A Simulation-based Aid for Organisational Decision-making

Souvik Barat

, Vinay Kulkarni

, Tony Clark

and Balbir Barn

Tata Consultancy Services Research, Pune, India

Shefﬁeld Hallam University, London, U.K.

Middlesex University, London, U.K.

Keywords:

Organisational Decision Making, Enterprise Modelling Languages, Meta Modelling, Simulation.

Abstract:

Effective decision-making of modern organisation often requires deep understanding of various aspects of

organisation such as organisational goals, organisational structure, business-as-usual operational processes.

The large size of the organisation, its socio-technical characteristics, and fast business dynamics make this

endeavor challenging. Industry practice relies on human experts for comprehending various aspects of organi-

sation thus making organisational decision-making a time-, effort- and intellectually-intensive endeavor. This

paper proposes a model-based simulation approach to organisational decision-making. We illustrate how this

is applied to a real life problem from software service industry.

1 INTRODUCTION

Modern organisations need to respond to a variety of

changes while operating in a dynamic environment.

In order to minimise undesirable consequences such

as prohibitive costs of erroneous decisions and lack

of opportunities for later course correction, there is

a need for a-priori judicious evaluation of the avail-

able courses of action. The decision-makers are thus

expected to understand, analyze and correlate exist-

ing information about various aspects of enterprise

such as organisational goals, organisational structure,

operational processes, change drivers and their inﬂu-

ences on overall organisation. Large and complex

organisational structure, and inherent socio-technical

characteristics of the organisation (McDermott et al.,

2013) all contribute to the complexity of organisa-

tional decision-making.

Current industry practice relies mostly on human

experts for decision-making with spreadsheet, word

processors, and diagram editors being the most pop-

ular tools used for capturing the relevant information

about enterprise. The informal nature of the informa-

tion means the power, rigour, and speed of sophisti-

cated analysis due to automation cannot be brought

to bear upon the decision-making problem. As a re-

sult, the quality of the solution is largely dependent

on knowledge and experience of human experts in-

volved in the decision-making process. When this is

coupled with the sheer volume, heterogeneity of the

information, and the complexity of a dynamic envi-

ronment then the analysis is further untenable. Pro-

vision of solutions that are able to stitch together a

coherent, consistent and integrated view from these

pieces is challenging for decision-makers.

Enterprise Modelling (EM) tries to reduce the

complexity of organisational decision-making with a

range of concepts, languages and tools for represent-

ing and analyzing the aspects of organisation. For

instance, the Zachman framework (Zachman et al.,

1987) advocates six aspects namely why, what, how,

who, when and where for comprehensive representa-

tion of an organisation. Thus it can be argued that

complete speciﬁcation of enterprise is possible using

Zachman framework, however, no automated analy-

sis support is available. Examination of existing EM

reveals some interesting observations. Languages ca-

pable of specifying all the relevant aspects of enter-

prise for organisational decision-making lack support

for automated analysis (e.g., Archimate (Iacob et al.,

2012), IEM (Bernus and Schmidt, 2006) and EEML

(Krogstie, 2008)). Languages capable of automated

analysis only cater for a subset of the relevant as-

pects for decision-making (e.g., BPMN

, i* (Yu et al.,

2006), and System Dynamics (Meadows and Wright,

2008)). Co-simulation using a relevant subset of EM

languages can be a pragmatic solution (Barjis, 2008).

For instance, i* (to specify the why aspect), BPMN (to

www.omg.org/spec/BPMN/2.0/

Barat, S., Kulkarni, V., Clark, T. and Barn, B.

A Simulation-based Aid for Organisational Decision-making.

DOI: 10.5220/0005992401090116

In Proceedings of the 11th International Joint Conference on Software Technologies (ICSOFT 2016) - Volume 2: ICSOFT-PT, pages 109-116

ISBN: 978-989-758-194-6

109

Figure 1: Abstraction of organisational decision-making.

specify the how aspect) and Stock-n-ﬂow (to specify

the what aspect) can be used to come up with the nec-

essary and sufﬁcient speciﬁcation which is amenable

for analysis albeit in parts (Kulkarni et al., 2015). Hu-

man expertise is still required for the analysis of the

problem, selection of the appropriate EM technique

and the integration of the technology into a consis-

tent whole (Barjis, 2008). This intellectual challenge

is further exacerbated due to paradigmatically diverse

nature of the three languages and issues of interoper-

ability of the various tools.

