EMBED-SoSE: Drawing a Cyber-physical System of Systems

Andr

e A. S. Ivo

1,2 a

, Sheila G. Ribeiro

2 b

and Fatima Mattiello-Francisco

2 c

National Center for Monitoring and Alert Natural Disasters (CEMADEN), SJCampos, SP, Brazil

National Institute for Space Research (INPE), SJCampos, SP, Brazil

Keywords:

Cyber-physical Systems, Internet of Things, System Engineering, Model-based System Engineering.

Abstract:

The general availability of a robust communication infrastructure facilitates and encourages the connection

between everything and anything (Internet of Things-IoT and Cyber-Physical Systems-CPS). The composition

of CPSs is a reality. Every day new systems are created from arrangements of legacy systems, referred to as

CP system of systems (CPSoS). The constituent systems of a CPSoS have very peculiar characteristics, which

hinder traditional systems engineering techniques. In addition, it is common for these systems to present many

intangible aspects, which further increases the complexity, especially in the representation of their design.

This work proposes EMBED-SoSE, an approach to the engineering of critical systems based on shared visual

thinking, which aims to facilitate and reduce the cognitive effort in the systems design process.

1 INTRODUCTION

The ease of interconnecting pre-existing standalone

computer systems leverages availability and provides

a new set of services to users. Such a composition

of independent and autonomous systems is called the

“System of Systems”.

Technologies based on the Internet of Things (In-

ternet of Things (IoT)) are considered, by some au-

thors, classic examples of SoS. In general, the IoT

strives to facilitate the connection of Everything (IoE)

and Anything (IoA) via the Internet as a highly com-

plex SoS set (Bojanova et al., 2014).

Speciﬁcally, IoT focuses on the integration of

smart devices, where each element can be considered

an independent system, with sensors and actuators

that directly observe and inﬂuence the physical envi-

ronment, with the potential to provide new disruptive

services to our society (Bojanova et al., 2014).

The diffusion of IoT with many connected devices

consolidated the Cyber-Physical System (CPS) con-

cept. The Cyber component controls the Physical

component because the cyber component has some

“intelligence” or, at least, some strategy to lead the

physicist to reach a predeﬁned goal. However, the

term is used indiscriminately and has become a ref-

https://orcid.org/0000-0001-6192-7705

https://orcid.org/0000-0002-0565-843X

https://orcid.org/0000-0002-0796-3868

erence to designate any system that interacts with

the physical environment, which is not true (Carreira

et al., 2020). Having identiﬁable cyber and physical

components is not enough for a system to be consid-

ered cyber-physical; not every IoT is a CPS. So, what

characterizes a CPS? It is well accepted that CPS con-

sists of a large-scale SoS, with a large number of com-

ponents (Constituent Systems (CS)), interconnected,

highly adaptive, and that includes human and socio-

technical systems (Carreira et al., 2020). More appro-

priately, these large-scale systems are called Cyber-

Physical System of Systems (CPSoS).

Emergent behavior is a fascinating and, at the

same time, a disturbing phenomenon in SoS. Accord-

ing to Maier (1998), the operational independence of

CSs and emergent behavior is noted as a common

characteristic of SoS (Maier, 1998). This feature,

which only becomes active or visible when CSs start

to cooperate, can provide new disruptive services to

our society. On the other hand, risks are associated

with unexpected or unintentional behavior resulting

from the combination of systems that have individu-

ally complex behavior.

As in traditional systems, CPSoS, in a higher de-

gree of complexity, can be represented by a set of tan-

gible and intangible aspects that provide the delivery

of actual or perceived values to satisfy the business’s

needs. The intangibles represent the major engineer-

ing and ﬁnancial efforts required to design and main-

tain many of today’s complex systems. Tangible and

Ivo, A., Ribeiro, S. and Mattiello-Francisco, F.

EMBED-SoSE: Drawing a Cyber-physical System of Systems.

