A Model Driven Method to Design Educational Cyber Physical Systems

Samia Bachir

1,2

, Laurent Gallon

, Philippe Aniorte

and Angel Abenia

Universite de Pau et des Pays de l’Adour, E2S UPPA, LIUPPA, Mont-de-Marsan, France

Computer Sciences Laboratory of Le Mans University, LIUM – EA 4023, Le Mans University, IUT de Laval,

Keywords:

Model Driven Engineering, Internet of Things, Cyber Physical Systems, Education, Educational Cyber

Physical System, SysML.

Abstract:

Instructional design is a major concern in TELE (Technology Enhanced Learning Environments) research,

especially since the beginning of the Covid-19 health crisis. Since the beginning of this crisis, emergency

remote teaching has been widely used. Accordingly, the primary objective in these circumstances is not

to re-create a robust educational ecosystem, but rather to provide adapted access to instructional support,

learning materials, services and objects. However, design connectedness in such environments is still required

regarding the emergence of IoT (Internet of things) and CPS (Cyber Physical Systems) in everyday life and

thus in educational environments. In this paper, we propose a model-driven engineering method for the design

of Educational Cyber Physical Systems (ECPS). Our method deals with the separation of concerns when it

comes to considering a Platform Independent Model (educational aspect) and a Platform Description Model

(connected aspect). This practice could then be adopted in order to design further environments by adapting

the required models.

1 INTRODUCTION

IoT (Internet of Things) and CPS (Cyber Physical

Systems) have overwhelmed several domains and are

gaining increasing attention. Research on connected

environments often focus on speciﬁc domains which

may result in a gap in the development between par-

ticular domains and others. For instance, interests in

Industry 4.0 exist because of the mass consumption

society we live in. For that reason, the industrial ﬁeld

has expeditiously moved through several revolutions

to attend the industry 4.0 (Hermann et al., 2016).

The buzzword IoE (Internet of Everything) has,

over the last years, registered a tremendous increase

in use. The connected world does not only mean con-

nected things, it also includes services, humans and

data. In our research work, we investigate the next

revolution at an educational level by exploring the

core technologies of Industry 4.0 (IoT/E, CPS) for

University 4.0. Although several research papers in

the literature have had the interest of connected ped-

agogical environments, many challenges with regard

to educational system policies remain, in particular at

a university level.

The purpose of this paper is to deﬁne and

model Educational Cyber Physical Systems (ECPS)

based on a model driven engineering (MDE) method

through meta-modeling. A meta model describes a

speciﬁc domain model through a modeling language

ezivin, 2005). To do so, we ﬁrst start by deﬁn-

ing the meta-model of CPS within a PDM (Platform

Dependent Model) perspective, then, we deﬁne the

meta-model of the educational environment from a

PIM (Platform Independent Model) perspective and

ﬁnally, we implement a model fusion through ATL

rules (Atlas Transformation Language) which will re-

sult in an ECPS model.

The remainder of this paper is structured as fol-

lows. Section 2 presents the related works covering

the proposed DSLs (Domain Speciﬁc Languages) by

describing IoT, CPS environments and IoT in educa-

tion. Section 3 presents our contribution regarding

both the deﬁnition and the design method of ECPS.

In order to clarify our work, an illustrative case study

will highlight the different concepts, in section 4.

Conclusions and future works will resume the paper.

Bachir, S., Gallon, L., Aniorte, P. and Abenia, A.

A Model Driven Method to Design Educational Cyber Physical Systems.

DOI: 10.5220/0010516401250134

In Proceedings of the 16th International Conference on Software Technologies (ICSOFT 2021), pages 125-134

ISBN: 978-989-758-523-4

125

2 RELATED WORKS

To the best of our knowledge, ECPS are not addressed

in the literature. Only the use of connected things in

education are provided in the research ﬁeld but with-

out any modeling perspective. The corresponding re-

lated works will be presented later. As a result, we

rely on IoT and CPS related works. In the literature,

one may ﬁnd several deﬁnitions of IoT and CPS to

such a degree that they could be considered as look-

alike concepts, however, they are not. More notably, it

was stated in (Guth et al., 2016) that CPS can be used

as a synonym of IoT since both terms have recently

been mentioned coincidentally. Nonetheless, they are

technically highly linked. In this section, we clarify

both deﬁnitions and give a glance of their proposed

models in Software Architecture (SA).