A simulation-based approach to organisational

decision-making may offer a pragmatic solution.

However, simulation is known to deliver in situations

where mechanistic world view holds (Barjis, 2008)

whereas modern enterprises are socio-technical sys-

tems (McDermott et al., 2013) bringing additional

dimensions such as uncertainty, autonomicity and

adaptability to the problem space.

We propose a pragmatic model-based simula-

tion approach for analyzing organisations as socio-

technical systems. This analysis centric approach

hinges on: (i) the necessary and sufﬁcient information

for decision-making to exist in machine processable

manner, (ii) machinery for effective processing of this

information, and (iii) a method to enable repetitive

use of the machinery at the hands of knowledgeable

users. This paper addresses tenets (i) and (ii) while

hinting at (iii). This paper also describes a model-

based realization of the proposed approach. The ap-

proach is illustrated using an example from the soft-

ware services industry. The principal contributions of

this paper are: (i) a modelling abstraction for precise

and comprehensive representation of organisations as

socio-technical systems, (ii) application of a simu-

lation technique to support organisational decision-

making through repeatable scenario playing.

2 MOTIVATION

Organisational decision-making activity is an iterative

process for selecting appropriate course of actions for

achieving the desired goals of organisation.

The process starts with identifying possible goals

or course of actions or both. The successive steps

deal with validation of selected course of actions

against goals, ranking alternatives options, and selec-

tion of alternatives over the other. Our visualization

of decision-making is depicted in Fig. 1. As shown

in the ﬁgure, an organisation (O) comprises four ba-

sic elements Data (D), Structure (S), Processes (P)

and Goals (G) i.e., O = <D, S, P, G>. The data D

describes the current state of organisation and a set of

past states of interest, process P describes collection

of Business As Usual (BAU) behaviors, and goals G

specify the desired intention of organisation. Struc-

turally an organisation (i.e., S of O) is a composition

of interacting socio-technical units where each unit

can further be visualized as <D, S, P, G>tuple. An

organisation or its constituent unit manages its goals

G; goals G affect organisational BAU behaviours P;

and organisational BAU behaviours P is accountable

for state change leading to data update D. The avail-

able data D determines whether the stated goals G

are achieved or not. The meaningful state variables

used for evaluation of goals are called the Measures

(M). The Levers (L) are appropriate course of actions

the decision-maker can take for achieving the stated

goals. A lever is essentially the speciﬁcation of a

change to structure, process, goals or any combina-

tion thereof.

Thus, in this formulation, decision-making is

human-guided exploration of design space wherein

a set of levers L

select

from the available levers L are

selected for application, their effect on the relevant

set of measures is observed, the desired goals are

(re)evaluated using the new values of measures - this

loop continues till either the desired goals are met

or the desired goal is changed thus starting off an-

other iterate-till-saturate process. Critically, the abil-

ity to specify inﬂuence of a lever on a set of measures

is the key. The socio-technical nature of an enter-

prise and the inability to have complete understand-

ing of the problem space make speciﬁcation of lever-

to-measure inﬂuence in pure mathematical terms very

hard. Therefore, simulation seems to be the pragmatic

recourse.

2.1 Tenets of Decision-making

We argue that an organisation can be understood well

by knowing what an organisation is, how it operates

and why it is so (Kulkarni et al., 2015). Further clarity

can be obtained by considering organisational respon-

sibilities and understanding the who (i.e., responsible

stakeholders) aspect of the organisation. Therefore,

we consider the four aspects, namely what, how, why

and who, as necessary and sufﬁcient for specifying

ICSOFT-PT 2016 - 11th International Conference on Software Paradigm Trends

110

Figure 2: Conceptual model of organisation.

data, structure, process and goals of an organisation.

This is broadly aligned with the Zachman framework

(Zachman et al., 1987) except for the where and when

aspects which we believe are mostly subsumed within

what and how aspects.