DOI: 10.5220/0011082600003176

In Proceedings of the 17th International Conference on Evaluation of Novel Approaches to Software Engineering (ENASE 2022), pages 485-493

ISBN: 978-989-758-568-5; ISSN: 2184-4895

485

intangible are, linguistically, terms that refer to ele-

ments that may or may not be touched. A car or a

satellite, for example, are tangible, while the project

plan of these systems, although they can be docu-

mented, has an incorporeal existence – therefore, in-

tangible.

The formulation of emergent behaviors can be de-

scribed by the interaction between intangible aspects

of the CSs, which can deliver a new service that will

undoubtedly be intangible from the systems engineer-

ing point of view. Intangible aspects increase the

complexity and cognitive effort to design and develop

a CPSoS free of ambiguity.

The considerable cognitive effort required to un-

derstand, document, and design a complex system

such as CPSoS is intrinsically linked to the difﬁculty

of representing and communicating intangible aspects

to project stakeholders. According to Wujec (2009),

to deﬁne the meaning of something, our brain creates

a series of mental models based on the experiences of

each individual; that is, each engineer will be able to

create his materialization of the intangible aspects of

the system. As a result, each project stakeholder will

create inﬁnite possibilities for materializing the same

intangible aspect, undoubtedly imprecise, uncertain,

and full of gaps.

In general, non-functional requirements are ad-

dressed by traditional Systems Engineering (SE) and

its successors such as Model-Based System Engi-

neering (MBSE); however, the intangible aspects of

a complex system such as CPSoS do not receive

the same attention due to the difﬁculty in expressing

them. The concern with the intangible aspects is la-

tent in Systems Engineering (SE) and methods that

can be more effective in understanding, communicat-

ing, documenting, and designing a system that deliv-

ers practical value to the business.

In order to meet this need, and considering that

emerging behaviors pose the greatest challenges for

CPSoS design, this work presents the EMergent

BEhavior-Driven System of Systems Engineering

(EMBED-SoSE). The main objective of EMBED-

SoSE is to bring awareness of emergent behavior to

decrease the cognitive effort to represent intangible

aspects dynamics and evolution to CPSoS design, to

establish a solid conceptual model that provides: a

well-deﬁned overview to describe CPSoS, to help in-

vestigate emerging behavior and thus create an incre-

mental and iterative cycle of analysis and develop-

ment to support CPSoS and its maintenance.

The EMBED-SoSE a methodological approach to

Systems Engineering (SE) based on two (2) princi-

ples: i) to guide engineers in a design concerned with

emergent behaviors, one of the CPSoS/SoS charac-

teristics that most impose intangible aspects. ii) sup-

port Systems Engineering (SE) through the paradigm

of shared visual thinking, with the primary objective

of reducing the cognitive effort necessary to under-

stand the intangible aspects of the project and trans-

form them into collective knowledge. This work in-

troduces the new method and is limited to presenting

only the ﬁrst of the processes (EMBED-SoSE Discov-

ery Canvas).

The main objective of EMBED-SoSE is to bring

awareness of emergent behavior to decrease the cog-

nitive effort to represent dynamics and evolution to

CPSoS design, to establish a solid conceptual model

that provides: a well-deﬁned overview to describe

CPSoS, to help investigate emerging behavior and

thus create an incremental and iterative cycle of anal-

ysis and development to support CPSoS and its main-

tenance.

This article is organized as follows: Section 2 con-

cerns the theoretical basis of the article, brieﬂy dis-

cussing the main challenges in a CPSoS; Section 3

presents the main related works; Section 4 presents

how the proposed approach (EMBED-SoSE) empow-

ers systems engineering; Section 5 presents its ap-

plication in a real case of CPSoS, the National Cen-

ter for Monitoring and Alert Natural Disasters (CE-

MADEN); Finally, Section 6 presents the Conclu-

sions and Future Work.

2 CPSoS CHARACTERISTICS

CPSoS are complex to build due to the inherent com-

plexity of the problems they are intended to solve. In

general, they consist of coordinating actions to opti-

mize multiple (possibly contradictory) objectives in

systems with a large number of sensors and actuators.

The other source of complexity is more of an acciden-

tal nature and has to do with the heterogeneity of the

systems that constitute it (Carreira et al., 2020).