2.1 Internet of Things (IoT)

Various deﬁnitions of IoT can be reviewed in (Muc-

cini and Moghaddam, 2018). For each deﬁnition, a

speciﬁc view point or perspective is considered. In

fact, some research works (e.g (Khan et al., 2012),

(Tyagi, 2016)) highlight the machine-to-machine in-

teractions when qualifying this paradigm. Others fo-

cus on the real-time collaboration and interaction

(Syed et al., 2016).

More generic deﬁnitions are also given in (Roman

et al., 2013) where the authors consider IoT as a

worldwide network of interconnected entities. In (Na-

vani et al., 2017), IoT is seen as the ability to connect,

communicate with, and remotely manage an incalcu-

lable number of networked and automated devices. It

is transforming our physical world into a giant infor-

mation system (Fortino et al., 2017). Also, a deﬁni-

tion given in (Nunes et al., 2017) characterizes IoT as

an ecosystem that interconnects physical objects with

telecommunication networks, joining the real world

with the cyber space and enabling the development

of new kinds of services and applications. We notice

that this latter represents IoT as Cyber Physical Sys-

tems (the next section speciﬁes more what CPS are).

In our work, we consider that IoT is getting an ex-

panded dimension of its things. Currently, we deal

with Everything. And so, IoE includes not only smart

machines (things), but also humans, data and Services

(CISCO, 2013). Interactions are then established be-

tween all these pillars of the IoE. Thus, IoE could be

deﬁned as an infrastructure of a smart world where

CPS could carry out high-performing computational

capacities.

Regarding IoT modeling in SA, a review (Muccini

and Moghaddam, 2018) about IoT architectural styles

notes that the layered style was dominant in the differ-

ent research works. Cloud-based and service-oriented

styles also have an important presence. We believe

that data-oriented architecture is not sufﬁciently de-

veloped in the literature. Object-oriented architecture

(in the sense of smart-object) could also be considered

as a way to IoT architecture.

For instance, the work of (Fortino et al., 2017) fo-

cuses on designing IoT service models which, accord-

ing to the authors, are the real IoT drivers. Each IoT

Service is designed as a composition of the Service

Model (details about attributes and relationships de-

scribing the IoT Service itself) and the Service Proﬁle

(details about the process implementing the service).

The LAURA architecture (Teixeira et al., 2020) is yet

another service-oriented architecture for IoT. Its aim

is to implement changes in business models by taking

into account numerous contextual elements. Mean-

while, various research works are in the pursuit of

standardising IoT reference Architecture (e.g. (Bauer

et al., 2013), (Guth et al., 2016), etc).

Accordingly, an IoT Reference Model (IoT ARM)

widely provides the concepts and deﬁnitions on which

IoT architectures can be built (Bauer et al., 2013).

IoT ARM could be considered as a mature and well-

deﬁned reference model that provides a common

structure and guidelines to deal with the core as-

pects of developing, using and analysing IoT sys-

tems. However, the emphasis is laid on service and

thing (device) modeling. Some research works have

adopted this reference model in order to design and

analyze IoT applications. For example, a SysML pro-

ﬁle was proposed in (Costa et al., 2016) to be used by

IoT application engineers where a veriﬁcation of QoS

properties is presented. This work itself was followed

by (Hussein et al., 2017) to propose model driven de-

velopment adaptive IoT systems.

Towards an open architecture, authors in (Vogel

and Gkouskos, 2017) pointed out some design prin-

ciples (ﬂexibility, customizability, and extensibility)

that should be taken into account in order to control

and guarantee the system evolution over time.

The trend in IoT architecting is highly directed

towards service modeling. However, we are facing

new paradigm such as *aaT (Everything-as-a-Thing)

(Maamar et al., 2018) or IoE. In our work, we are

interested in a ”four-oriented” architecture for IoE. In

the next section, we explore CPS. Then, we deﬁne our

proposed CPS modeling language to build upon ideas

from some related works.

ICSOFT 2021 - 16th International Conference on Software Technologies

126

2.2 Cyber Physical Systems (CPS)

CPS are deﬁned as transformative technologies for

managing interconnected systems between its phys-

ical assets and computational capabilities (Baheti and

Gill, 2011). They monitor, analyse and control the

physical processes and act accordingly. They are con-

sidered as System of Systems (SoS).

So far, architecting CPS is still in its enfancy.

There are presently no standardized or reference mod-

els for such a system. But, SA is currently gaining im-

portant attention of researchers. Looking through the

last 5-years proceedings of the most recognised con-

ference on Software Architecture (ECSA (European

conference on Software Architecture), ICSA (Interna-

tional Conference on Software Architecure), Models

(International Conference on Model Driven Engineer-

ing Languages and Systems), etc.) has allowed us to

select a small number of interesting works gathering

both SA and CPS modeling.

