Using Modelling and Simulation as a Service (MSaaS) for Facilitating

Flexibility-based Optimal Operation of Distribution Grids

Moritz Stüber, Lukas Exel and Georg Frey

Chair of Automation and Energy Systems, Saarland University, Campus A5 1, 66123, Saarbrücken, Germany

Keywords:

Modelling and Simulation as a Service (MSaaS), Simulation as a Service (SIMaaS), Service-oriented

Architecture (SOA), Flexibility, Power-to-Heat.

Abstract:

In this contribution, initial results of a research project on modelling and simulation as a service (MSaaS)

within the context of optimal operation of distribution grids by exploiting ﬂexibility are summarized. Based

on a brief description of service-oriented architecture (SOA), deﬁnitions and key aspects of MSaaS as well as

open conceptual and technical challenges are outlined. Some speciﬁc aspects and challenges are then related

to a possible application: a service that predicts the ﬂexibility that a power-to-heat system can offer is imple-

mented as an instance of MSaaS. The calculations are based on the current state of the system and a model

deﬁned in the equation-based, object-oriented language Modelica, as well as historical data when available.

Thereby, an accurate and reliable representation of the ﬂexiblity becomes available that facilitates its use for

system balancing, energy market participation and grid operations. This functionality is exposed through a

representational state transfer (REST)-based service interface, designed to allow for the straightforward inte-

gration with existing systems and other virtual resources. The service architecture and initial results of the

implementation are described.

1 INTRODUCTION

The term “modelling and simulation as a service

(MSaaS)” describes the effort to provide users with

modelling and simulation (M&S) functionality over

the internet by using a service-oriented architecture

(SOA) to structure applications and cloud computing

technology to efﬁciently host them (Cayirci, 2013b).

Before introducing MSaaS in more detail by elabora-

ting proposed beneﬁts and key aspects in section 2, a

brief introduction to M&S as well as an introduction

to SOA will be given below; an introduction to cloud

computing can be found in Cayirci (2013b, section 2)

and Mckee et al. (2017a, section 3).

From a systems engineering perspective, M&S

can be seen as a way to calculate physical quanti-

ties in technical devices that is particularly suited for

analysing complex, multi-domain systems. In the

context of modelling and simulation using equation-

based, object-oriented languages (EOOLs) such as

Modelica

, a model generally consists of a system

of implicit differential algebraic equations (DAEs)

which can be numerically approximated by a suitable

https://modelica.org/

solver for a given set of parameters and boundary con-

ditions. Based on the resulting trajectory, a user can

predict the behaviour of the real or imagined system

that the model represents. Compared to experiments

with prototypes, M&S offers faster and safer experi-

mentation as well as access to internal states that can-

not be measured in a real experiment (Cellier, 2013).

Due to the acausal formulation of the models, EOOLs

allow for fast assembly of complex system models by

reusing available component models.

The ﬁrst step in creating a model is ﬁnding an

abstraction of a system that is suitable for the pur-

pose of the model. In-depth domain knowledge is

required in order to understand and describe the re-

levant properties of the system at hand. Second, the

model needs to be implemented in a formal model-

ling language, which requires expertise in the model-

ling formalisms, languages and algorithms used be-

cause attention to the numerical properties of the mo-

del and the use of structuring mechanisms that ensure

the reusability of the model must be paid. M&S en-

vironments such as Dymola

provide support regar-

ding the implementation of the models by checking

https://www.3ds.com/products-services/catia/

products/dymola/

Stüber, M., Exel, L. and Frey, G.

Using Modelling and Simulation as a Service (MSaaS) for Facilitating Flexibility-based Optimal Operation of Distribution Grids.

DOI: 10.5220/0006899106230630

In Proceedings of the 15th International Conference on Informatics in Control, Automation and Robotics (ICINCO 2018) - Volume 2, pages 613-620

ISBN: 978-989-758-321-6

613

for syntactic mistakes, providing access to model li-

braries, offering the possibility to graphically create

models by assembling component models, translating

the models to executable form, triggering their execu-

tion and allowing users to analyse the results. Howe-

ver, it remains the user’s responsibility to make sense

of the calculated trajectories and to ensure that they

are valid. In practice, obtaining accurate parameter

values and measurement data for veriﬁcation and va-

lidation of developed models represents a major chal-

lenge.