CPSoS are considered a class of SoS and inherit

all their characteristics and pose new challenges in

forming a system. CPSoS, according to Lee (2008),

are integrations of computing and physical processes.

Embedded computers and networks monitor and con-

trol physical processes, often with feedback loops

in which physical processes affect computations and

vice versa. In other words, CPSoS uses the cyber

component to interact with physical processes in or-

der to add new resources to physical systems (Lee,

2008) (Tr

ols et al., 2021).

SoS is the set of independent systems cooperating

with a common goal. Independent systems are techni-

cally known as Constituent Systems (Constituent Sys-

ENASE 2022 - 17th International Conference on Evaluation of Novel Approaches to Software Engineering

486

tems (CS)).

The ISO/IEC/IEEE 21839 (ISO, 2019) stan-

dard provides a deﬁnition of SoS and CS:

(a) System of Systems (SoS): Set of systems or

elements of the system that interact to provide a

unique capacity that none of the constituent systems

can realize on their own. Deﬁnition extensible to

CPSoS. (b) Constituent Systems (CS): Constituent

systems can be part of one or more SoS. Each

constituent is an independent and useful system in its

own right, having its own development, goals, and

management resources, but it interacts within the SoS

to provide the unique capability of SoS. Deﬁnition

extensible to CPS.

Emergent behavior is noted as a common feature

of SoS(Maier, 1998). Kopetz (2016) uses a quote at-

tributed to Aristotle to describe the fascinating emerg-

ing phenomenon: “The whole is greater than the sum

of its parts” (Kopetz et al., 2016). The interactions

of the “parts” can generate a “whole” with properties

that go beyond those of any of its constituent “parts”.

Moreover, this property makes SoS a system capable

of breaking down some barriers of knowledge, gen-

erating situations/opportunities that were hitherto un-

known.

We use the term “phenomenon” to refer to emer-

gent structure, behavior, or property. In many cases,

these phenomena are described and explained only

after their discovery, which in most cases are acci-

dental. Formulating all emergent phenomena is an al-

most impossible task and requires much cognitive ef-

fort and ﬁnancial resources, and indeed, the system’s

dynamism would lead us to fail in this mission. The

ﬁrst appearance of an emergent phenomenon is often

a surprise to an observer (Kopetz et al., 2016). Fig.

1 shows a scheme for the classiﬁcation of emergent

phenomena.

Figure 1: Emerging phenomena classiﬁcation (Kopetz et al.,

2016).

While this “phenomenon” is explainable in prin-

ciple, we may not be able to explain or predict this

behavior in practice due to our ignorance of the full

scope of CPSoS, the precise temporal interactions be-

tween CS and communication channels hidden be-

hind the interfaces of a CS (Kopetz et al., 2016).

2.1 Challenges

In the CPSoS project, not all combinations allowed by

Fig. 1 are of interest; in fact, we are particularly in-

terested in the domain of behavior, that is, behavioral

emergence. Fig. 2 classiﬁes the emergent behavior of

a CPSoS from the point of view of the consequences

of this behavior in the general mission of a CPSoS

and the prediction or awareness we can have about

the emergence of emergent behavior.

Figure 2: Emergent behavior consequences(Kopetz et al.,

2016).

The expected and beneﬁcial emergent behavior is

the normal case (quadrant 1) that results from a con-

scious design effort. Expected detrimental emergent

behavior can be avoided by adhering to proper design

rules (quadrant 2). Unexpected and beneﬁcial emer-

gent behavior is a positive surprise (quadrant 3). The

problematic case is quadrant 4, unexpected detrimen-

tal emergent behavior (Kopetz et al., 2016). In safety-

critical CPSoSs, unexpected detrimental emergent be-

havior can cause a catastrophic accident.

A conscious design discipline aims to move, as

knowledge progresses, more and more emergent phe-

nomena from quadrant 4 to quadrant 2, where mea-

sures can be taken to mitigate, eliminate or prevent

the harmful emergence (Kopetz et al., 2016).

