Conceptual Modelling for Simulation Part I: Definition and Requirements

Author(s): S. Robinson
Source: The Journal of the Operational Research Society, Vol. 59, No. 3 (Mar., 2008), pp.
278-290
Published by: Palgrave Macmillan Journals on behalf of the Operational Research Society
Stable URL: http://www.jstor.org/stable/30132749
Accessed: 21-06-2016 12:25 UTC

REFERENCES
Linked references are available on JSTOR for this article:
http://www.jstor.org/stable/30132749?seq=1&cid=pdf-reference#references_tab_contents
You may need to log in to JSTOR to access the linked references.

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
http://about.jstor.org/terms

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted
digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about
JSTOR, please contact support@jstor.org.

Operational Research Society, Palgrave Macmillan Journals are collaborating with JSTOR to
digitize, preserve and extend access to The Journal of the Operational Research Society

This content downloaded from 194.47.65.106 on Tue, 21 Jun 2016 12:25:32 UTC
All use subject to http://about.jstor.org/terms
Journal of the Operational Research Society (2008) 59, 278-290 C 2008 Operational Research Society Ltd. All rights reserved. 0160-5682/08 $30.00

www.palgravejournals.com/jors

Conceptual modelling for simulation Part I:

definition and requirements
S Robinson*

University of Warwick, Coventry, UK

Conceptual modelling is probably the most important aspect of a simulation study. It is also the most difficult
and least understood. Over 40 years of simulation research and practice have provided only limited information
on how to go about designing a simulation conceptual model. This paper, the first of two, discusses the
meaning of conceptual modelling and the requirements of a conceptual model. Founded on existing literature, a
definition of a conceptual model is provided. Four requirements of a conceptual model are described: validity,
credibility, utility and feasibility. The need to develop the simplest model possible is also discussed. Owing
to a paucity of advice on how to design a conceptual model, the need for a conceptual modelling framework
is proposed. Built on the foundations laid in this paper, a conceptual modelling framework is described in the
paper that follows.
Journal of the Operational Research Society (2008) 59, 278-290. doi:10.1057/palgrave.jors.2602368
Published online 24 January 2007

Keywords: discrete-event simulation; model development; conceptual model; validation

Introduction This somewhat ad hoc approach does not seem satisfactory for
such an important part of the simulation modelling process.
Conceptual modelling is the process of abstracting a model
This paper is the first of two papers that attempt to bring
from a real or proposed system. It is almost certainly the
more clarity to the area of conceptual modelling for simula-
most important aspect of a simulation project. The design of
tion. The issue is addressed first by defining the meaning of
the model impacts all aspects of the study, in particular the
conceptual modelling and establishing the requirements of a
data requirements, the speed with which the model can be
conceptual model. These are the subjects of this paper. Hav-
developed, the validity of the model, the speed of experimen-
ing provided a foundation for conceptual model development,
tation and the confidence that is placed in the model results.
the paper that follows describes a framework for developing
A well-designed model significantly enhances the possibility
a conceptual model (Robinson, 2007).
that a simulation study will be a success.
The domain of interest for this discussion is primarily in
Although effective conceptual modelling is a vital aspect of
the use of discrete-event simulation for modelling operations
a simulation study, it is probably the most difficult and least
systems or operating systems. 'An operating system is a con-
understood (Law, 1991). There is surprisingly little written
figuration of resources combined for the provision of goods
on the subject. It is difficult to find a book that devotes more
or services' (Wild, 2002). Wild identifies four specific func-
than a handful of pages to the design of the conceptual model.
tions of operations systems: manufacture, transport, supply
Neither are there a plethora of research papers, with only a
and service. This is one of the prime domains for simula-
handful of well-regarded papers over the last four decades.
tion in operational research. We might refer to it as 'business
A search through the academic tracks at major simulation
oriented' simulation while interpreting business in its widest
conferences on discrete-event simulation reveals a host of
sense to include, for instance, the public sector and health.
papers on other aspects of simulation modelling. There are,
Models in this domain tend to be of a relatively small scale,
however, very few papers that give any space to the subject
with a project life-cycle of normally less than six months
of conceptual modelling.
(Cochran et al, 1995). The models are generally developed
The main reason for this lack of attention is probably due
by a lone modeller acting as an external or internal consultant.
to the fact that conceptual modelling is more of an 'art'
Sometimes the models are developed on a 'do-it-yourself' ba-
than a 'science' and therefore it is difficult to define methods
sis with a domain expert carrying out the development. This
and procedures. Whatever the reason, the result is that the
is somewhat different to the nature of simulation modelling in
art of conceptual modelling is largely learnt by experience.
the military domain, another major application of simulation
in operational research, where models tend to be of a much
* Correspondence: S Robinson, Warwick Business School, University of
Warwick, Coventry CV4 7AL, UK. larger scale and where they are developed by teams of peo-
E-mail: stewart.robinson @ warwick.ac.uk ple (Robinson, 2002). Although the focus is on discrete-event