For instance, (Muccini and Sharaf, 2017) propose

CAPS architecture framework for modeling and sim-

ulation of situational-aware CPS which deals with dif-

ferent architectural views of CPS in order to model

the software, hardware and physical space views.

These three proposed modeling views are linked by

two auxiliary views. Modeling the hardware as well

as the physical space is a low-level speciﬁcation of the

concepts we may ﬁnd in the IoT environment models.

The structural aspect in CAPS SA could also be con-

sidered in this way. The Behavioral aspect is globally

deﬁned as a set of actions and events which are highly

related to the domain model application.

A proposed framework was also published in

(Group et al., 2016) in order to develop a CPS anal-

ysis methodology and vocabulary to describe it. One

of the various addressed issues mentioned by the au-

thors was data exchange that is considered as a promi-

nent dimension of a CPS operation. In a recent study

(Kirchhof et al., 2020), further related works on mod-

eling CPS are also presented (e.g ThingML, Mon-

tiArc).

The difference between IoT and CPS is conspic-

uously clear now. The former is highly linked to en-

vironmental issues (hard devices, networking, etc) as

a paradigm. The latter concerns systems in a soft-

ware and digital level (digital twins (Kirchhof et al.,

2020),computational capabilities, control, decision-

making). Yet, both subjects are thoroughly connected

in order to construct a smarter world.

Figure 1 shows the borders among this plethora of

concepts in the composition of CPS.

Figure 1: Main components of Cyber Physical Systems (re-

vised from (de Amorim Silva and Braga, 2018)).

2.3 Educational Connected

Environments

As far as connected environments are concerned,

many research works have had the interest of the

added values of IoT in education. Related works

about educational modeling languages are presented

later in this paper. Several value propositions

(Bagheri and Movahed, 2016) are empowering, di-

rectly or not, students’ achievements. Interesting ef-

forts have been conducted to improve the educational

ecosystem based on this paradigm. We classify them

according to the way IoT are used. We opt for the

following classiﬁcation :

• Learning/Teaching of Internet of Things: this

classiﬁcation concerns the teaching and learning

of IoT as a learning subject. The aim is to

teach/learn the different core knowledge of the

subject like in (Sackey et al., 2017). Education

4.0 could be classiﬁed in this type of IoT use. It

focuses on teaching/learning IoT to prepare future

professionals who will have the required compe-

tences and skills of the subject and then will be

able to work on an IoT equipped environment like

Industry 4.0.

• Learning/Teaching by Internet of Things: this

classiﬁcation concerns the use of Internet of

Things as an artifact to acquire other knowledge.

(Yuqiao and Kanhua, 2016) is an example of re-

search works which focus on such an aspect. Ex-

periment based Learning is one of the conducted

pedagogies that could be viewed as a way to serve

knowledge through the bias of smart object ma-

nipulation.

• Learning/Teaching based on Internet of Things:

This perspective is addressed in some research

works regarding the monitoring of students’

A Model Driven Method to Design Educational Cyber Physical Systems

127

healthcare or in classroom access control, like in

(Palma et al., 2014). Learning analytic techniques

are adopted as a means to provide feedback to the

different stakeholders.

• Another category could be drawn according to

further uses of IoT in the educational context,

which are not directly linked to learning and

teaching. These applications focus on energy

management, enhancing safety, improving com-

fort, etc. (Bagheri and Movahed, 2016).

According to this classiﬁcation, we consider that there

is still a great deal of work to do in order to directly

improve educational processes based on IoT. Mod-

eling an educational connected environment is still

needed. We thus adopt the positioning of our work

on the third categorisation of IoT application (Learn-

ing/teaching based on IoT). This does not exclude the

possible association with the other purposes of IoT

applications.

3 EDUCATIONAL CYBER

PHYSICAL SYSTEMS

3.1 Deﬁnition

Domains and applications of CPS were identiﬁed in

previous studied works (robotics domain, electric ve-

hicles, supervisory system, federal embedded system,

sensor and actuator network, and smart grid). We thus

consider that such an educational application could

also be viewed as a supervisory system aiming at

gathering real-time data from connected educational

systems in order to constantly monitor and make deci-

sions about not only the adaptation of curriculum for

students but also other educational oriented decisions.