This boils down to two primary requirements for

supporting organisational decision-making: (i) the

ability to capture why, what, how and who aspects of

an organisation, in a formal manner and (ii) the ability

to perform what-if and if-what analyses of the formal

speciﬁcation.

Decision makers expect help not only in identi-

fying the candidate set of levers to be applied at a

given state but also with quantitative as well as qual-

itative estimation of application of the selected levers

towards achievement of the stated goals.

3 PROPOSED SOLUTION

We propose a pragmatic approach to improve the cur-

rent state of organisational decision-making to help

decision-makers to analyze various what-if scenario

of a decision-making problem. We now describe a

model-based realisation of simulation-based analysis

approach. Firstly, we propose a conceptual model for

representing the why, what, how and who aspects of

organisation in a localised relatable manner. Further

we reﬁne this conceptual model to an implementa-

tion model and provide simulation semantics to en-

able what-if scenario playing thus enabling human-

guided exploration of solution space with enhanced

certainty. We argue how the proposed implementa-

tion model meets the desired tenets, describe an im-

plementation strategy, and outline a packaging that

practitioners may ﬁnd effective in real-life industry-

scale situations.

3.1 Conceptual Model

From an external stakeholder perspective, an organ-

isation can be viewed as something that responds to

a set of events as it goes about achieving its stated

goals. Organisations consist of many autonomous

units, organised into dynamically changing hierar-

chical groups, operating concurrently, and managing

goals that affect their behaviour. We describe struc-

ture and behaviour of an organisation using a small

set of concepts and their relationships as depicted in

Fig. 2.

A Unit is an autonomous self-contained functional

unit with high coherence and low external coupling.

It exposes Goals stating its intention, and it interacts

with environment through a set of In-Events and Out-

Events. Internally it contains a Behaviour, a set of

Internal Events and a Type Model. The type model

describes the schema for representing current and pre-

vious states of the organisation, i.e. Data and History.

A Unit may make use of several contained Units in or-

der to meet the promised goals. The contained units

can interact with each other to delegate their respon-

sibilities to others; a unit can also participate in hi-

erarchical composition structure to accomplish wider

goals of the organisation, e.g., a larger unit or an

organisation. A Unit has a set of Levers and Mea-

sures where levers are parameters that can be used

for conﬁguration purposes, and measures are mean-

ingful state variables that are exposed to the environ-

ment. Conceptually, the elements Unit, unit relation

and nesting capability represent the structure S, the

Event and Behaviour represent process P, Data and

History represent data D, Goal represent the goal G

of organisation O. On the other hand elements Unit,

Event, Data, History and nesting capability of Unit

are capable of specifying the what aspect, Goal spec-

iﬁes the why aspect, Behaviour speciﬁes the how as-

pect and Unit, as individual, speciﬁes the who aspect

of an organisation. Event helps to capture reactive na-

ture of Unit, the intent is captured using Goal, mod-

ularity is achieved through Unit, autonomy is possi-

ble due to the concept of Internal Event, and compo-

sition can be speciﬁed using nesting relation. Also,

Unit is adaptable as it can construct and reconstruct

its structure; modular as it encapsulates the structure

and behaviour of an organisation; intentional as it has

its own goals; and compositional as it can be an as-

sembly of Units.

We draw from a set of existing concepts to come

up with the unit abstraction. Modularization and re-

ﬂective unit hierarchy are taken from fractal compo-

nent models (Barros et al., 2009). Goal-directed reac-

tive and autonomous behaviour can be traced to agent

A Simulation-based Aid for Organisational Decision-making

111

Figure 3: Implementation Meta Model.

behaviour . Deﬁning states in terms of a type model

is borrowed from UML

. An event driven architec-

ture (Michelson, 2006) supports ﬂexible interactions

between components, and the concept of intentional

modelling (Yu et al., 2006) is adopted to enable spec-

iﬁcation of component goals.

3.2 Implementation

Fig. 3 describes the specialisation of the conceptual

model of Fig. 2 for implementation and simulation se-

mantic purposes. It can be read as follows.