This content downloaded from 194.47.65.106 on Tue, 21 Jun 2016 12:25:32 UTC
All use subject to http://about.jstor.org/terms
 S Robinson-Conceptual modelling for simulation 279

simulation for modelling operations systems, this is not to say

Line B
that the concepts do not have wider applicability. Final Dress Hot Test

In this paper, the meaning of the term conceptual model is

discussed in relation to existing definitions. A refined defi-
Head Line
nition of a conceptual model is then given and the scope Line A

of conceptual modelling is defined. There is a pause for

thought concerning the purpose of a conceptual model before
a discussion on the requirements of a conceptual model. The Figure 1 Schematic showing the layout of the South Wales
engine assembly plant.
paper finishes with a brief review of the guidance that is
available for conceptual modelling. The prime contributions
of this paper are to provide a definition of a conceptual model
and to identify the requirements for a conceptual model. A, the engine is loaded to a Line B platen to continue the
Throughout the paper, three roles in a simulation study are assembly process. The empty Line A platen is washed and
assumed: returned so a new engine block can be loaded. At the end of
Line B, completed engines are off-loaded and move to the Hot
* The Clients: The problem owners and recipients of the Test facility. In Hot Test, engines are rigged to test machines,
results. run for a few minutes and monitored. Engines that pass Hot
* The Modeller: The developer of the model. Test move to the Final Dress area for completion. Engines
* Domain Experts: Experts in the domain being modelled that fail Hot Test are rectified and then completed.
who provide data and information for the project. The majority of the operations on the three main assembly
lines consist of a single automatic machine. Some operations
These roles do not necessarily imply individual or separate require two parallel machines due to the length of the machine
people. There are often many clients and domain experts cycle, while a few other operations are performed manually.
involved in a simulation study. In some situations one of the At various points along the line there are automatic test sta-
clients or domain experts may also act as the modeller. tions. When an engine fails the test, it is sent to an adjoining
Before exploring the meaning of conceptual modelling, let rework station, before returning to be tested again. All the op-
us begin with an example that highlights how more than one erations are connected by a powered roller conveyor system.
(conceptual) model can be developed of the same system. The key components are the engine block, head, crankshaft,
cam shaft and connecting rods. These are produced at nearby
production facilities, delivered to the main assembly plant
Example: Modelling the Ford Motor Company South
and stored line-side ready for assembly. Because various
Wales engine assembly plant
engine derivatives are made on the assembly line, a range of
I had been called in to carry out some simulation modelling component derivatives needs to be produced and stored for
of the new engine assembly plant that Ford Motor Company assembly. The result was the concern over scheduling the
(Ford) was planning to build in South Wales. Faced with a production and the storage of these key components.
meeting room full of engineers I started, as normally I would, As with all such projects, time for developing and using
by asking what was the problem that they wished to address. the model was limited. It was important, therefore, to devise
There was a unanimous response: 'Scheduling! We are not a model that could answer the questions about scheduling
sure that there is enough space by the line to hold sufficient key components as quickly as possible while maintaining a
stocks of the key components. Obviously the schedules we satisfactory level of accuracy.
run on the key component production lines and on the main In considering the nature of the problem, it was clear that
engine assembly line will affect the inventory we need to the key issue was not so much the rate at which engines
hold'. After further questioning it was clear that they saw this progressed through the assembly line, but their sequence. The
as the key issue. In their view, there was no problem with initial sequence of engines was determined by the production
achieving the required throughput, especially because they schedule, but this sequence was then disturbed by engines
had designed a number of similar lines previously. being taken out for rework and by the presence of parallel
The engine assembly line was planned to consist of three machines for some operations. Under normal operation the
main assembly lines (with well over 100 operations), a Hot parallel machines would not cause a change in the sequence
Test facility and a Final Dress area. Figure 1 provides a of engines on the line, but if one of the machines breaks down
schematic of the line. On the first line (Line A), engine blocks for a period, then the engines queuing for that machine would
are loaded onto platens (metal pallets on which engines move be delayed and their sequence altered.
around the conveyor system) and then pass through a series It was recommended that the simulation model should rep-
of operations. On the Head Line various components are resent in detail those elements that determined the sequence
assembled to the head before the complete sub-assembly is of engines on the main assembly line, that is, the schedule,
joined with the engine block on Line A. On leaving Line the test and rework areas, and the parallel machines. All other