Applications of M&S include dimensioning of

technical devices, ﬁnding control strategies and as-

sessing the quality and feasibility of ideas as well

as control and training applications (hardware-in-the-

loop, human-in-the-loop). For example, M&S could

be used to predict the energy generated by a photo-

voltaic system or to compare different possibilities for

increasing the own use of energy within a household

from a technical and economical perspective. In this

contribution, it is used to determine the ﬂexibility of a

power-to-heat (PtH) system, which is then used as an

input to a larger process.

SOA is “a paradigm for organizing and utilizing

distributed capabilities that may be under the control

of different ownership domains” (OASIS, 2006). The

ISO/IEC standard 18384 (ISO/IEC_18384-1:2016, E)

deﬁnes SOA as “an architectural style that supports

service orientation”, where service orientation is de-

ﬁned as an “approach to designing systems in terms of

services and service-based deployment” and a service

is deﬁned as a “logical representation of a set of acti-

vities that has speciﬁed outcomes, is self-contained,

may be composed of other services, and is a ‘black-

box’ to consumers of the service”. In other words, a

service describes the capability, the speciﬁcation and

an offer to perform work for someone and can also

be seen as a mechanism to match the capability of

the service provider with the need of the service con-

sumer. SOA is a way of structuring and offering a

functionality that promotes reuse, growth and intero-

perability by focusing on tasks and business functi-

ons and acknowledging the existence of ownership

boundaries. The OASIS Reference Model for Ser-

vice Oriented Architecture (OASIS, 2006) identiﬁes

six major concepts of SOA and describes their pro-

perties and relations: visibility, service description,

interaction, contracts & policies, real world effect and

execution context.

Provided that a service is visible to the consumer,

interaction with the service results in a retrieval of in-

formation and/or in a change of the shared informa-

tion about affected entities (real world effect). The-

reby, it contributes to the goal of the service consu-

mer, which is generally unknown to the service pro-

vider. The interaction with a service is deﬁned by an

information model and a behaviour model, which are

exposed through the service interface. The service in-

terface is formally described in the service descrip-

tion, which includes applicable contracts and policies.

The infrastructure elements, process entities and po-

licy agreements of a speciﬁc service interaction are

called execution context. The execution context may

evolve during an interaction, for example if the ser-

vice participants agree to encrypt the following com-

munication. Visibility, interaction and real world ef-

fect can be seen as the dynamic elements of a service

interaction as they result from the actions of the ser-

vice participants. In contrast, service description, po-

licies and execution context can be seen as supporting

aspects that are mostly deﬁned by the service provi-

der.

According to OASIS (2006, section 2.3), SOA is

well suited for developing complex, yet manageable

software systems that scale well and allow remote

users to choose the services that they need. However,

implementing a SOA in a robust way based on stan-

dards and best practices, fully exploiting the under-

lying concepts, can be challenging(Rodríguez et al.,

2016) . Furthermore, systematic approaches for cre-

ating value by transitioning to a service-oriented ar-

chitecture as well as concepts and algorithms for the

automated composition of services are missing.

2 M&S as a SERVICE

By combining the concepts and tools of SOA and

cloud computing with the functionality of modelling

and simulation, it is hoped to help a wide audience

make informed decisions when confronted with com-

plex questions: the various methods, languages and

tools summarized under the term modelling and simu-

lation allow ﬁnding quantitative information about the

behaviour of a system by numerically approximating

the behaviour of a virtual representation of that sy-

stem; SOA and cloud computing allow users to access

this capability over the internet.

Currently, factors that prevent a more widespread

use of modelling and simulation amongst engineers

include a lack of knowledge regarding concepts, mo-

delling languages, algorithms and tools, the cost of

tools and model libraries as well as problems with the

usability of user interfaces and data formats. Also,

current M&S environments are not designed to be

used as part of a larger process. Factors preventing the

general public from using modelling and simulation

include missing awareness of its capabilities, missing

ICINCO 2018 - 15th International Conference on Informatics in Control, Automation and Robotics

614

background knowledge, missing feeling for the steps

required to arrive at a solution and the challenge of

using complex M&S software environments.