Organisation is a Unit that comprises a set of Units

and strives to accomplish its stated Goal. It does so by

responding to Events taking place in its environment

(In-Events), processing them (as speciﬁed in Spec),

and by interacting with other external Units in terms

of Events raised/responded (OutEvents). A Unit may

choose not to expose all events to the external world

(InternalEvents). Spec (associated with Unit through

behavioralSpec) is a declarative speciﬁcation of event

processing logic i.e., behaviour of the Unit. Thus,

looking outside-in, a Unit is a Goal-directed agent that

receives events (InEvents), processes them, and raises

events (OutEvents) to be processed by other Units.

Also, Unit is a parameterized entity whose structure

and behaviour can be altered through Levers. The

lever speciﬁcation is a Spec that connects Unit with

leverSpec. Internally, a Unit has current and historic

states that comprise the instances of Event, Unit, As-

sociations and Attributes. The model provides two

abstractions namely Snapshots and Value to encapsu-

late instances. A Unit may choose not to expose the

entire state to the external world (InternalState). A

Unit interacts with other Units in a-priori well-deﬁned

http://www.omg.org/spec/UML/

Role-playing manner. TypeModel provides a type

system for structural as well as behavioural aspects of

a Unit. A language deﬁned using the metamodel from

Fig. 3, is termed as ESL (Enterprise Simulation Lan-

guage) and has a meaning that is deﬁned with respect

to trace semantics expressed as a sequence of Snap-

shots. Sequence of snapshots describes the history

and current state (i.e. D) of the organisation (O). Each

unit has a speciﬁcation that describes the organisa-

tional behaviour (P) and a unit conforms to a structure

(S). A behavioural speciﬁcation (Spec associated with

behaviouralSpec) is a predicate over traces that must

be true for the projection of the overall organisation-

trace that relates to the appropriate unit. Each unit has

a goal (G) that governs its intent and behaviour. The

semantics of goals are predicates over state traces or

snapshots (i.e. D). Selected snapshots and slots can

be marked as measure (M) for quantitative measure-

ments. Unlike speciﬁcations, goals need not be true

for every legal behavioural-trace: a goal may fail and

it is the job of each individual unit to perform actions

in order to achieve its goal. In a simulation scenario,

the lever l ∈ L can be selected though appropriate pa-

rameter value to lever Specs or modifying lever Specs.

The observation of measures (M) is evaluation of ap-

propriate snapshot values. Thus the scenario playing

formulation i.e., M = P

<t0−t1>

, L

select

(O)), is es-

sentially setting initial simulation value (D

), select-

ing L

select

from possible levers L, executing BAU pro-

cess P for duration t1-t0 and evaluating of appropriate

snapshot values.

ESL was prototyped by extending an existing

event-driven language LEAP (Clark and Barn, 2013)

with the concepts borrowed from actor model of

computation (Hewitt, 2010), multi-agent systems

(Van Harmelen et al., 2008), and goals (Yu et al.,

2006). These concepts and their augmentation with

ICSOFT-PT 2016 - 11th International Conference on Software Paradigm Trends

112

Figure 4: Models of Software Service Provisioning Organisation (SSPO).

conventional class models and temporal logic closely

match the required features speciﬁed in Fig. 3. The

ESL and simulation engine for ESL is implemented

using DrRacket

4 ILLUSTRATIVE EXAMPLE

In this section we evaluate our approach by present-

ing a modelling and sample decision-making scenario

http://racket-lang.org/

of a software service-provisioning organisation. We

consider an organisation that earns revenue by devel-

oping bespoke software for its customers. The or-

ganisation bids for various software projects in re-

sponse to request for proposals (RFPs). Once a bid is

won, the organisation initiates and executes projects

using tried-and-tested process. This business as usual

(BAU) scenario of the organisation is driven by high

level goals of securing leadership position in terms of

business volume, proﬁtability and customer satisfac-

tion.

The implementation of this case study example is

A Simulation-based Aid for Organisational Decision-making

113

detailed and can be seen to approximating to real life.