Still, our knowledge of CPSoS can remain lim-

ited, and our ignorance of them can hardly be reduced

sufﬁciently, especially when considering COTS com-

ponents and legacy constituent systems. Most CPSoS

are built by incorporating LEGACY and COTS about

which very little is known and where the ﬂow of in-

formation is often quite hidden (Kopetz et al., 2016).

3 RELATED WORKS

Authors have presented works intending to reduce the

complexity and the cognitive effort necessary to un-

derstand the behavior of a SoS through the applica-

tion of appropriate simpliﬁcation strategies (Kopetz,

2011) (Ceccarelli et al., 2016). The considerable cog-

nitive effort required to understand how a large SoS

works the primary engineering and ﬁnancial effort re-

quired to design and maintain many of today’s SoSs.

The main works/projects that have produced advances

EMBED-SoSE: Drawing a Cyber-physical System of Systems

487

in the CPSoS engineering process are presented be-

low:

1. COMPASS - (2014): The Comprehensive

Modeling for Advanced Systems of Systems (COM-

PASS) project was ofﬁcially completed in 2014. The

proposed approach was based on the relationships and

guarantees of the constituent systems that are explic-

itly recorded in a formal language (“COMPASS Mod-

eling Language” (CML)) and analyzed by new tools

that exploit the formality of the semantics of the CML

to assist in the analysis and guarantee of SoS proper-

ties (Holt et al., 2014); 2. DANSE - (2015): The

“Designing for Adaptability and evolution in System

of Systems Engineering” (DANSE) project was ofﬁ-

cially concluded in 2015. The DANSE project de-

veloped a set of methodology and tools for the tech-

nical management of an SoS. DANSE’s methodol-

ogy and tools are mainly based on the frameworks

DoDAF, MoDAF and NAF (Mangeruca et al., 2013);

3. AMADEOS - (2016): The project “Architec-

ture for Multi-criticality Agile Dependable Evolution-

ary Open System-of-Systems” (AMADEOS) had as

the main objective of the AMADEOS project was

to bring awareness of time, dynamics, and evolu-

tion to design of SoS, to establish a solid concep-

tual model that provides: a well-deﬁned language to

describe SoS, to investigate emergent behavior; out-

line a generic architectural framework; and an SoS

design methodology that is supported by modeling

tools (Bondavalli et al., 2016); 4. NIST - Frame-

work for Cyber-physical Systems - (2017): The

CPS Framework was developed by the Cyber-Phys-

ical Systems Public Working Group (CPS PWG), an

open public forum established by the National Insti-

tute of Standards and Technology (NIST). The ulti-

mate goal of the CPS Framework is to provide a com-

mon language to describe interoperable CPS architec-

tures so that these systems can interoperate within and

across domains, allowing the formation of an SoS;

5. MPM4CPS - Multi-paradigm Modeling for Cy-

ber-physical Systems (2018): The “Multi-Paradigm

Modeling for Cyber-Physical Systems” (MPM4CPS)

project was ofﬁcially completed in 2018. In 2020 the

book “Foundations of Multi-Paradigm Modeling for

Cyber-Physical Systems” was published, compiling

the design results, which is based on the fact that there

is no super formalism to support the multiple design

dimensions of a CPSoS, were to design effectively,

engineers (in the role of modelers) need to be versed

in multiple formalism’s (Carreira et al., 2020).

Table 1 demonstrates the comparison between the

projects and emphasizes that the EMBED-SoSE does

not intend to replace any of the proposals of the

projects but rather to cover the gap regarding the con-

Table 1: Projects comparison.

Characteristic

Project EMBED

1 2 3 4 5 SoSE

Artefacts models X X X X X

Framework design X X X X X X

Emergent behavior

awareness

X X

Visual project plan X

Concern with In-

tangible aspects

cern with intangible aspects and bring emergent be-

havior’s awareness through visual project plan and

models. The EMBED-SoSE must use frameworks

and models to generate many of the system artifacts

necessary for a CPSoS, and in this way, combining

the projects can offer great resources for system engi-

neers.