MSaaS attempts to address the aforementioned

problems by introducing a new layer of abstraction:

instead of dealing with methods and data in the form

of models, parameters and unprocessed simulation re-

sults, a service consumer interacts with an interface

designed for speciﬁc tasks/business functions. Thus,

the needed functionality is separated from the techni-

cal prowess and infrastructure required to implement

it. The service interaction results in the retrieval of

information as deﬁned by Rowley (2007): “[. . .] in-

formation is deﬁned in terms of data, and is seen to be

organized or structured data. This processing lends

the data relevance for a speciﬁc purpose or context,

and thereby makes it meaningful, valuable, useful and

relevant”.

MSaaS is deﬁned as a “means of delivering va-

lue to customers to enable or support modelling and

simulation (M&S) user applications and capabilities

as well as to provide associated data on demand wit-

hout the ownership of speciﬁc costs and risks” by

the specialist team MSG-131 at NATO (MSG-131,

2015)v. Cayirci (2013b) sees MSaaS as a “model for

provisioning modelling and simulation (M&S) servi-

ces on demand from a cloud service provider (CSP),

which keeps the underlying infrastructure, platform

and software requirements/details hidden from the

users”, which aligns well with the previous deﬁnition

but highlights the use of cloud computing technolo-

gies. This aspect is important for distinguishing the

current research activities from earlier attempts, sum-

marized under the term web-based simulation (WBS),

as retraced by Wang and Wainer (2016, section 2).

Consequently, MSaaS is one form of software as a

service (SaaS) (Cayirci, 2013b; Mckee et al., 2017a)

that “inherits” the general beneﬁts and challenges of

cloud computing, but also offers opportunities pecu-

liar to the ﬁeld of M&S.

The proposed value of MSaaS (Cayirci, 2013b;

MSG-131, 2015, section 2.6) lies in an increase in

usability and functionality on the one hand and a pos-

sible decrease in costs on the other hand. Increased

usability and functionality are expected to stem from

better accessibility of M&S resources, their reuse for

reproducing results and the creation of new functio-

nality through composition of resources:

• The accessibility of M&S resources is increa-

sed through broad network access that allows on-

demand self-service while having very light re-

quirements on the client (hardware, operating sy-

stem (OS), software).

• Accessible resources that hide their complexity

and implementation from the user (encapsulation)

make the reuse of knowledge possible, for exam-

ple the provision of product models by companies.

• Also, providing an interface to an operational si-

mulation environment can enhance the credibility

of a simulation study by ensuring its replayability

and makes sure that a user can always access the

latest version of the software.

• Implementing new functionality by composition

of services is facilitated by provisioning M&S re-

sources as a service. Composition can reduce the

time and effort needed for implementation and ex-

tend the scope of M&S resources by using them in

conjunction with functionality that is not normally

available in an M&S environment.

A decrease in costs can be achieved by measu-

ring the service usage, which allows pay-per-use op-

tions, by scaling the infrastructure according to de-

mand and by increasing the degree of capacity uti-

lization through resource pooling. Furthermore, the

service consumer does not need to invest in hardware,

software and personnel, which could increase the wil-

lingness of companies to try to integrate M&S into

their workﬂow.

Researchers have worked on combining M&S

with web technologies for more than 20 years by

imagineering the possibilities (Fishwick, 1996; Mc-

kee et al., 2017a), proposing architectures for im-

plementation (Cubert and Fishwick, 1997; Al-Zoubi

and Wainer, 2011; Ribault and Wainer, 2012; Shekhar

et al., 2016), pondering risk, trust and accountability

(Cayirci, 2013a) and investigating the composability

of M&S services (Tolk, 2013; Tolk and Mittal, 2014).

Below, we will elaborate what we perceive as the key

aspects of MSaaS.

Service Interface. The service interface deﬁnes the

possibilities for interaction with a service. It com-

bines the underlying functionality with the ser-

vice’s execution context. The possibilities and li-

mitations of the latter deﬁne the degree to which

the ideas and proposed beneﬁts of SOA and cloud

computing applied to M&S can be realized. This

applies to both the general characteristics of cloud

computing and internet technologies as well as the

speciﬁc characteristics of the chosen service ar-

chitecture.