This content downloaded from 194.47.65.106 on Tue, 21 Jun 2016 12:25:32 UTC
All use subject to http://about.jstor.org/terms
280 Journal of the Operational Research Society Vol. 59, No. 3

operations could be simplified by grouping sections of the line model could have been developed, which could have an-
that consisted of individual machines and representing them swered both sets of questions. This, however, would have
as a queue with a delay. The queue capacity needed to equate taken much longer to develop and it would certainly have
to the capacity of that section of the line. The delay needed run much slower, restricting the extent of the experimentation
to be equal to the time it took for an engine to pass through possible. Anyway, the need for the second model was only
the section of the line, allowing for breakdowns. This would identified as a result of indications about throughput from the
give a reasonable approximation to the rate at which engines first model. Up to that point, a throughput model seemed
would progress through the facility. Of course, the operations unnecessary.
where the key components are assembled to the engine need A more fundamental question that should be asked is if
to be modelled in detail, along with the line-side storage areas very different models can be developed of the same system,
for those components. how can a modeller determine which model to use? Indeed,
Further to this, it was noted that detailed models of the key how can a modeller develop a model design, or a set of model
component production lines already existed. Alternative pro- designs from which to select? The only clue that comes
duction schedules for each line could be modelled separately from the example above is the importance of the modelling
from the engine assembly line model and the output from objectives in determining the nature of the model. Beyond
these models stored. The outputs could then be read into the this, modellers need some means for determining what to
engine assembly line model as an input trace stating the com- model. This process of taking a real-world situation and
ponent derivatives and their time of arrival at the assembly from it designing a model is often referred to as conceptual
line. Some suitable delay needed to be added to allow for modelling.
the transportation time between the key component lines and
the main assembly line. It was also unnecessary to model the
What is conceptual modelling?
Hot Test and Final Dress, as all of the key components have
been assembled prior to reaching these areas. Conceptual modelling is about abstracting a model from a
As a result of these simplifications, the model could be real or proposed system. All simulation models are simpli-
developed much more quickly and the final model ran much fications of reality (Zeigler, 1976). The issue in conceptual
faster, enabling a greater amount of experimentation in the modelling is to abstract an appropriate simplification of real-
time available. The model fulfilled its objectives, sizing the ity (Pidd, 2003). This provides some sense of what concep-
line side storage areas and showing that shortages of key com- tual modelling is, but only in the most general of terms. How
ponents were unlikely. What the model did suggest, however, can the terms conceptual model and conceptual modelling be
was that there may be a problem with throughput. more precisely defined? Existing literature may shed some
Although the scheduling model indicated a potential prob- light on this topic.
lem with throughput, it did not contain enough detail to give In general, the notion of conceptual modelling, as expressed
accurate predictions of the throughput of the engine assembly in the simulation and modelling literature, is vague and ill-
line. As a result, a second model was developed with the ob- defined, with varying interpretations as to its meaning. What
jective of predicting and helping to improve the throughput of seems to be agreed is that it refers to the early stages of a
the facility. This model represented each operation in detail, simulation study. This implies a sense of moving from the
but on this occasion did not represent the arrival and assem- recognition of a problem situation to be addressed with a
bly of key components. It was assumed that the key compo- simulation model to a determination of what is going to be
nents would always be available, as had been suggested by modelled and how. Balci (1994) breaks the early parts of a
the scheduling model. simulation study down into a number of processes: problem
The second (throughput) model indeed confirmed that the formulation, feasibility assessment of simulation, system and
throughput was likely to fall significantly short of that required objectives definition, model formulation, model representa-
by Ford and identified a number of issues that needed to be tion and programming. Which of these is specifically included
addressed. Over a period of time, by making changes to in conceptual modelling is not identified. What is clear from
the facility and performing further simulation experiments, Balci and other authors, for instance Willemain (1995), is
improvements were made such that the required throughput that these early stages of a modelling study are not just vis-
could be achieved. ited once, but that they are continually returned to through
This example demonstrates how two very different sim- a series of iterations in the life-cycle of a project. As such,
ulation models can be developed of the same system. But conceptual modelling is not a one-off process, but one that is
which model was the right one? The answer is both, since repeated and refined a number of times during a simulation
both answered the specific questions that were being asked study.
of them. Underlying the differences between the models was Zeigler (1976) sheds some light on the subject by identify-
the difference in the modelling objectives. Neither simula- ing five elements in modelling and simulation from the 'real
tion model would have been useful for meeting the objectives system' through to the 'computer' (the computer-based sim-
of the other model. Of course, a single all encompassing ulation model). In between is the 'experimental frame', 'base

