.promfg.2017.09.093
Schierholt, K., 2001. Process configuration: Combining the principles of product configuration and
process planning. Artificial Intelligence for Engineering Design, Analysis and Manufacturing:
AIEDAM 15, 411–424.
Schlapfer, R.C., Koch, M., Merkofer, P., 2015. Industry 4.0 Challenges and solutions for the digital
transformation and use of exponential technologies.
Science, P.C. of A. on, Technology (U.S.), 2014. Report to the President, Accelerating U.S. Advanced
Manufacturing. Executive Office of the President, President’s Council of Advisors on Science
and Technology.
Segovia, D., Mendoza, M., Mendoza, E., González, E., 2015. Augmented Reality as a Tool for
Production and Quality Monitoring. Procedia Computer Science 75, 291–300.
https://doi.org/10.1016/j.procs.2015.12.250
SenTryo, 2017. The 4 industrial revolutions [WWW Document]. URL https://www.sentryo. net/the-4-
industrial-revolutions/ (accessed 6.17.19).
STEARNS, P.N., 2020. The industrial revolution in world history. Routledge.
Stock, T., Seliger, G., 2016. Opportunities of sustainable manufacturing in industry 4.0. Procedia Cirp,
40:536:541.
Strategy&, 2015. Small automation, big benefits [WWW Document]. PwC. URL
https://www.strategyand.pwc.com/gx/en/functions/technology-strategy/big-benefits.html
(accessed 3.5.20).
Strayer, R.W., Nelson, E., 2009. Ways of the World: A Brief Global History. Bedford/St. Martin’s.
Sung, T.K., 2018. Industry 4.0: A Korea perspective. Technological Forecasting and Social Change 132,
40–45. https://doi.org/10.1016/j.techfore.2017.11.005
Tao, F., Qi, Q., Liu, A., Kusiak, A., 2018. Data-driven smart manufacturing. Journal of Manufacturing
Systems 48, 157–169. https://doi.org/10.1016/j.jmsy.2018.01.006
245
Thames, L., Schaefer, D., 2016. Software-defined cloud manufacturing for industry 4.0. Procedia cirp,
52, 12-17.
Theorin, A., 2012. Service-oriented Process Control with Grafchart and the Devices Profile for Web
Services 6.
Theorin, A., Bengtsson, K., Provost, J., Lieder, M., Johnsson, C., Lundholm, T., Lennartson, B., 2017a.
An event-driven manufacturing information system architecture for Industry 4.0. International
Journal of Production Research 55, 1297–1311.
https://doi.org/10.1080/00207543.2016.1201604
Theorin, A., Bengtsson, K., Provost, J., Lieder, M., Johnsson, C., Lundholm, T., Lennartson, B., 2017b.
An event-driven manufacturing information system architecture for Industry 4.0. International
Journal of Production Research 55, 1297–1311.
https://doi.org/10.1080/00207543.2016.1201604
Weyer, S., Schmitt, M., Ohmer, M., Gorecky, D., 2015. Towards Industry 4.0 - Standardization as the
crucial challenge for highly modular, multi-vendor production systems. IFAC-PapersOnLine
48, 579–584. https://doi.org/10.1016/j.ifacol.2015.06.143
Wilamowski, B.M., 2005. Welcome to the IEEE Transactions on Industrial Informatics, a New Journal
of the Industrial Electronics Society. Presented at the IEEE Transactions on Industrial
Informatics, 1(1), 1-2.
Wonderware enhances Operations and Performance Software [WWW Document], 2008. .
automation.com. URL https://www.automation.com/en-us/products/product18/wonderware-
enhances-operations-and-performance-sof (accessed 2.15.21).
Wonderware System Platform - Fondation de l’Industrie du Futur [WWW Document], 2017. .
Wonderware. URL https://www.wonderware.fr/produit/supervision-et-controle/ihm-
supervision-et-controle/system-platform-2 (accessed 5.5.20).
Xu, L.D., Xu, E.L., Li, L., 2018a. Industry 4.0: state of the art and future trends. International Journal
of Production Research 56, 2941–2962. https://doi.org/10.1080/00207543.2018.1444806
Xu, L.D., Xu, E.L., Li, L., 2018b. Industry 4.0: state of the art and future trends. International Journal
of Production Research 56, 2941–2962. https://doi.org/10.1080/00207543.2018.1444806
Yang, H.-L., Chang, T.-W., Choi, Y., 2018. Exploring the Research Trend of Smart Factory with Topic
Modeling. Sustainability 10, 2779. https://doi.org/10.3390/su10082779
Yu, Z., Ouyang, J., Li, S., Peng, X., 2017. Formal modeling and control of cyber-physical manufacturing
systems. Advances in Mechanical Engineering 9, 168781401772547.
https://doi.org/10.1177/1687814017725472
Zdravković, M., Zdravković, J., Aubry, A., Moalla, N., Guedria, W., Sarraipa, J., 2018. Domain
framework for implementation of open IoT ecosystems. International Journal of Production
Research 56, 2552–2569. https://doi.org/10.1080/00207543.2017.1385870
246
Zhang, J., Yao, X., Zhou, J., Jiang, J., Chen, X., 2017. Self-Organizing Manufacturing : Current status
and Prospect for Industry 4.0. Presented at the 5th International Conference on Enterprise
Systems (ES), Beijing, China.