The detail has considered various kinds of projects,

different execution strategies and resource catego-

rization derived from industry. But in the interest

of size, here we consider a part of the case-study

by limiting to a simple project classiﬁcation and a

relatively non-disruptive strategy for illustration pur-

poses. Hence, the environment is characterised us-

ing a representative classiﬁcation with customers of-

fering four different kinds of projects – High Mar-

gin High Risk (HMHR) project, Low Margin Low Risk

(LMLR) project, Medium Margin High Risk (MMHR)

and Medium Margin Low Risk (MMLR). The strategy

is implemented using four levers namely “Increase

Win Rate”, “New Opportunity Stream”, “Increase Re-

source Strength” and “Improve resource Skill”. Mod-

els and decision-making process for selecting the

levers with best potential for achieving the desired or-

ganisational goal are illustrated below.

4.1 Models

The model of the software service provisioning or-

ganisation is illustrated in Fig. 4. The organisation is

visualized as a unit (SSPO unit) with four in-events,

four out-events and an organisational goal as shown in

Fig. 4 a. The key elements of the model are illustrated

below:

Goal speciﬁcation: SSPO unit targets a primary

goal namely “Securing Leadership Position”. This

goal is decomposed into three sub-goals namely

“Increase Business Volume”, “Increase Proﬁtability”

and “Improve Customer Satisfaction” to support bet-

ter qualitative and quantitative measurements. The

“Increase Proﬁtability” sub-goal is further decom-

posed into two sub-goals namely “Increase Revenue”

and “Reduce Expenditure”. These goals and sub-

goals are described using predicates where terms

are ﬁnally associated with TypeModel shown in

Fig. 4 c. For instance, “Increase Business Vol-

ume” is associated with “business volume” attribute

of “Sales Record” class, “Improve Customer satis-

faction” is associated with two attributes of “Project

Delivery” class - “project completed ontime” and

“project completed with delay” (where these two at-

tributes contribute positively and negatively respec-

tively towards “Improve Customer satisfaction”), “In-

crease Revenue” is associated with “revenue” of “Ac-

count” class and “Reduce Expenditure” is associated

with “expenditure” of Account class of the Type-

Model.

In and Out Events: The SSPO unit inter-

acts with environment by receiving “rfp(RFP)”

event, “bidResponse(BidResponse)” event, and “pay-

ment(Payment)” event from environment (in partic-

ular from customers), and responding “bid(Bid)”

event and “deliver(Deliverable)” event to the cus-

tomers. It also receives “join(Resource)” event from

various sources, sends “offers(Offer)” event to the

Resources who are outside of SSPO organisation,

and send Resources to the environment using “sep-

arates(Resources)” event for resign, termination and

retirement. Events use TypeModel for specifying

their parameters.

Internal events and organisation structure: In-

ternal units, internal events and their interactions

are depicted using component model in Fig. 4 b.

The ﬁgure shows four sub-units namely “Sales”,

“Delivery”, “Account” and “Resource Management”

units with their in and out events. Interactions be-

tween SSPO and sub-unit, and between sub-units are

also illustrated. For example “allocation(Resource)”

and “deAllocation(Resource)” events are the in-

teractions between “Delivery” unit and “Resource

Management” unit whereas the delegation of event

“rfp(RFP)” is the interaction between SSPO and sub-

unit. The interaction and structure can be static or dy-

namic. For example the “CustomerProject” is a unit

is created once a Bid is won by SSPO unit and it re-

solves after producing “deliver(Deliverable)” event.

Behaviour: a simpliﬁed behaviour of SSPO unit

is depicted using state diagrams in Fig. 4 d. The

behaviour shows the transformation and life-cycle

RFP. SSPO unit receives “RFP” through “rfp(RFP)”

event; it then delegates to “Sales” unit; “Sales”

unit works on “RFP” and transforms it to “Bid”;

the “Bid” is transformed into a “CustomerProject”

when a bid is won by SSPO (the intimation re-

ceives through “bidResponse(Bidresponse)”); inter-

nally the “CustomerProject” goes through many

states and ﬁnally dissolves by responding event “de-

liver(Deliverables)” and deallocating resource using

“deAllocate(Resource)”.