This content downloaded from 194.47.65.106 on Tue, 21 Jun 2016 12:25:32 UTC
All use subject to http://about.jstor.org/terms
 S Robinson--Conceptual modelling for simulation 281

model' and 'lumped model'. The experimental frame is the model. The paper describes a plethora of views, but con-
limited set of circumstances under which the real system is cludes by identifying two types of conceptual model. A
observed, that is, specific input-output behaviours. The base domain-oriented model that provides a detailed represen-
model is a hypothetical complete explanation of the real sys- tation of the problem domain and a design-oriented model
tem, which is capable of producing all possible input-output that describes in detail the requirements of the model. The
behaviours (experimental frames). The base model cannot be latter is used to design the model code. Meanwhile, Haddix
fully known since full knowledge of the real system cannot (2001, www.sisostds.org, accessed February 2006) points
be attained. For instance, almost all systems involve some out that there is some confusion over whether the conceptual
level of human interaction that will affect its performance. model is an artefact of the user or the designer. This may,
This interaction cannot be fully understood since it will vary to some extent, be clarified by adopting the two definitions
from person-to-person and time-to-time. above.

In the lumped model, the components of a model are The approach of military simulation modellers can be quite
lumped together and simplified. The aim is to generate a different to that of those working in business-oriented simula-
model that is valid within the experimental frame, that is, re- tion (Robinson, 2002). Military simulations often entail large-
produces the input-output behaviours with sufficient fidelity. scale models developed by teams of software developers.
The structure of the lumped model is fully known. Return- There is much interest in model reuse and distributed simula-

ing to the example of human interaction with a system, in a tion, typified by the High Level Architecture (DMSO, 2005,
lumped model specific rules for interaction are devised for http://www.dmso.mil/public/transition/hla/, accessed Febru-
example a customer will not join a waiting line of more than ary 2006). Business-oriented simulations tend to be smaller
ten people. in scale, involve lone modellers normally using a visual in-
Nance (1994) separates the ideas of conceptual model and teractive modelling system (Pidd, 2004), and the models are
communicative model. The conceptual model exists in the often thrown-away on completion of a project. Interest in dis-
mind of a modeller, the communicative model is an explicit tributed simulation is moderate, mostly because the scale and
representation of the conceptual model. He also specifies that life-time of the models does not warrant it (Robinson, 2005).
the conceptual model is separate from model execution. In As a result, although the definition and requirements for con-
other words, the conceptual model is not concerned with how ceptual modelling may be similar in both these domains, some
the computer-based model is coded. Fishwick (1995) takes account must be made of the differences that exist.

a similar view, stating that a conceptual model is vague and In summary, the discussion above identifies some key facets
ambiguous. It is then refined into a more concrete executable of conceptual modelling and the definition of a conceptual
model. The process of model design is about developing and model:

refining this vague and ambiguous model and creating the

model code. In these terms, conceptual modelling is a sub- * Conceptual modelling is about moving from a problem
set of model design, which also includes the design of the situation, through model requirements to a definition of
model code. what is going to be modelled and how.
The main debate about conceptual modelling and its def- * Conceptual modelling is iterative and repetitive, with the
inition has been held among military simulation modellers. model being continually revised throughout a modelling
Pace has lead the way in this debate and defines a conceptual study.
model as 'a simulation developer's way of translating mod- * The conceptual model is a simplified representation of the
elling requirements.., into a detailed design framework..., real system.
from which the software that will make up the simulation * The conceptual model is independent of the model code or
can be built' (Pace, 1999, www.sisostds.org, accessed Febru- software (while model design includes both the conceptual
ary 2006). In short, the conceptual model defines what is to model and the design of the code (Fishwick, 1995)).
be represented and how it is to be represented in the simula- * The perspective of the client and the modeller are both
tion. Pace sees conceptual modelling as being quite narrow important in conceptual modelling.
in scope viewing objectives and requirements definition as
precursors to the process of conceptual modelling. The con- It is clear, however, that complete agreement does not exist
over these facets.
ceptual model is largely independent of software design and
implementation decisions. Pace (2000a, www.sisostds.org,
accessed February 2006) identifies the information provided
A definition of a conceptual model
by a conceptual model as consisting of assumptions, algo-
rithms, characteristics, relationships and data. Following the discussion above, Figure 2 defines a conceptual
Lacy et al (2001, www.sisostds.org, accessed February model as shown by the area within the dashed ellipse. It also
2006) further this discussion reporting on a meeting of the places it within the wider context of a simulation study as de-
Defence Modelling and Simulation Office (DMSO) to try fined in Robinson (2004). Figure 2 shows four key processes
and reach a consensus on the definition of a conceptual in the development and use of a simulation model: conceptual