We deﬁne what we call Educational Cyber Phys-

ical Systems (ECPS) as Cyber Physical Systems ap-

plied to the world of Education. An ECPS is a set

of objects, services, data and humans, that collabo-

rate to carry out a teaching and learning scenario in a

short/medium/long term context (Classroom, Course,

and Curriculum). For the best of our knowledge, Edu-

cational Cyber Physical System are not yet emerging

in the scientiﬁc community.

We consider that ECPS consists in the instruc-

tional design to occur in a connected environment

gathering learners, teachers, admin, learning object

(physical and virtual), technological object, data, etc

in order to collect, analyse data and make decision

according to the learning and teaching context, in

other words, educational purpose in IoE environment.

Based on (Lee et al., 2015), we published, in a pre-

vious work, a generic architecture of ECPS (Bachir

et al., 2019). In this paper, we detail the modeling of

pedagogical situations in IoE environment.

3.2 Method

(Costa et al., 2016) identiﬁed some design challenges

related to IoT system. We believe that they remain

the same for IoE system. They concern (i) the het-

erogeneity of hardware devices and software compo-

nents; (ii) the lack of mechanisms to address multiple

stakeholders’ concerns; and, (iii) lack of a method to

design IoT applications. In our work, we provide a

method for the design ECPS. As stated in (Ramsin

and Paige, 2008) and according to the (OMG, 2000),

a method, in software engineering, consists of : (1)

a modeling language : a set of modeling conventions

(syntax and semantics); and (2) a process.

Our aim is to propose a method to deﬁne and

model ECPS based on MDE. MDE is a software de-

velopment method that uses its core models not only

as inputs but also as outputs in order to reduce the gap

between problem domain and solution domain. It is

more general than the set of standards and practices

recommended by the OMG’s MDA proposal (Kurtev

et al., 2006). It handles separation and combination

of various kinds of concerns (such as platform models

and code generation) in software engineering in order

to bridge the gap between design and implementation.

Domain Speciﬁc Languages (DSLs) are the com-

plementary part of MDE which is considered as the

main means to deﬁne models. DSLs are languages

tailored for a speciﬁc application domain. In contrast

to general-purpose languages, they offer substantial

gains in expressiveness and ease of use (Mernik et al.,

2005). We will also deﬁne DSLs to explore their dif-

ferent parts (concrete syntax, abstract syntax, and se-

mantics) and beneﬁt from model generation features.

We adopt the classical “Y development cycle”

from MDA as the process of our method, as illus-

trated in Figure 2. The objective of this process, ac-

cording to (B

ezivin, 2005), is to enable the generation

of Platform Speciﬁc Models (PSMs) from Platform

Independent Models (PIMs) (speciﬁc to the business

model) and Platform Description Models (PDMs).

Aiming a higher ﬂexibility, the goal is also to pro-

duce software assets that are resilient to changes in

technologies. MDA stresses on the importance of

PIMs since they are supposed to survive the constant

changes in software technologies that are transformed

into new PSMs (Kurtev et al., 2006). In contrast,

PDM are highly linked to the kind of platforms (con-

nected in our case). In an educational purpose, we

ICSOFT 2021 - 16th International Conference on Software Technologies

128

Figure 2: Modeling process.

can deploy the same pedagogical scenario in different

platforms (something we have noticed during lock-

downs). In this case, the Y development cycle is a

real advantage, allowing us to redeﬁne only the PDM

model, with no change in the educational one (PIM).

Furthermore, several pedagogical situations can be

modeled in the same environment, only by changing

the PIM model.

In this section, we present different meta models

according to these different perspectives. In the in-

terest of simplicity, the concepts are depicted without

cardinalities. Examples will be provided for further

elucidation of the proposed models.

3.3 CPSML

We begin by presenting the PDM perspective as men-

tioned in Figure 2 (1). We propose a speciﬁc model-

ing language, called CPSML (Cyber Physical System

Modeling Language) in order to design a connected

environment regardless the domain model. CPSML

is a SysML proﬁle. According to OMG, SysML is

a general-purpose modeling language for systems en-

gineering. In turn, it is considered as a UML proﬁle

where the former (e.g Block diagram) is an extended

customization of the latter (Class diagram).

SysML is a highly referenced modeling language

among the scientiﬁc community. It handles both

structural and behavioral perspectives in the modeling

of speciﬁc system features. It has proved its efﬁciency

and re-usability. CPS is a system of systems where

the components within a system interact between each

other or with other systems. So it is crucial that our

proposal takes into account both view points.