Zhang, L.L., Rodrigues, B., 2009. A tree unification approach to constructing generic processes. IIE
Transactions 41, 916–929. https://doi.org/10.1080/07408170903026049
Zheng, L.Y., Dong, H.F., Vichare, P., Nassehi, A., Newman, S.T., 2008. Systematic modeling and
reusing of process knowledge for rapid process configuration. Robotics and Computer-
Integrated Manufacturing 24, 763–772. https://doi.org/10.1016/j.rcim.2008.03.006
Zhong, R.Y., Xu, X., Klotz, E., Newman, S.T., 2017. Intelligent Manufacturing in the Context of
Industry 4.0: A Review. Engineering 3, 616–630. https://doi.org/10.1016/J.ENG.2017.05.015
Zhou, K., Taigang Liu, Lifeng Zhou, 2015. Industry 4.0: Towards future industrial opportunities and
challenges, in: 2015 12th International Conference on Fuzzy Systems and Knowledge
Discovery (FSKD). Presented at the 2015 12th International Conference on Fuzzy Systems and
Knowledge Discovery (FSKD), IEEE, Zhangjiajie, China, pp. 2147–2152.
https://doi.org/10.1109/FSKD.2015.7382284
Zunino, C., Valenzano, A., Obermaisser, R., Petersen, S., 2020. Factory Communications at the Dawn
of the Fourth Industrial Revolution. Computer Standards & Interfaces 71, 103433.
https://doi.org/10.1016/j.csi.2020.103433
247
Résumé
L'industrie 4.0 est un système dynamique et intégré permettant de contrôler l'ensemble de la chaîne de valeur du
cycle de vie des produits. L'intégration verticale et horizontale et la fusion des composants physiques et virtuels
sont essentielles, de même que la mise en œuvre des tendances technologiques telles que les systèmes cyber
physiques et l'Internet des objets en tant que catalyseurs clés d'un haut niveau d'intégration à l'échelle mondiale.
L'introduction de l'Industrie 4.0, caractérisée par des processus de fabrication intelligents et connectés, intéresse
de plus en plus les entreprises qui s'efforcent de s'adapter à la demande croissante du marché et des clients.
L'objectif de ce projet de doctorat est d'aider les entreprises à évoluer de la fabrication automatisée traditionnelle
vers la numérisation du processus de fabrication en proposant un nouveau modèle d'intégration de l'Industrie 4.0.
La variété et la complexité des aspects de l'Industrie 4.0, ainsi que les défis techniques et de gestion doivent être
pris en compte pour fournir aux entreprises des approches d'intégration cohérentes sur la base d’un nouveau
Framework d'intégration. L'expérimentation a été réalisée à la plateforme technologique 4.0 des Arts et Métiers,
campus de Paris. Le modèle développé et le Framework proposé ont été vérifiés grâce à un travail intensif
d’expérimentation et de mise en œuvre à la plateforme 4.0.
Mots-clés : Industrie 4.0, Systèmes cyber-physiques (CPS), Internet des objets (IoT), Modèles d'intégration,
Transformation numérique, Framework d'intégration, Flexibilité, Modularité, Déploiement de données, Robotique
avancée, Système d'exécution de fabrication.
Abstract
Industry 4.0 is a dynamic and integrated system for employing control over the entire value chain of the lifecycle
of products. Vertical and horizontal integration and fusion of physical and the virtual components are critical for
Industry 4.0, as well as technology trends implementation such Cyber Physical Systems and Internet of Things as
key enablers of high level of integration at a global scale. Introducing Industry 4.0 as the smart and connected
manufacturing processes enabler is increasingly attracting companies trying hard to adapt to the growing market
and customer demands. Due to the emergence of Industry 4.0 concept, the aim of this PhD thesis is to help
companies to evolve from traditional automated manufacturing to the digital manufacturing level by proposing a
novel integration model of Industry 4.0. The variety and the complexity of Industry 4.0 aspects, along with
technical and managerial challenges are to be taken into consideration to provide companies with successful
integration approaches along with a novel integration framework. The experimentation following this research has
been realized at the platform 4.0 at Arts et Métiers Institute of Technology, Paris. The developed model and the
proposed framework have been verified through an intensive implementation and integration work among the
experimentations done at the industrial platform 4.0.
Keywords: Industry 4.0, Cyber-Physical Systems (CPS), Internet of Things (IoT), Integration Models, Digital
Transformation, Integration Framework, Flexibility, Modularity, Data Deployment, Advanced Robotics,
Manufacturing Execution System.
248