4 EMBED-SoSE: DRAWING A

CPSoS

In general, the traditional design of a system is de-

veloped and documented linearly; that is, a group of

artifacts is generated during the project life cycle to

support the solution design. However, the relation-

ship between these artifacts is not immediately evi-

dent from a cognitive point of view. When seeking

to understand the solution, an engineer must interpret

and analyze the generated artifacts, and in this pro-

cess, the answer to doubts may be dozens of pages

ahead. Although the project plan and its artifacts

are developed linearly, the relationship between its

components is multiple, parallel, simultaneous, and

branching. In this way, to understand the solution as a

whole, it is necessary to create a global view with all

the artifacts, which can be pretty complex in systems

such as CPSoS.

Speciﬁcally, in CPSoS, this complexity is multi-

plied by the number of constituent systems and the

possibilities of emergent behaviors. In this way, the

proposal to facilitate the cognition of complex sys-

tems is to represent them through shared visual mod-

els based on visual thinking. The proposal is to have

a model of the system created collectively and, in this

way, prevent each engineer or stakeholder from mate-

rializing their version of the system. In addition, the

proposal uses the strategy of inﬂuencing engineers in

a design concerned with emergent behavior; that is,

the method leaves this concern latent throughout the

design process, activating the unconscious creative

process. The principle is the same used in sublimi-

nal marketing messages. A subliminal message is any

ENASE 2022 - 17th International Conference on Evaluation of Novel Approaches to Software Engineering

488

stimulus or information exposed to a receiver imper-

ceptibly at a conscious level in an attempt to inﬂuence

opinions and decisions.

According to Wujec (2009), the better we under-

stand the functioning of the human brain and how it

creates meaning and meaning, the better we will be

able to communicate and share information; that is,

we will facilitate cognition.

In this way, this work proposes the EMergent

BEhavior-Driven System of Systems Engineering

(EMBED-SoSE), an interactive, visual, and contin-

uous method to facilitate cognition in the process of

development and maintenance of a system of systems,

as represented in the cycle Fig. 3a.

In general, as proposed by the EMBED-SoSE life

cycle, the formation of the System of Systems (SoS)

will be directly inﬂuenced by its constituent systems

(CS) and by the systemic thinking about the emer-

gence behavior. The processes predicted in an itera-

tive and continuous cycle must be executed and sup-

ported by visual tools, passing through the processes:

(a) Discovery: Its objective is to drive the discov-

ery and evaluation of the solution. This process is

one of the main ones responsible for activating the

creative process of engineers and stakeholders in the

search for the solution; (b) Requirements: Its ob-

jective is to document and facilitate the requirements

elicitation and assessment of a SoSs; (c) Design:

It aims to build the solution design based on the

previous processes; (d) V&V: Includes the veriﬁca-

tion and validation process of the proposed solution;

(e) Development: Solution development proposed in

the previous processes; (f) Integration: Solution in-

tegration to pre-existing SoSs; (g) Operation: In this

process, monitoring, evaluation, and maintenance ac-

tivities of the SoS solution must be performed.

In general, all processes must be based on the

EMBED-SoSE visual thinking paradigm, which de-

ﬁnes that the entire engineering process must, ﬁrst of

all, evaluate the SoS operation concept after keeping

mind in a Think Emergent Behavior, Draw Solutions,

Look and Share, as described Fig. 3b.

This work proposes the EMBED-SoSE Discov-

ery Canvas to guide the discovery process through a

visual tool where all engineers and stakeholders can

collaborate to construct a SoSs or CPSoS. EMBED-

SoSE Discovery Canvas is presented by Fig. 4a,

and in addition to considering the premises of vi-

sual thinking, it takes into account the characteris-

tics and challenges of a CPSoS, as described in the 2

and 2.1. In this way, the Discovery process represents

one of the most critical stages of the EMBED-SoSE

life cycle, which, in addition to supporting all pro-

cesses, is based on the EMBED-SoSE visual thinking

paradigm. The ﬁrst canvas model used within compa-

nies was created by Swiss Alexandre Osterwalder, the

Business Model Canvas (Osterwalder and Pigneur,

2010). The model has become famous throughout the

corporate world and has several applications.