We extend the block diagram by specifying the

block concept through different stereotypes, as speci-

ﬁed in Figure 3. According to OMG, blocks are mod-

ular units of system description. Each block deﬁnes

a collection of features to describe a system or other

element of interest. A stereotype is a generalization

or speciﬁcation of a predeﬁned metaclass. To do so,

we implemented our CPSML proﬁle, in the Papyrus

modeling tool, in order to customize SysML in the

Eclipse Modeling Framework (EMF). The idea is to

allow the designer to specify the nature of the used

blocks (object, service, human, data) which refer to

the main components of CPS illustrated in Figure 1,

thus representing all the aspects in a connected envi-

ronment. For instance, in an educational context, the

proposed stereotypes are:

• ”Human” refers to the students, teachers, admin-

istrators or other stakeholders who are involved

in the learning/teaching processes. This knowl-

edge about user proﬁle is essential to create per-

sonal and professional linkages by incentivizing

collaboration and cooperation. As it is stated, peo-

ple themselves will become nodes on the Internet,

with both static information and a constantly emit-

ting activity system.

• ”Object” refers not only to physical devices that

can establish connection with the Internet and uti-

lize sensors to capture environment information

(as it is presented in (de Amorim Silva and Braga,

2018)) or actuators to act on the environment, but

also to smart learning resources that can estab-

lish connection with the Internet (e.g computer,

telepresence-robot, etc.).

• ”Service” refers to how people and objects must

interact to generate data that can be transformed

into usable knowledge through service invocation.

According to (Fortino et al., 2017), each IoT Ser-

vice is designed as a composition of the Service

Model (details about attributes and relationships

describing the IoT Service itself) and the Service

Proﬁle (details about the process implementing

the service).

• ”Data” refers to all the different data ﬂows com-

ing from the different elements of the connected

system. Cloud computing and learning analyt-

ics are examples of technologies for data man-

agement and its transformation into information.

Data could be a structured, semi-structured, or

non-structured in a connected environment.

Likewise, we extend the activity diagram to add Au-

tonomic Computing (AC) (IBM, 2006) features and

specify action types regarding CPS control, analysis

and computational capabilities. AC covers different

aspects of self-management that implement MAPE

control loops (Monitor, Analyse, Plan and Execute)

which collect details from the system and act accord-

ingly. More precisely, we extend the Action con-

cept (which represents the behavior of the system) in

SysML by adding both Analyzing Action and Moni-

toring Action stereotypes. We believe that they play

A Model Driven Method to Design Educational Cyber Physical Systems

129

Figure 3: CPS Modeling Language.

Figure 4: Educational Modeling Language for University 4.0.

an important role throughout the design process of the

connected environment such as to represent the auto-

nomic properties of the CPS. Other types of actions

like ’decision-making’ could be simply designed as

Action with the SysML concept. In such systems, we

also deal with events. For this purpose, it is also im-

portant to explicitly model them. In CPSML, Event is

modeled as a stereotype of the Object Node.

3.4 EML4.0

In the literature, modeling education has taken sev-

eral dimensions. Various research works have been

conducted since the early 2000s especially with the

emergence of online learning via, for instance, Moo-

dle platforms. One may consider an educational

modeling language according to a particular perspec-

tive. There are some research works that design

learning through Activity Modeling Language (e.g.

E2ML (Botturi, 2003), ColeML (Martel et al., 2006)).

Others design learning through Content Structuring

Language (e.g. PoEML (Caeiro-Rodr

ıguez, 2008),

(SCORM, 2003)). Some others focus on both per-

spectives, Activity and Content Structuring Lan-

guages, (e.g. IMS LD (Koper and Olivier, 2004),

CoUML (Derntl and Motschnig-Pitrik, 2008)). Fur-

thermore, there are modeling languages which con-

cern rather evaluation modeling (e.g (QTI, 2005) ).

In our work, we build upon ideas from Activity

and Content Structuring Language, in order to ensure

ICSOFT 2021 - 16th International Conference on Software Technologies

130

Figure 5: ECPS Modeling Language.

both structural and behavioral aspects of educational

modeling in a learning process. For this purpose,

we adopt IMS-LD (Instructional Modeling System -

Learning Design) as it is an IMS standard. As pre-

sented in Figure 4, we extend IMS-LD with further

concepts. Some of the latters are inspired from (Ruiz

et al., 2018) regarding presence mode and user role

in a learning activity. We also upgrade the ”Environ-

ment” concept in order to support any type of TELS

(ITS, MOOC, Moodle, CSCL, etc.).