Service Architecture. The service architecture

structures an implementation and comprises

the underlying design concepts, for example

representational state transfer (REST), the use

of standards such as OpenAPI or functional

mockup interface (FMI), the software tools and

Using Modelling and Simulation as a Service (MSaaS) for Facilitating Flexibility-based Optimal Operation of Distribution Grids

615

middleware, the deployment technology as well

as the available computing power.

Service Composition The full potential of transitio-

ning to using SOA for providing M&S capabilities

cannot be realized when limiting the implementa-

tions to providing existing M&S functionality in

its current form over the internet as a web applica-

tion. An increase in value stems from new use ca-

ses and target audiences for M&S technology and

new functionality for users which can be achie-

ved by service composition. Examples are pre-

sented by Wang and Wainer (2016); Wainer and

Wang (2017); Mckee et al. (2017b). However,

service composition represents a conceptual and

technical challenge because a consistent and via-

ble semantics needs to be guaranteed across ser-

vices in order to achieve user acceptance of the

proposed solutions. Tolk and Mittal (2014), the-

refore, propose that a consistent representation of

truth can be seen as the deﬁnition of composabi-

lity and state that it can be achieved by alignment

of data and the orchestration of the individual ser-

vices.

Contracts and Policies Providing functionality for a

remote user makes an explicit and comprehensive

consideration of the implications on security, pri-

vacy, intellectual property and accountability ne-

cessary (Cayirci, 2013b, section 4, 5). This me-

ans that an analysis of risk and trust needs to be

performed and adequate countermeasures need to

be implemented. Ideally, the countermeasures are

integrated into the service architecture by design.

From the perspective of a SOA, contracts and po-

licies are the mechanisms for dealing with the afo-

rementioned. In general, contracts are negotiated

between the service participants whereas policies

are enforced by the service provider. They go-

vern the willingness of a service to perform a re-

quest and can be related to infrastructure-oriented

as well as business-oriented matters. Two exam-

ples that represent the need for business-oriented

policies and highlight the importance of conside-

ring the deﬁnition of policies and contracts tho-

roughly are privacy and intellectual property.

Signiﬁcant conceptual and technological challenges

have to be addressed before the full potential of MS-

aaS can be understood and, ultimately, realized. Con-

sequently, cloud-based M&S or MSaaS was identiﬁed

as one of the “grand challenges for modelling and si-

mulation” by Taylor et al. (2015, section 4).

Conceptual challenges include questions such as

“what makes sense?”, “which granularity of services

makes sense?”, “how can SOA and cloud computing

be exploited to full capacity?”, “how can trust re-

garding the reliability of results be induced?”, “how

can trust regarding the use of sensitive data be indu-

ced?” and “how can a consistent representation of

truth be guaranteed across services?” as well as “how

to derive viable business models?”. On the technical

side, it is necessary to combine the expertise, tools

and methods from M&S, SOA and cloud computing

in order to ﬁnd a software architecture that facilitates

the sustainable development of M&S services. Addi-

tionally, solutions for formulating and enforcing con-

tracts and policies and ensuring the security of the ser-

vice as well as its reachability need to be found.

In the next section, an envisioned application of

MSaaS for facilitating the ﬂexibility-based optimal

operation of distribution grids is described.

3 DETERMINATION OF THE

FLEXIBILITY OF A

POWER-TO-HEAT SYSTEM

The increasing share of energy generated by solar and

wind power plants imposes challenges on the electri-

cal power grid. Because of the dependency of these

power plants on the weather, utilities have to ﬁnd

ways to guarantee the stability of the grid despite the

volatile nature of the energy generation. One possi-

ble way of stabilizing the grid is to use loads ﬂexi-

bly and consume power when it is generated, in ot-

her words using ﬂexibility for optimal operation of

the grid. Flexibility is deﬁned as “the changes in con-

sumption/injection of electrical power from/to the po-

wer system from their current/normal patterns in re-

sponse to certain signals, either voluntarily or manda-

tory” (SG-CG/M490/L, 2014, section 5.1).

3.1 Scenario

As an exemplary application, the energy consump-

tion of a building complex is investigated with respect

to possiblities to adapt the consumption of electrical

energy on demand. In the system under investiga-

tion, two such possibilities exist: ﬁrst, the warm water

can be generated by using submersion heaters instead

of the normal gas heating system, which represents a

load that can be switched on and off ﬂexibly. The sub-

mersion heaters are mounted inside the storage tank.