The EMBED-SoSE Discovery Canvas was struc-

tured to lead and activate the unconscious creative

process of engineers and stakeholders, facilitating

cognition and inducing the search for emergent be-

haviors. Based on the challenges presented in the

section 2.1, and focused on transforming unexpected

detrimental emergent behavior into known and ex-

pected cases, the process proposed by the EMBED-

SoSE Discovery Canvas must activate and facilitate

the neurological mechanisms by which our brain cre-

ates its meanings and seeks to solve problems (Kopetz

et al., 2016).

Inspired on Roam (2009), it is possible to visu-

ally clarify any problem through a visual classiﬁca-

tion of six basic questions (who, what, how much,

where, when, how and why?).

EMBED-SoSE Discovery Canvas was structured

based on Roam’s (2009) observations on problems

structuring, in “WHERE”, “WHAT” and “HOW” to

facilitate cognition, in the same way to make the con-

sumption of information neurologically as natural as

possible, as shown in Fig. 4b.

In summary, the EMBED-SoSE Discovery Can-

vas is a large screen where information about the sys-

tem architecture will be positioned to create an inte-

grated and shared view of all possible information of

interest. The screen has been divided into “WHY?”,

“WHERE?”, “HOW?” and “WHAT?”.

The “WHY?” section deﬁnes the motivation for

which the CPSoS should exist; that is, important in-

formation that should support the existence and evo-

lution of the CPSoS is structured. This section is di-

vided into: Justiﬁcation(Rationale), Objective, Bene-

ﬁts, Strategic Needs, Information Needs, and general

rules of the CPSoS.

In the “WHERE?” section, the environments

where the CSs are located are deﬁned. Understanding

the environment where the constituent systems (CS)

are located helps anticipate future needs and design

appropriate mechanisms (e.g., architectural) that al-

low adapting the CPSoS adequately to changes in the

environment.

The “HOW?” section deﬁnes how the CPSoS for-

mation process will be, that is, what are the con-

stituent systems (CS), the interfaces, and their infor-

mation, as well as how this information will ﬂow.

The “WHAT?” section deﬁnes what supports the

formation of the CPSoS. CSs Goals are the common

and conﬂicting goals between the CS; and CSs Ca-

EMBED-SoSE: Drawing a Cyber-physical System of Systems

489

(a) Life Cycle. (b) Visual Thinking Paradigm.

Figure 3: EMBED-SoSE.

(a) Canvas. (b) Explain.

Figure 4: EMBED-SoSE Discovery Canvas.

pabilities and restrictions of each participant system.

This section must also capture the main emerging be-

haviors present in the CPSoS.

In the CSs Goals section, stakeholder needs

should be captured as goals. Goals can be captured

at various levels of abstraction, from high-level strate-

gic objectives to low-level technical objectives. High-

level goals are reﬁned and modeled as AND-OR goal

trees. Goals can represent functional and quality ex-

pectations. This approach can also be valuable in

identifying conﬂicting goals between constituent sys-

tems (CS) and between the CPSoS itself and the con-

stituent systems (CS). It is essential to identify and

understand these conﬂicts because individual systems

often prioritize achieving their own goals.

In the CS Capabilities section, the goals of each

provided capability must be mapped, in addition to

the capabilities and constraints themselves. The map-

ENASE 2022 - 17th International Conference on Evaluation of Novel Approaches to Software Engineering

490

ping of capacity goals can represent advantages and

help in the CPSoS deﬁnition process. Goals provide

a good starting point for identifying the capabilities

needed at the system level and relating the problem

domain to the solution domain. Capabilities are re-

sponsible for providing resources for interaction be-

tween CSs.

In the life cycle proposed by the EMBED-SoSE

Life Cycle, Discovery is the ﬁrst and most impor-

tant of the processes to understand the formation of

a CPSoS and directing the development of collective

knowledge to the requirements gathering phase that

must follow the same philosophy as Discovery.