This content downloaded from 194.47.65.106 on Tue, 21 Jun 2016 12:25:32 UTC
All use subject to http://about.jstor.org/terms
282 Journal of the Operational Research Society Vol. 59, No. 3

Real world

(problem situation)
.0

Conceptual Model

Modelling and
general project
objectives
Solutions/

understanding
Aodel content:
Experimental Accepts Provides
scope and nses
factors level of detail

In.uts

Computer
model

Figure 2 The conceptual model in the simulation project life-cycle (revised from Robinson, 2004).

modelling, model coding, experimentation and implementa- of the model and its use (eg requirements for the flexibil-
tion. The outcome of each process is, respectively, a con- ity of the model, run-speed, visual display, ease-of-use and
ceptual model, a computer model, solutions to the problem model/component reuse). The definition of objectives is seen
situation and/or a better understanding of the real world and as intrinsic to decisions about the conceptual model. The
improvements to the real world. The double arrows illustrate Ford example above highlighted how different modelling ob-
the iterative nature of the process and the circular diagram jectives led to different models. Similarly, the general project
illustrates the potential to repeat the process of improvement objectives can affect the nature of the model. A shorter time-
through simulation a number of times. Missing from this di- scale, for instance, may require a simpler conceptual model
agram are the verification and validation activities involved than would have been devised had more time been available.

in a simulation study. These are carried out in parallel with For this reason, the objectives are included in the definition
each of the four processes outlined in Figure 2. For a more of the conceptual model.
detailed description of this life-cycle and model verification Including the modelling objectives as part of the definition
and validation, see Robinson (2004). of a conceptual model is at odds with Pace (1999). He sees
Based upon an understanding of the problem situation, the objectives and requirements definition as separate from the
which sits outside the conceptual model, the conceptual model conceptual model. The author's view is that while understand-
is derived. This model is only a partial description of the real ing the problem situation and the aims of the organization
world, but it is sufficient to address the problem situation. The lies within the domain of the real world (problem situation),
double arrow between the problem situation and objectives the modelling objectives are specific to a particular model and
signifies the interplay between problem understanding and modelling exercise. Different modelling objectives lead to
modelling. While the conceptual model reflects the under- different models within the same problem situation, as in the
standing of the problem situation, the process of developing Ford example. As a result, the modelling objectives are in-
the conceptual model also changes the understanding of the trinsic to the description of a conceptual model. Without the
problem situation. In particular, the nature of the questions modelling objectives, the description of a conceptual model
that the modeller asks during conceptual modelling can lead is incomplete.
to new insights on behalf of the clients and domain experts. The inputs (or experimental factors) are those elements of
At a greater extreme, ideas derived purely from conceptual the model that can be altered to effect an improvement in,
modelling may be implemented in the real system, changing or better understanding of, the problem situation. They are
the actual nature of the problem situation. determined by the objectives. Meanwhile, the outputs (or
The conceptual model itself consists of four main compo- responses) report the results from a run of the simulation
nents: objectives, inputs (experimental factors), outputs (re- model. These have two purposes: first, to determine whether
sponses) and model content. Two types of objective inform the modelling objectives have been achieved; second, to point
a modelling project. First, there are the modelling objec- to reasons why the objectives are not being achieved, if they
tives, which describe the purpose of the model and mod- are not.

elling project. Second, there are general project objectives Finally, the model content consists of the components that
which include the time-scales for the project and the nature are represented in the model and their interconnections. The