Second, there is a solar photovoltaic system which

can be throttled on demand. Consequently, the overall

system represents a buffered ﬂexibility source which

is controllable within bounds.

In order to reliably determine how much energy

ICINCO 2018 - 15th International Conference on Informatics in Control, Automation and Robotics

616

can be consumed by the PtH system and/or not gene-

rated by the solar panels for a given time frame, mo-

delling and simulation of the overall system is neces-

sary due to its complexity: among other factors, the

ﬂexibility that can be offered by the PtH-components

depend on the current state of the system, the maxi-

mum allowed temperature of the water, the tempera-

ture of the fresh water, the warm water consumption,

and the heat losses to the environment; the ﬂexibi-

lity of the photovoltaic system mainly depends on the

weather conditions, but for example also on the im-

plemented operating strategies et cetera.

As a ﬁrst step, a very simple model of the buil-

ding was implemented using the modeling language

Modelica. It mainly consists of a model for warm wa-

ter generation and -storage and the control logic that

ensures that whenever the PtH-components are active,

the water is not heated by the gas heating system. The

level of the thermal energy storage, deﬁned as the dif-

ference of the current temperature of the water inside

the tank to the minimum temperature divided by the

allowed temperature range, is introduced as a charac-

teristic value. Parameters of the used models were set

to plausible values, but there is no direct relation to

a real building; currently, the purpose of the model is

to provide plausible data as part of a proof-of-concept

implementation. The electric power generated by the

solar power plant as well as warm water consumption,

heating demand and consumption of electrical power

are deﬁned as inputs to the system model.

The ﬂexibility arising from throttling the solar

power plant is calculated by subtracting the electric

energy consumption inside the building from the ge-

nerated power. The ﬂexibility provided by the PtH-

components is calculated as follows: the model of

the building is used as the plant model within a

closed control loop, using the level of the thermal

energy storage as the measured process variable and

100 % as the value of the desired set-point. The PtH-

components are activated and the power they are re-

quired to provide to the system in order to always

keep the level of the thermal energy storage at the

desired set-point is deﬁned as the ﬂexibility they can

offer.

3.2 Service Architecture

Providing an answer to the question “how much

energy can be consumed/not generated on demand by

the system in the next 24 h?” represents the capability

that shall be offered as a service in order to enable

the service consumer to offer the calculated ﬂexibi-

lity to potential users. The service shall also integrate

necessary auxiliary steps, such as querying forecasts

for weather and user behaviour, thus hiding the under-

lying complexity of the calculation from the service

consumer.

Consequently, the service is realized as an in-

stance of service composition from a technical point

of view: while the endpoints of the service that pro-

vides the calculation of the ﬂexibility for a certain sy-

stem are speciﬁc to the needs of the service consumer,

different services are called in the background and the

results are combined such that the desired functiona-

lity follows, as shown in Figure 1.

First, the necessary input for the simulation of the

building are collected, namely the forecast for the

electric power generated by the solar power plant,

which in turn requests an up-to-date weather fore-

cast before triggering the simulation of the under-

lying model, as well as forecasts for the behaviour of

the residents (warm water consumption, heating de-

mand, consumption of electric power). Then, the ac-

tual simulation is triggered by sending a request to a

simulation as a service (SIMaaS)-instance.

For each of the developed services, a formal ser-

vice description needs to be found based on the des-

cription of the desired service functionality. The ser-

vice description “represents the information needed in

order to use a service” (OASIS, 2006, section 3.3.1)

and can comprise information on the service’s reacha-

bility, functionality, its interface and applicable poli-

cies. It exposes the set of capabilities offered by the

service provider to a remote audience, but its expres-

sivity and therefore the degree to which these capabi-

lities can be exposed is deﬁned by the service archi-

tecture and the underlying architectural style. Con-

sequently, the development of the service description

and the development of the service architecture go

hand in hand. In the presented case, the service in-

terface adheres to the architectural style REST due to

its advantages over other approaches, following the

argumentation presented by Wang and Wainer (2016,

section 2).