5 APPLICATION EXAMPLE

This section presents an example for improving the

national natural disaster alert system, a signiﬁcant

CPSoS maintained by Brazil’s Federal Government.

The National Center for Monitoring and Early

Warning of Natural Disasters (Cemaden) aims to

monitor and issue alerts of natural threats in mapped

risk areas of Brazilian municipalities. Besides that,

Cemaden also conducts research and technological

innovations to improve its early warning systems.

(Brazil, 2011). Cemaden operates 24 hours a day,

without interruption, monitoring 958 municipalities

classiﬁed as vulnerable to natural disasters (mainly

ﬂoods and landslides).

In general, the national ﬂood risk monitoring sys-

tems are unconnected to the city siren and contain-

ment reservoirs’ ﬂoodgates control systems. Thus,

Fig. 5 presents the EMBED-CPSoS Discovery Can-

vas application for the design of the Flood Prevention

System that connects the systems.

Considering the above, the construction of the

EMBED-SoSE Discovery Canvas requires nothing

more than a few materials available in the ofﬁce:

those adhesive blocks known as post-it notes and ﬂip

chart sheets (Format A1). The sheet in A1 format,

segmented according to Figure 4a will be used as can-

vas. The sheet should be large enough for stakehold-

ers to meet around it and collaborate in the construc-

tion process. In practice, the Canvas is built as the

concepts represented by the post-its are placed on the

sheet orderly and create the necessary relationships

between these concepts. Post-its offer an explicit re-

striction of writing space, which leads to more objec-

tive writing.

The ﬁrst step in the example in Fig. 5 is to iden-

tify and ﬁll in the “WHY?” section. Start the pro-

cess by identifying the problem to be solved. De-

scribe Justiﬁcation/rationale, objectives, and beneﬁts.

Also, in the “WHY?” section, describe the needs and

rules/restrictions. The second step is identifying the

constituent systems (CS) candidates to contribute to

the solution in the “HOW?” section. From the iden-

tiﬁcation of the system, also are identify the environ-

ments where they are associated in the “WHERE?”

section. Still in the “HOW?” section, the next step is

to identify which integration interfaces are available

for each identiﬁed CS, besides how information can

ﬂow between systems and their interfaces. Finally,

the “WHAT?” deﬁnes what supports the formation of

the CPSoS.

The discovery process is iterative, where the vi-

sual creation process allows activating the creativity

in the group. In this way, the discovery canvas for

the ﬂood prevention system was built collectively and

visually. Allowing Cemaden engineers to develop the

design of the solution in a shared way, where the main

difﬁculties related to the integration interfaces could

be discussed and evaluated more efﬁciently than the

traditional engineering model.

6 CONCLUSIONS

EMBED-SoSE empowers systems engineering (SE)

with the ability to generate a maturity proﬁle for criti-

cal and complex systems like CPSoS. In addition, the

new paradigm based on visual thinking proposed by

this work can generate a radical change in the systems

engineering process, allowing the collective construc-

tion of a more straightforward and more effective de-

sign for the integration of heterogeneous systems such

as CPSoS without sacriﬁcing rigor.

The paradigm based on the visual thinking of

EMBED-SoSE, when concerned with emerging be-

haviors that deliver value to the business, will con-

sequently be concerned with the main intangible as-

pects of the system and how to materialize them col-

lectively, which allows the creation of shared knowl-

edge, reducing the cognitive effort to understand the

project among the Stakeholders.

The ability to simplify the engineers’ cognitive

process allows for a much more straightforward as-

sessment of system-level behavior and, depending on

accuracy, can simulate emerging CPSoS behaviors

early in the project lifecycle. In addition, unwanted

behaviors and failures are highly likely to be identi-

ﬁed earlier in the project.

According to Fig. 2 (Kopetz et al., 2016), emer-

gent behavior can also be unexpected, and the effort

becomes to make it known and expected. In this way,

the EMBED-SoSE Discovery Canvas process must be

interactive, incremental, and collaborative; that is, it

EMBED-SoSE: Drawing a Cyber-physical System of Systems

491

Figure 5: Flood Prevention System Discovery Canvas.

must be revisited to discover new emerging behaviors.