EML4.0 corresponds to (2) in Figure 2. IMS-LD

is a speciﬁcation implementation based on an XML

Schema. To the best of our knowledge, there is no

standardised tool for it. Many attemps have been con-

ducted in order to offer a required tool to design learn-

ing according to the IMS speciﬁcation. But, they are

not open-source and cannot support extensions. As

we are in a model driven approach implemented with

EMF, we realized an EML4.0 modeling tool from

scratch in this framework.

3.5 ECPSML

In our research, we adopt a MDE method in order to

design ECPS. The latter is deﬁned and build based

on both CPS and EML4.0 meta models. We pro-

pose ECPS as a SysML proﬁle too, as shown in Fig-

ure 5. ECPS integrates structural concepts coming

from both CPS and pedagogical meta-models. For

Figure 6: ATL transformation rules.

instance, Object could be either technological (from

(1)) or pedagogical (from (2)). From a behavioral

perspective, learning activity is a type of action in a

SysML proﬁle.

The generation of our ECPS model will be au-

tomatically handled by the ATL module. ATL is a

model transformation language (Allilaire and Idrissi,

2004). It is integrated in EMF which, in addition to its

modeling framework features, provides facility model

generation building tools and other applications based

on the different XMI (XML Metadata Interchange)

models.

Rules are the heart of ATL transformations.

A Model Driven Method to Design Educational Cyber Physical Systems

131

They describe how output elements (ECPS) are pro-

duced from input elements based on the input meta-

model(s). Input models could be multiple which is the

case of this work (CPSML, EML4.0). Each rule ex-

presses a mapping between a multiple input element

and an output element. Figure 6 presents parts of our

ATL rules.

Figure 7: Parts of the CPS model elements within a

connected learning environment (Monitoring Action (light

grey), Analyzing Action (blue), Object (orange)), Data (yel-

low), Human (dark grey), other activities regarding decision

making (white)).

4 CASE STUDY

The aim of this section is not to represent a full

study case, but rather to illustrate the generation of

the ECPS model from both CPS and EML4.0 mod-

els through transformation rules. Also, due to limited

space, the above will be simply presented through ac-

tivity diagrams illustrating the different elements. The

case study was implemented and tested in EMF by

deﬁning input models as dynamic instances.

We present both structured and behavioral aspects

of the different scenarios inspired from observations

conducted during the ﬁrst lockdown (March - April

2020) at our university with bachelor students, in a

”Cyber Security initiation” subject, in order to collect

and analyse their logs. In the CPS model, monitor-

ing and analyzing activities are highlighted (Figure 7).

Two control loops (MAPE) were implemented, each

of which is linked to some parts of the pedagogical

Figure 8: Parts of the EML4.0 model elements within a

learning scenario (Learning Activity (blue), Learning Ob-

ject (yellow), Outcome (dark grey), TELS (orange),other

activities regarding the pedagogical scenario (white)).

scenario presented in the Figure 8.

The ﬁrst MAPE control loop aims at conﬁguring

the CPS by detecting the telepresent students and de-

ploying a telepresent robot or a visio in order to ensure

their involvement. The presence of the other students

is controlled by reading their student card with the

RFID reader. The second one monitors the CPS dur-

ing the collaborative activity, by collecting and ana-

lyzing students discord logs. Time was noted between

both control loops.

Regarding this, we focus on the domain model

scenario of some learning activities of the ”Cyber Se-

curity” knowledge-to-be-taught. This example is built

upon teaching strategies adopted during the lock-

down. Figure 8 presents the pedagogical scenario

which starts by checking students’ presence (handled

by the MAPE 1). Then a brainstorming activity is

followed in order to introduce ”Cyber Security” con-

cepts for students. After this, the students are asked

to work collaboratively on different projects (handled

by the MAPE 2) to ﬁnally present them and move to

an evaluation activity.

Figure 10 shows the ECPS model, obtained by fu-

sioning the two previous models, using the transfor-

ICSOFT 2021 - 16th International Conference on Software Technologies

132

Figure 9: Parts of the ECPS model generated through the

transformation rules given in Figure 6.

Figure 10: ECPS model.

mation rules described previously. A part of the XMI

generated model is showed in Figure 9 where techno-

logical objects are taken from the CPSML model and

pedagogical ones from EML4.0 model.

5 CONCLUSIONS

The objective of this paper is to introduce a model-

based engineering methodology for Educational Cy-

ber Physical Systems, focusing on the design phase,

which aims to help educators to develop IoE ap-

plications to fully achieve the beneﬁts of this new

paradigm in teaching and learning. The approach

comprises a method and three modeling tools, which

are aligned with proven concepts from system engi-

neering and software architecture literature. Our pro-

posal ensures reusability of the proposed modeling

languages (CPSML, EML4.0, ECPSML). They are

languages tailored for a speciﬁc application domain.