This content downloaded from 194.47.65.106 on Tue, 21 Jun 2016 12:25:32 UTC
All use subject to http://about.jstor.org/terms
 S Robinson-Conceptual modelling for simulation 283

content can be split into two dimensions (Robinson, 1994): code considerations from decisions about the conceptual
design.
* The scope ofthe model: The model boundary or the breadth The definition does not place the conceptual model at a
of the real system that is to be included in the model. specific point in time during a simulation study. This re-
* The level of detail: The detail to be included for each flects the level of iteration that may exist in simulation work.
component in the model's scope. A conceptual model may reflect a model that is to be de-
veloped, is being developed or has been developed in some
The model content is determined, in part, by the inputs and software. The model is continually changing as the simula-
outputs, in that the model must be able to accept and interpret tion study progresses. Whatever stage has been reached in
the inputs and to provide the required outputs. The model a simulation study, the conceptual model is a non-software-
content is also determined by the level of accuracy required. specific description of the model as it is understood at that
More accuracy generally requires a greater scope and level of point in time. That said, the prime interest of this paper is
detail.
in the role of the conceptual model during conceptual mod-
While making decisions about the content of the model, elling, which implies it is describing a computer model that
various assumptions and simplifications are normally intro- is yet to be developed, or at least the development is not yet
duced. These are defined as follows:
complete.

* Assumptions are made either when there are uncertainties

or beliefs about the real world being modelled. Conceptual modelling defined
* Simplifications are incorporated in the model to enable Put simply, conceptual modelling is the process of creating
more rapid model development and use, and to improve the conceptual model. Based on the definition given above
transparency.
this requires the following activities:

Assumptions and simplifications are identified as separate

* understanding the problem situation (a precursor to con-
facets. Assumptions are ways of incorporating uncertainties
ceptual modelling),
and beliefs about the real world into the model. Simplifica-
* determining the modelling and general project objectives,
tions are ways of reducing the complexity of the model. As
* identifying the model outputs (responses),
such, assumptions are a facet of limited knowledge or pre-
* identify the model inputs (experimental factors),
sumptions, while simplifications are a facet of the desire to
* determining the model content (scope and level of detail),
create simple models.
identifying any assumptions and simplifications.
Based on these ideas a conceptual model is defined as
follows:
These activities are explored in more detail in the paper that
follows (Robinson, 2007). This suggests a general order in
The conceptual model is a non-software specific description of which the elements of a conceptual model might be deter-
the computer simulation model (that will be, is or has been mined. There is likely to be a lot of iteration forwards and
developed), describing the objectives, inputs, outputs, content, backwards between these activities. Further to this, there is
assumptions and simplifications of the model.
iteration between conceptual modelling and the rest of the
process of model development and use (Robinson, 2004).
This definition adds the point that the conceptual model is Having said that the conceptual model is independent of the
non-software specific in line with the views of the other au- modelling software, it must be recognized that there is an in-
thors described above. Considerations as to how the model terplay between the two. Since many modellers use the soft-
code will be developed (whether it be a spreadsheet, special- ware that they are familiar with, it is possible (although not
ist software or a programming language) should not dominate necessarily desirable) that methods of representation and lim-
debate around the nature of the model that is required to ad- itations in the software will cause a revision to the conceptual
dress the problem situation. Conceptual modelling is about model. Continued learning during model coding and exper-
determining the right model, not how the software will be imentation may cause adjustments to the conceptual model
implemented. as the understanding of the problem situation and modelling
In saying this, it must be recognized that many simulation objectives change. Model validation activities may result in
modellers only have access to one or possibly two simulation alterations to the conceptual model in order to improve the
tools. As a result, considerations of software implementation accuracy of the model. Availability, or otherwise, of data may
will naturally enter the debate about the nature of the concep- require adjustments to the conceptual model. All this implies
tual model. This is recognized by the double arrow, signify- a great deal of iteration in the process of modelling and the
ing iteration, for the model coding process in Figure 2. What requirement to continually revise the conceptual model. This
this definition for a conceptual model aims to highlight is the iteration is illustrated by the double arrows between the stages
importance of separating as far as possible detailed model in Figure 2.