Measures: measures are state variables or condi-

tion over state variables. The attributes which are

used in measures are highlighted in TypeModel of

Fig. 4 c. The measures within SSPO unit are shown

in Fig. 4 b. We consider 6 measures for SSPO

unit - “Business Volume”, “Revenue”, “Expense”,

“Proﬁtability” and “Customer Satisfaction”. “Busi-

ness Volume” measure represents the slot value “busi-

ness volume” attribute of “Sales Record” class, “Rev-

enue” measure represents the slot value of “revenue”

attribute of “Account” class, and “Expenditure“ mea-

sure represents the slot value of “expenditure“ at-

tribute of “Account” class of TypeModel shown in

Fig. 4 c. Similarly, the “Proﬁtability” represents the

conditions on the slot values of “revenue” and “ex-

ICSOFT-PT 2016 - 11th International Conference on Software Paradigm Trends

114

Figure 5: Decision Making and Simulation Results.

penditure” attributes of “Account” class, and “Cus-

tomer Satisfaction” is for representing the condition

on the slot values of “project completed ontime” and

“project completed with delay” attributes of “Project

Delivery” class.

Levers: the levers are the condition over events

and its parameters in the context of behaviour. In

this example, we consider 4 basic levers “Increase

Win Rate”, “New Opportunity Stream”, “Increase Re-

source Strength” and “Improve Resource Skill”.

4.2 Decision-making

The decision-making process is about ﬁnding possi-

ble levers (L), evaluating them with respect to organ-

isational goals and sub-goals (G), and selecting a set

of levers (L

select

∈ L) that have the best potential to

achieve the goal G. Simulation-based what-if scenario

playing forms a cornerstone of this process.

Fig. 5 a shows simulation needs in a consolidated

form (upper portion) and the rest of the sub-ﬁgures in

Fig. 5 represent the simulation outputs from DrRacket

based ESL prototype. Goal and sub-goals G form

columns of the table (depicted in Fig. 5 a) with levers

L forming the rows. Each cell of the table represents a

what-if scenario for a lever l ∈ L on goal g ∈ G where

the impact of a lever on a goal needs to be computed

using simulation run. For example, ﬁrst row in the

table corresponds to “Increase Win Rate” lever. As

can be seen, this lever has a positive impact on “Busi-

ness Volume” sub-goal, marginally positive impact on

“Revenue”, “Expense” and “Proﬁtability” sub-goals,

and eventual negative impact on “Customer Satisfac-

tion” sub-goal. As a result, nothing conclusive can be

said about impact of “Increase Win Rate” lever on the

overall goal of “Secure Leadership Position”. The left

half of ﬁgure Fig. 5 b depicts initial state of the organ-

isation in terms of RFPs received, RFPs responded,

RFPs won (i.e., “Business Volume”), On-time de-

livery, Delayed delivery, Project execution pipeline

build-up etc without applying any levers. Signiﬁcant

points to be noted are: all projects are delivered on

time, and there is no project execution pipeline build-

up. Right half of Fig. 5 b depicts details of simula-

tions carried out to determine the impact of “Increase

Win Rate” lever on various sub-goals. As can be seen,

A Simulation-based Aid for Organisational Decision-making

115

“Business Volume” increases by about 30% but there

is signiﬁcant increase in the number of projects de-

livered with a delay some of which leads to penal-

ties. As a result, proﬁts do not increase in the same

proportion as increase in “Business Volume”. Also,

build-up in project execution pipeline is a concern

that can lead to customer dissatisfaction that can po-

tentially impact overall goal adversely. Fig. 5 c and

Fig. 5 d. depict impact of levers “New Opportunity

Stream” and “Increase Resource Strength” on the var-

ious sub-goals. Comparison of ﬁgures Fig. 5 b. and

Fig. 5 c. shows the “Proﬁtability” of “New Opportu-

nity Stream” is much higher than the “Proﬁtability”

of “Increase Win Rate” however the factors associ-

ated with negative “Customer Satisfaction” are also

high. On other hand, “Increase Resource Strength”