They offer substantial gains in expressiveness, ease

of use and beneﬁt from model generation features.

Our ﬁrst perspective is to test the genericity of our

method by changing the educational model and deal-

ing with the same CPS model and vice-versa. There-

fore, we could validate the fact of acting on either

PIM or PDM in order to generate the required educa-

tional environment. Still, our second perspective is to

adapt this method to model other environments (smart

grid, smart building, medical environment, etc.). Our

aim is also to study another case in an other domain,

in order to generate the required model by acting on

the domain model and transformation rules.

REFERENCES

Allilaire, F. and Idrissi, T. (2004). Adt: Eclipse develop-

ment tools for atl. In Proceedings of the Second Euro-

pean Workshop on Model Driven Architecture (MDA)

with an emphasis on Methodologies and Transforma-

tions. Citeseer.

Bachir, S., Gallon, L., Abenia, A., Aniorte, P., and Expos-

ito, E. (2019). Towards autonomic educational cyber

physical systems. In SmartWorld, Ubiquitous Intelli-

gence & Computing, pages 1198–1204. IEEE.

Bagheri, M. and Movahed, S. H. (2016). The effect of the

internet of things (iot) on education business model. In

2016 12th International Conference on Signal-Image

Technology & Internet-Based Systems (SITIS). IEEE.

Baheti, R. and Gill, H. (2011). The impact of control tech-

nology: Cps. IEEE Control Systems Society.

Bauer, M., Bui, N., De Loof, J., Magerkurth, C., Nettstr

ater,

A., Stefa, J., and Walewski, J. W. (2013). Iot reference

model. In Enabling Things to Talk. Springer.

ezivin, J. (2005). On the uniﬁcation power of models.

Software & Systems Modeling, 4(2).

Botturi, L. (2003). E2ml-educational environment model-

ing language. In EdMedia+ Innovate Learning. Asso-

ciation for the Advancement of Computing in Educa-

tion (AACE).

Caeiro-Rodr

ıguez, M. (2008). poeml: A separation of con-

cerns proposal to instructional design. In Handbook

A Model Driven Method to Design Educational Cyber Physical Systems

133

of visual languages for instructional design: theories

and practices. IGI Global.

Costa, B., Pires, P. F., Delicato, F. C., Li, W., and Zomaya,

A. Y. (2016). Design and analysis of iot applications:

a model-driven approach. In 2016 IEEE 14th Intl Conf

on Dependable, Autonomic and Secure Computing,

(DASC/PiCom/DataCom/CyberSciTech). IEEE.

de Amorim Silva, R. and Braga, R. T. (2018). An acknowl-

edged system of systems for educational internet of

everything ecosystems. In Proceedings of the 12th

ECSA: Companion Proceedings.

Derntl, M. and Motschnig-Pitrik, R. (2008). couml: A vi-

sual language for modeling cooperative environments.

In Handbook of visual languages for instructional de-

sign: Theories and practices. IGI Global.

Fortino, G., Russo, W., Savaglio, C., Viroli, M., and Zhou,

M. (2017). Modeling opportunistic iot services in

open iot ecosystems. In WOA.

Group, C. P. S. P. W. et al. (2016). Framework for cyber-

physical systems, release 1.0. Report, National Insti-

tute of Standards and Technology.

Guth, J., Breitenb

ucher, U., Falkenthal, M., Leymann, F.,

and Reinfurt, L. (2016). Comparison of iot platform

architectures: A ﬁeld study based on a reference ar-

chitecture. In 2016 Cloudiﬁcation of the Internet of

Things (CIoT). IEEE.

Hermann, M., Pentek, T., and Otto, B. (2016). Design prin-

ciples for industrie 4.0 scenarios. In System Sciences

(HICSS). IEEE.

Hussein, M., Li, S., and Radermacher, A. (2017). Model-

driven development of adaptive iot systems. In MOD-

ELS (Satellite Events).

IBM (2006). An architectural blueprint for autonomic com-

puting. IBM White Paper, 31.

Khan, R., Khan, S. U., Zaheer, R., and Khan, S. (2012). Fu-

ture internet: the internet of things architecture, pos-

sible applications and key challenges. In 2012 10th

international conference on frontiers of information

technology. IEEE.