This content downloaded from 194.47.65.106 on Tue, 21 Jun 2016 12:25:32 UTC
All use subject to http://about.jstor.org/terms
284 Journal of the Operational Research Society Vol. 59, No. 3

The purpose of a conceptual model * Helps build the credibility of the model.
* Guides the development of the computer model.
In reflecting on the purpose of a conceptual model, one might
* Forms the basis for model verification and guides model
question whether it is necessary to have one at all. Indeed,
validation.
some might argue that the power of modern simulation soft-
* Guides experimentation by expressing the objectives, ex-
ware negates the need for conceptual modelling. Such soft-
perimental factors and responses.
ware enables a modeller to move straight from developing an
* Provides the basis of the model documentation.
understanding of the problem situation to creating a computer
* Can act as an aid to independent verification and validation
model.
when it is required.
Albeit that this argument appears to have some credence, it
* Helps determine the appropriateness of the model or its
ignores the fact that whatever practice a modeller might em-
parts for model reuse and distributed simulation (Pace,
ploy for developing the model code, decisions still have to be
2000b).
taken concerning the content and assumptions of the model.
Modern simulation software does not reduce this level of de-
Overall the conceptual model, if clearly expressed, provides
cision making. What the software can provide is an envi-
a means of communication between all parties in a simula-
ronment for the more rapid development of the model code,
tion study: the modeller, clients and domain experts (Pace,
enhancing the opportunities for iteration between conceptual
2002). In so doing it helps to build a consensus, or least an
modelling and model coding, and facilitating rapid prototyp- accommodation, about the nature of the model and its use.
ing. This does not negate the need for conceptual modelling,
but simply aids the process of model design. It also highlights
the point that conceptual modelling is not a one-off step, but Requirements of a conceptual model
part of a highly iterative process, particularly in relation to
In designing a conceptual model it would be useful to have
model coding.
a set of requirements in mind. These could provide a basis
Indeed, the power of modern software (and hardware) and
against which to determine whether a conceptual model is
the wider use of distributed processing may actually have
appropriate. Indeed, Pritsker (1987) says that 'modelling is a
increased the need for effective conceptual modelling. Salt
difficult process because we do not have measurable criteria
(1993) and Chwif et al (2000) both identify the problem of
for evaluating the worth of a model'. In conceptual modelling,
the increasing complexity of simulation models; a result of
it may be difficult to identify a complete set of measurable
the 'possibility' factor. People build more complex models
criteria, since the model is purely descriptive at this stage.
because the hardware and software enable them to. While this
That said, a sense of requirements, even if they are more
may have extended the utility of simulation to problems that
qualitative, would be helpful.
previously could not have been tackled, it also breads a ten-
So what are the requirements for an effective conceptual
dency to develop overly complex models. There are various
model? This question is first answered by describing four
problems associated with such models including extended de-
main requirements after which the overarching need to keep
velopment times and onerous data requirements. This trend
the model as simple as possible is discussed.
to develop ever more complex models has been particularly
Assessment criteria for models have been discussed by a
prevalent in the military domain (Lucas and McGunnigle,
number of authors, for instance, Gass and Joel (1981), Oren
2003). Indeed, it could be argued that there are some ad-
(1981, 1984), Robinson and Pidd (1998) and Balci (2001).
vantages in only having limited computing capacity; it forces
The majority of this work is in the domain of large-scale mil-
the modeller to carefully design the model! As a result of
itary and public policy models; Robinson and Pidd is an ex-
the possibility factor it would seem that careful design of the
ception. Furthermore, the criteria focus on assessing models
conceptual model is more important than ever.
that have been developed rather than on the assessment of
Beyond the general sense that careful model design is im-
conceptual models.
portant, there are a number of reasons why a conceptual model
In terms of criteria for conceptual models in operational re-
is important to the development and use of simulation models.
search there has been little reported. Willemain (1994), who
Pace (2003, www.sisostds.org, accessed February 2006) puts
investigates the preliminary stages of operational research in-
this succinctly by stating that the conceptual model provides a
terventions, briefly lists five qualities of an effective model:
roadmap from the problem situation and objectives to model
validity, usability, value to the clients, feasibility and apt-
design and software implementation. He also recognizes that
ness for the clients' problem. Meanwhile, Brooks and Tobias
the conceptual model forms an important part of the docu-
(1996a) identify eleven performance criteria for a good model.
mentation for a model. More specifically a well-documented
Requirements are also briefly discussed by Pritsker (1986),
conceptual model:
Henriksen (1988), Nance (1994), and van der Zee and van der
* Minimises the likelihood of incomplete, unclear, in- Vorst (2005). Outside of operational research there are some
consistent and wrong requirements (Borah, 2002, discussions, for instance, Teeuw and van den Berg (1997,
www.sisostds.org, accessed February 2006; Pace, 2002). http://osm7.cs.byu.edu/ER97/workshop4/, accessed February

This content downloaded from 194.47.65.106 on Tue, 21 Jun 2016 12:25:32 UTC
All use subject to http://about.jstor.org/terms
 S Robinson-Conceptual modelling for simulation 285

2006) who discuss the quality of conceptual models for busi- model validity as a perception of the modeller. It also main-
ness process reengineering. tains the notion that a model is built for a specific purpose,
Based on the discussions by simulation modellers and op- which is common to most definitions of validity.
erational researchers, here it is proposed that there are four Credibility is similar to validity, but is taken from the per-
main requirements of a conceptual model: validity, credibil- spective of the clients rather than the modeller. The credibility
ity, utility and feasibility. Table 1 shows how the requirements of the conceptual model is therefore defined as:
discussed in the literature relate to these.