According to Verborgh et al. (2015, section 3.1),

REST is “a style for system architectures, dictating

several architectural constraints, rather than a techno-

logy, or architecture in itself”. Web services adher-

ing to REST are called RESTful web services and

their service interface is also referred to as REST-

API (Rodríguez et al., 2016, section 2). REST ap-

plies to client-server systems, based on the idea that

the resulting separation of concerns facilitates scala-

bility and longevity. The constraints characterizing

REST concern the identiﬁcation of resources, the ma-

nipulation of resources through representations, the

exchange of self-descriptive messages (stateless inte-

ractions) and the use of hypermedia as the engine of

Using Modelling and Simulation as a Service (MSaaS) for Facilitating Flexibility-based Optimal Operation of Distribution Grids

617

Figure 1: The functionality of the service that provides the desired information is realized by combining several services.

application state (HATEOAS); they are intended to

result in a uniform interface (Verborgh et al., 2015,

section 3.3).

In the abstract, a REST-API gives access to a set

of resources. Resources are unique conceptual en-

tities of the service interface; the semantics of a re-

source always remains the same. Conceptual entities

of the application domain are mapped to resources

when deﬁning the service functionality. Resources

can be considered as nouns that are uniquely identi-

ﬁed by URIs. Interaction with resources occurs by

applying a set of verbs deﬁned by the hypertext trans-

fer protocol (HTTP) speciﬁcation to these nouns. Ap-

plying an HTTP-verb to a resource results in a tran-

saction of a resource representation that is deﬁned by

the (hyper)media type agreed upon by the service par-

ticipants. Therefore, the same service can be used to

integrate with a variety of existing systems by imple-

menting the appropriate resource representation.

Deﬁning the resources provided by a given ser-

vice is not sufﬁcient for describing the service inter-

face, which deﬁnes the speciﬁcs of how to interact

with the service in order to invoke its functionality.

As the exchange of messages represents the “primary

mode of interaction with a service” (OASIS, 2006,

section 3.2.2), the structure and semantics of the data

that needs to be exchanged have to be deﬁned as well,

resulting in the service’s information model. The cor-

responding behaviour model deﬁnes the possible in-

teractions with the service as well as their temporal

relationships and properties and also needs to be des-

cribed.

In order to describe these details in a standardized

way that is readable by both humans and machines,

a web service description language should be used.

Because the software under development is descri-

bed in terms of resources, actions and representations

instead of objects, methods and attributes, a web ser-

vice description language promotes the realization of

SOAs. The resulting document, the service descrip-

tion, serves as documentation for humans and can be

used to automatically derive server stubs, thus signi-

ﬁcantly reducing the amount of code that needs to be

implemented manually. Therefore, the interfaces of

all developed services are described according to the

OpenAPI-speciﬁcation

The performance of a service instance is deﬁned

by the characteristics of the implementation and the

characteristics resulting from deploying it on a certain

hardware using a certain technology. Thus, based on

the answer to the question “which performance cha-

racteristics do I envision for the service?”, an answer

to the question “how do I deploy the service in or-

der to fulﬁl them?” needs to be found. Aspects to

be considered include the formulation, measurement

and testing of key performance indicators (KPIs) as

well as the scalability of the chosen solution. Since

MSaaS is described as the combination of M&S, SOA

and cloud computing, using infrastructure as a ser-

vice (IaaS), platform as a service (PaaS) and SaaS as

well as associated technologies like Docker in order

to achieve the characteristics associated to “the cloud”

also need to be considered.

Typically, M&S environments and -tools have ex-

tensive dependencies. In order to avoid problems due

to those dependencies and in order to clearly sepa-

rate the different modules of a service, it was deci-

ded to use Docker as a software container platform.

After explicitly deﬁning the dependencies of a soft-

ware once, Docker guarantees that the software runs

no matter where it is deployed and has a number of

additional advantages, such as its suitability for agile

software development processes.

https://github.com/OAI/OpenAPI-Speciﬁcation

ICINCO 2018 - 15th International Conference on Informatics in Control, Automation and Robotics

618

3.3 Discussion

The predictions delivered by the developed service

are currently not yet applicable in a real practical ap-

plication. However, implementing the functionality

as a service increases both the accessibility of simu-

lation resources and the reuse of knowledge: instead

of using a complex and typically expensive M&S-

environment, which requires manual work and/or

scripting, the service consumer sends a simple HTTP-

request to a web API. This facilitates the reuse of

an analysis, such as the calculation of the ﬂexibility,

within a larger process. Therefore, also the reuse of

knowledge and capabilities is promoted. For exam-

ple, a SIMaaS-instance is used by both the service that

predicts the power generated by a solar power plant as

well as by the service that predicts the ﬂexibility.