shows positive impact on “Customer Satisfaction” but

with an additional cost that brings down “Proﬁtabil-

ity”. Thus, as can be seen from the ﬁrst four rows of

ﬁgure Fig. 5 a, no lever individually can ensure the

overall goal of “Secure Leadership Position” can be

achieved. As a result, one has to explore what impact

a combination of these levers can have. For example

one can evaluate the combination of levers “Increase

Win Rate” and “Increase Resource Strength” or levers

“New Opportunity Stream” and “Increase Resource

Strength”. Fig. 5 e. shows impact of levers “Increase

Win Rate”, “New Opportunity Stream” and “Increase

Resource Strength” applied together. As can be seen

from Fig. 5 a, this conclusively leads to achievement

of the overall goal. Further simulation can be done to

ﬁne tune the options by deciding quantitative ﬁgures.

5 CONCLUSION

Organisational decision-making practice today relies

excessively on human expertise. This is primarily

due to unavailability of suitable technology support.

Available technology support is found inadequate ei-

ther in completeness of speciﬁcation of all relevant

aspects of decision-making or in analysis rigour or

both. This paper has presented a conceptual model,

the accompanying implementation model that forms

the basis of a high-level language and its simulation

semantics.

The approach has been illustrated with a substan-

tive example from the software services domain. We

have shown the example can be modeled and sim-

ulated leading to the ability to inﬂuence the strate-

gically selected measures. However, we recognise

that the current implementation model (Fig. 3) of

ESL is not sufﬁciently high-level for direct adop-

tion by decision-makers. Our immediate next step

is to develop high-level abstractions to support the

core concepts of Fig. 3 in a business facing manner.

In doing so, we will adopt language processing and

model transformation technology to enable support

for deﬁning domain speciﬁc languages geared for spe-

ciﬁc problems.

REFERENCES

Barjis, J. (2008). Enterprise, organization, modeling, simu-

lation: Putting pieces together. CEUR Workshop Pro-

ceedings, 338:1–5.

Barros, T., Ameur-Boulifa, R., Cansado, A., Henrio, L., and

Madelaine, E. (2009). Behavioural models for dis-

tributed fractal components. Annales des Telecommu-

nications/Annals of Telecommunications, 64(1-2):25–

43.

Bernus, P. and Schmidt, G. (2006). Architectures of in-

formation systems. In Handbook on Architectures of

Information Systems, pages 1–9. Springer.

Clark, T. and Barn, B. (2013). Goal driven architecture

development using leap. Enterprise Modelling and

Information Systems Architectures-An International

Journal, 8(1):40–61.

Hewitt, C. (2010). Actor model of computation: scal-

able robust information systems. arXiv preprint

arXiv:1008.1459.

Iacob, M., Jonkers, D. H., Lankhorst, M., Proper, E., and

Quartel, D. D. (2012). Archimate 2.0 speciﬁcation:

The open group. Van Haren Publishing.

Krogstie, J. (2008). Using eeml for combined goal and pro-

cess oriented modeling: A case study. CEUR Work-

shop Proceedings, 337:112–129.

Kulkarni, V., Barat, S., Clark, T., and Barn, B. (2015). To-

ward overcoming accidental complexity in organisa-

tional decision-making. In 17th Intl. Conf. on Model

Driven Engineering Languages and Systems (MoD-

ELS 15). IEEE.

McDermott, T., Rouse, W., Goodman, S., and Loper,

M. (2013). Multi-level modeling of complex socio-

technical systems. Procedia Computer Science,

16:1132–1141.

Meadows, D. H. and Wright, D. (2008). Thinking in sys-

tems: A primer. Chelsea Green Publishing.

Michelson, B. M. (2006). Event-driven architecture

overview. Patricia Seybold Group, 2.

Van Harmelen, F., Lifschitz, V., and Porter, B. (2008).

Handbook of knowledge representation, volume 1. El-

sevier.

Yu, E., Strohmaier, M., and Deng, X. (2006). Exploring

intentional modeling and analysis for enterprise ar-

chitecture. 10th IEEE International Enterprise Dis-

tributed Object Computing Conference Workshops.

Zachman, J. et al. (1987). A framework for information sys-

tems architecture. IBM systems journal, 26(3):276–

292.

ICSOFT-PT 2016 - 11th International Conference on Software Paradigm Trends

116