Kirchhof, J. C., Michael, J., Rumpe, B., Varga, S., and

Wortmann, A. (2020). Model-driven digital twin

construction: synthesizing the integration of cyber-

physical systems with their information systems. In

Proceedings of the 23rd ACM/IEEE International

Conference on Models.

Koper, R. and Olivier, B. (2004). Representing the learning

design of units of learning. Journal of Educational

Technology & Society, 7(3).

Kurtev, I., B

ezivin, J., Jouault, F., and Valduriez, P. (2006).

Model-based dsl frameworks. In Companion to the

21st ACM SIGPLAN Symposium on Object-oriented

Programming Systems, Languages, and Applications,

OOPSLA ’06, New York, NY, USA. ACM.

Lee, J., Bagheri, B., and Kao, H.-A. (2015). A cyber-

physical systems architecture for industry 4.0-based

manufacturing systems. Manufacturing letters, 3.

Maamar, Z., Faci, N., Sellami, M., Ugljanin, E., and Kajan,

E. (2018). Everything-as-a-thing for abstracting the

internet-of-things. In ICSOFT.

Martel, C., Vignollet, L., Ferraris, C., and Durand, G.

(2006). Ldl: a language to model collaborative learn-

ing activities. In Proc. of the 2006 World Confer-

ence on Educational Multimedia, Hypermedia and

Telecommunications (ED-MEDIA’2006).

Mernik, M., Heering, J., and Sloane, A. M. (2005). When

and how to develop domain-speciﬁc languages. ACM

Comput. Surv., 37(4).

Muccini, H. and Moghaddam, M. T. (2018). Iot architec-

tural styles. In European Conference on Software Ar-

chitecture. Springer.

Muccini, H. and Sharaf, M. (2017). Caps: Architecture de-

scription of situational aware cyber physical systems.

In 2017 IEEE ICSA. IEEE.

Navani, D., Jain, S., and Nehra, M. S. (2017). The internet

of things (iot): A study of architectural elements. In

2017 13th International Conference on Signal-Image

Technology & Internet-Based Systems (SITIS). IEEE.

Nunes, L. H., Estrella, J. C., Perera, C., Reiff-Marganiec, S.,

and Botazzo Delbem, A. C. (2017). Multi-criteria iot

resource discovery: a comparative analysis. Software:

Practice and Experience, 47(10).

Palma, D., Agudo, J., S

anchez, H., and Mac

ıas, M. (2014).

An internet of things example: Classrooms access

control over near ﬁeld communication. Sensors.

QTI, I. (2005). Ims global learning consortium: Ims ques-

tion & test interoperability: Ims question & test inter-

operability speciﬁcation.

Ramsin, R. and Paige, R. F. (2008). Process-centered

review of object oriented software development

methodologies. ACM Computing Surveys (CSUR).

Roman, R., Zhou, J., and Lopez, J. (2013). On the features

and challenges of security and privacy in distributed

internet of things. Computer Networks, 57(10).

Ruiz, A., Panach, J. I., Pastor, O., Giraldo, F. D., Arciniegas,

J. L., and Giraldo, W. J. (2018). Designing the didactic

strategy modeling language (dsml) from pon: An ac-

tivity oriented eml proposal. IEEE Revista Iberoamer-

icana de Tecnologias del Aprendizaje, 13(4).

Sackey, S. M., Bester, A., and Adams, D. (2017). Industry

4.0 learning factory didactic design parameters for in-

dustrial engineering education in south africa. South

African Journal of Industrial Engineering, 28(1).

Syed, M. H., Fernandez, E. B., and Ilyas, M. (2016). A

pattern for fog computing. In in the 10th Travelling

Conference on Pattern Languages of Programs.

Teixeira, S., Agrizzi, B. A., Pereira Filho, J. G., Rossetto,

S., Pereira, I. S. A., Costa, P. D., Branco, A. F., and

Martinelli, R. R. (2020). Laura architecture: To-

wards a simpler way of building situation-aware and

business-aware iot applications. Journal of Systems

and Software, 161.

Tyagi, N. (2016). A reference architecture for iot. Interna-

tional Journal of Computer Engineering and Applica-

tions, 10(I).

Vogel, B. and Gkouskos, D. (2017). An open architecture

approach: Towards common design principles for an

iot architecture. In Proceedings of the 11th ECSA.

Yuqiao, Y. and Kanhua, Y. (2016). Construction of distance

education classroom in architecture specialty based on

internet of things technology. International Journal of

Emerging Technologies in Learning (iJET), 11(05).

ICSOFT 2021 - 16th International Conference on Software Technologies

134