The use of service composition for implementing

a certain functionality has two major advantages:

ﬁrst, it allows deﬁning analyses in the problem dom-

ain without having to take care of how the individual

parts required for the analysis are implemented. For

example, obtaining the latest forecast for the solar po-

wer generation reduces to calling an application pro-

gramming interface (API), the complexity of hand-

ling the underlying model and auxiliary data such as

the weather forecast are hidden from the consumer

of this API. Second, using services makes it easy

to change them; for example, all forecast services

could be replaced by services that provide measure-

ment data, thus allowing to investigate the quality of

the forecast services and allowing to perform parame-

ter ﬁtting et cetera.

Many aspects of using MSaaS for providing in-

formation as part of a larger process have not been

considered yet. Most importantly, these include con-

sidering the interoperability of services both from a

conceptual and a technical point of view, and the sca-

lability of the developed solutions. Moreover, threats

to the security of the solutions as well as possible mi-

tigations need to be considered.

With respect to the service functionality, a process

for evolving model instances based on parameter es-

timation shall be devised and implemented in future

work: in the ﬁrst iteration, an initial set of parameters

is chosen. In subsequent iterations, the parameters

are optimized by automatically comparing the simu-

lation results to measurement data, which is linked to

the model instance and accessible from within the ser-

vice.

4 CONCLUSION

MSaaS represents a ﬁeld of study at the intersection

of M&S, information science and software engineer-

ing with the goal of increasing the accessibility of and

audience for the formalisms to describe knowledge

about a system and the algorithms used to perform

experiments based on the resulting models that were

found in M&S research. Concepts developed in in-

formation sciences, such as service-orientation, pro-

vide the necessary theoretical background to trans-

fer existing solutions to the new problem domain of

M&S and to identify and reason about challenges pe-

culiar to this domain, for example the composabi-

lity of M&S-services. Software engineering methods

and tools, in particular cloud computing, the concept

of representational state transfer (REST), API ﬁrst-

development and containerization enable the realiza-

tion of solutions that meet the functional and non-

functional requirements resulting from a speciﬁc ser-

vice concept.

An exemplary application of MSaaS, namely pro-

viding an accurate and reliable description of the ﬂex-

ibility of a PtH system that can be used to mitigate

unwanted states of the distribution grid in subsequent

processes, was outlined.

ACKNOWLEDGEMENTS

This work was supported by the SINTEG-project “De-

signetz” funded by the German Federal Ministry of

Economic Affairs and Energy (BMWi) under grant

03SIN224.

REFERENCES

Al-Zoubi, K. and Wainer, G. (2011). Distributed Simula-

tion Using RESTful Interoperability Simulation Envi-

ronment (RISE) Middleware, pages 129–157. Sprin-

ger Berlin Heidelberg, Berlin, Heidelberg.

Cayirci, E. (2013a). A joint trust and risk model for msaas

mashups. In Pasupathy, R., Kim, S.-H., Tolk, A., and

Kuhl, M. E., editors, Proceedings of the 2013 Winter

Simulation Conference (WSC).

Cayirci, E. (2013b). Modeling and simulation as a cloud

service: A survey. In Pasupathy, R., Kim, S.-H., Tolk,

A., and Kuhl, M. E., editors, Proceedings of the 2013

Winter Simulation Conference (WSC).

Cellier, F. E. (2013). The complexity crisis: Using modeling

and simulation for system level analysis and design. In

3rd International Conference on Simulation and Mo-

deling Methodologies, Technologies and Applications,

Reykjavik, Island.

Using Modelling and Simulation as a Service (MSaaS) for Facilitating Flexibility-based Optimal Operation of Distribution Grids

619

Cubert, R. M. and Fishwick, P. A. (1997). A framework for

distributed object-oriented multimodeling and simula-

tion. In Proceedings of the 29th Conference on Winter

Simulation, WSC ’97, pages 1315–1322, Washington,

DC, USA. IEEE Computer Society.