It is essential to clarify that the EMBED-SoSE

Discovery Canvas is not a ﬂowchart of the project

since the ﬂowchart represents a sequence of steps,

while what is essential in the Discovery Canvas are

the relationships between concepts. In some cases,

the Canvas itself can only represent a single and con-

sistent project plan, however in other projects, the

rigor of formalism may be required, so the Canvas

can be used as a primary tool to confront the logic of

the project and relate the concepts, serving as a basis

for the transcription of a project plan in the traditional

model required by formalism.

There is no doubt that knowledge about CPSoS

and its unexpected emergent behaviors can be lim-

ited; however, with a well-structured process, the

EMBED-SoSE Discovery Canvas proposes a way to

increase the ability to evidence these types of be-

haviors through stimulation and engineers’ creative

ability. In summary, EMBED-SoSE proposes an ap-

proach to systems engineering (SE), especially to

complex systems such as a CPSoS, which moves

away from the linearity of classical design and makes

the connections between the parts evident, reducing

the cognitive effort of engineers and stakeholders.

The next step for this work is to develop a case

study for proposal validation. In addition, the work

can generate opportunities for new research and the

development of tools based on the paradigm of visual

thinking.

REFERENCES

Bojanova, I., Hurlburt, G., and Voas, J. (2014). Imagineer-

ing an internet of anything. Computer, 47:72–77.

Bondavalli, A., Bouchenak, S., and Kopetz, H. (2016).

Cyber-physical systems of systems: foundations–

a conceptual model and some derivations: the

AMADEOS legacy, volume 10099. Springer.

Brazil (2011). Decreto nº 7.513 de 1º de julho de 2011.

Carreira, P., Amaral, V., and Vangheluwe, H. (2020).

Foundations of Multi-Paradigm Modelling for Cyber-

Physical Systems. Springer.

Ceccarelli, A., Bondavalli, A., Froemel, B., H

oftberger, O.,

and Kopetz, H. (2016). Basic Concepts on Systems of

Systems, page 1. Springer, Cham.

Holt, J., Perry, S., Hansen, F. O., Miyazawa, A., Kristensen,

K., and Hains, R. (2014). Final report on guidelines

for sos engineering. Technical report, COMPASS.

ISO (2019). ISO/IEC/IEEE 21839. International Organiza-

tion for Standardization, Genebra.

Kopetz, H. (2011). Real-time systems: design principles for

distributed embedded applications. Springer, 2 edi-

tion.

Kopetz, H., Bondavalli, A., Brancati, F., Fr

omel, B.,

oftberger, O., and Iacob, S. (2016). Emergence in

Cyber-Physical Systems-of-Systems (CPSoSs), pages

73–96. Springer, Cham.

Lee, E. A. (2008). Cyber physical systems: design

challenges. In Proceedings..., pages 363–369. In-

ternational Symposium on Object and Component-

Oriented Real-Time Distributed Computing (ISORC),

IEEE.

ENASE 2022 - 17th International Conference on Evaluation of Novel Approaches to Software Engineering

492

Maier, M. W. (1998). Architecting principles for systems-

of-systems. Systems Engineering, 1(4):267 – 284.

Mangeruca, L., Passerone, R., Etzien, C., Gezgin, T.,

Peikenkamp, T., Jung, M., Alexandre, A., Bullinga,

R., Imad, S., Honour, E., Paul, S., and Klaas, S.

(2013). Danse methodology v2. Technical report,

ALES.

Osterwalder, A. and Pigneur, Y. (2010). Business Model

Generation: A Handbook for Visionaries, Game

Changers, and Challengers. The Strategyzer Series.

Wiley.

ols, M. A., Mashkoor, A., and Egyed, A. (2021). Team-

oriented consistency checking of heterogeneous engi-

neering artifacts. In 2021 IEEE/ACM 43rd Interna-

tional Conference on Software Engineering: Compan-

ion Proceedings (ICSE-Companion), pages 250–251.

EMBED-SoSE: Drawing a Cyber-physical System of Systems

493