Fishwick, P. A. (1996). Web-based simulation: Some per-

sonal observations. In Proceedings of the 28th Con-

ference on Winter Simulation, WSC ’96, pages 772–

779, Washington, DC, USA. IEEE Computer Society.

ISO/IEC_18384-1:2016(E) (2016). Information technology

– reference architecture for service oriented architec-

ture (soa ra). ISO/IEC 18384-1:2016(E), ISO/IEC.

Mckee, D., Clement, S., Ouyang, X., Xu, J., Romano, R.,

and Davies, J. (2017a). The internet of simulation, a

specialisation of the internet of things with simulation

and workﬂow as a service (sim/wfaas). In 11th IEEE

International Symposium on Service-Oriented System

Engineering (SOSE 2017). IEEE.

Mckee, D., Clement, S., Xu, J., Romano, R., Lopez, J.,

and Battersby, D. (2017b). The internet of simulation:

Enabling agile model based systems engineering for

cyber-physical systems. In IEEE 12th System of Sys-

tems Engineering Conference (SoSE 2017), Waikoloa,

Hawaii, USA. IEEE.

MSG-131, S. T. (2015). Modelling and simulation as a ser-

vice: New concepts and service-oriented architectu-

res. Final Report AC/323(MSG-131)TP/608, North

Atlantic Treaty Organization NATO.

OASIS (2006). Reference model for service oriented archi-

tecture 1.0. Technical report.

Ribault, J. and Wainer, G. (2012). Using workﬂows and

web services to manage simulation studies (wip). In

Proceedings of the 2012 Symposium on Theory of

Modeling and Simulation - DEVS Integrative M&S

Symposium, TMS/DEVS ’12, pages 50:1–50:6, San

Diego, CA, USA. Society for Computer Simulation

International.

Rodríguez, C., Baez, M., Daniel, F., Casati, F., Trabucco,

J. C., Canali, L., and Percannella, G. (2016). Rest

apis: A large-scale analysis of compliance with princi-

ples and best practices. In Bozzon, A., Cudre-Maroux,

P., and Pautasso, C., editors, Web Engineering: 16th

International Conference, ICWE 2016, Lugano, Swit-

zerland, June 6-9, 2016. Proceedings, pages 21–39,

Cham. Springer International Publishing.

Rowley, J. (2007). The wisdom hierarchy: representations

of the dikw hierarchy. Journal of Information Science,

33(2):163–180.

SG-CG/M490/L (2014). Overview of the main concepts of

ﬂexibility management. Technical Report Version 3.0,

CEN/CENELEC/ETSI.

Shekhar, S., Abdel-Aziz, H., Walker, M., Caglar, F., Gok-

hale, A., and Koutsoukos, X. (2016). A simulation as

a service cloud middleware. Annals of Telecommuni-

cations, 71(3):93–108.

Taylor, S. J. E., Khan, A., Morse, K. L., Tolk, A., Yilmaz,

L., Zander, J., and Mosterman, P. J. (2015). Grand

challenges for modeling and simulation: simulation

everywhere–from cyberinfrastructure to clouds to ci-

tizens. SIMULATION, 91(7):648–665.

Tolk, A. (2013). Interoperability, composability, and their

implications for distributed simulation: Towards mat-

hematical foundations of simulation interoperability.

In 17th IEEE/ACM International Symposium on Dis-

tributed Simulation and Real Time Applications.

Tolk, A. and Mittal, S. (2014). A necessary paradigm

change to enable composable cloud-based m&s ser-

vices. In Proceedings of the 2014 Winter Simulation

Conference.

Verborgh, R., van Hooland, S., Cope, A. S., Chan, S., Man-

nens, E., and de Walle, R. V. (2015). The fallacy of

the multi-api culture: Conceptual and practical bene-

ﬁts of representational state transfer (rest). Journal of

Documentation, 71(2):233–252.

Wainer, G. and Wang, S. (2017). Mams: Mashup architec-

ture with modeling and simulation as a service. Jour-

nal of Computational Science, 21:113–131.

Wang, S. and Wainer, G. (2016). Modeling and simulation

as a service architecture for deploying resources in the

cloud. International Journal of Modeling, Simulation,

and Scientiﬁc Computing, 07(01):1641002.

ICINCO 2018 - 15th International Conference on Informatics in Control, Automation and Robotics

620