It is generally agreed that a valid model is one that is A perception, on behalf of the clients, that the conceptual model
sufficiently accurate for the purpose at hand (Carson, 1986). can be developed into a computer model that is sufficiently
However, since the notion of accuracy is of little meaning accurate for the purpose at hand.
for a model that has no numeric output, conceptual model
validity might be defined as:
The clients must believe that the model is sufficiently accurate.
Included in this concept is the need for the clients to be
A perception, on behalf of the modeller, that the conceptual convinced that all the important components and relationships
model can be developed into a computer model that is suffi- are in the model. Credibility also requires that the model
ciently accurate for the purpose at hand.
and its results are understood by the clients. Would a model
that could not be understood have credibility? An important
The phrase '... can be developed into a computer model...' factor in this respect is the transparency of the model which
is included in recognition that the conceptual model is a de- is discussed below.

scription of a model, not the computer model itself. Depend- Validity and credibility are seen as separate requirements
ing on the status of the simulation project, the conceptual because the modeller and clients may have very different
model may be describing a computer model that will be de- perceptions of the same model. Although a modeller may be
veloped, is being developed, or has been developed. satisfied with a conceptual model, the clients may not be. It
Underlying the notion of validity is the question of whether is not unusual for additional scope and detail to be added to
the model is 'right'. Note that this definition places conceptual a model, not because it improves its validity, but because it

Table 1 Requirements of a conceptual model related to those documented in the literature

Documented requirements

Proposed Pritsker Henriksen Nance (1994) Willemain Brooks and van der Zee and
requirements (1986) (1988) (1994) Tobias (1996a) van der Vorst
(2005)

Validity Valid Fidelity Model Validity Model describes Completeness

correctness Aptness for behaviour of interest
Testability client's problem Accuracy of the
model's results
Probability of
containing errors
Validity
Strength of
theoretical basis
of model

Credibility Understandable Ease of understanding Transparency

Utility Extendible Execution Adaptability Value to client Portability and ease

speed Reusability Usability with which model can
Ease of Maintainability be combined with
modification others

Feasibility Timely Elegance Feasibility Time and cost to build

model
Time and cost to run
model
Time and cost to
analyse results
Hardware requirements

This content downloaded from 194.47.65.106 on Tue, 21 Jun 2016 12:25:32 UTC
All use subject to http://about.jstor.org/terms
286 Journal of the Operational Research Society Vol. 59, No. 3

improves its credibility. Not that adding scope and detail to The four requirements described above are not mutually ex-
gain credibility is necessarily a bad thing, but the modeller clusive. For instance, the modeller's and clients' perspectives
must ensure that this does not progress so far that the model on model accuracy are likely to be closely aligned, although
becomes over complex. Simulation is particularly prone to not always. An infeasible model could not generally be de-
such a drift through, for instance, the addition of non-vital scribed as a useful model, although a conceptual model that
graphics and the logic required to drive them. is infeasible could be useful for aiding problem understand-
The third concept, utility, is defined as: ing. Albeit that these concepts are related, it is still useful to
identify them as four separate requirements so a modeller can
A perception, on behalf of the modeller and the clients, that be cognisant of them when designing the conceptual model.
the conceptual model can be developed into a computer model
that is useful as an aid to decision-making within the specified
context. The overarching requirement: keep the model simple

The overarching requirement is the need to avoid the devel-

Utility is seen as a joint agreement between the modeller and
the clients about the usefulness of the model. This notion
opment of an overly complex model. In general, the aim
should be:
moves beyond the question of whether the model is suffi-
ciently accurate, to the question of whether the model is use-
to keep the model as simple as possible to meet the objectives
ful for the context of the simulation study. Utility includes is- of the simulation study (Robinson, 2004).
sues such as ease-of-use, flexibility (ie ease with which model
changes can be made), run-speed and visual display. Where There are a number of advantages with simple models (In-
the model, or a component of the model, might be used again nis and Rexstad, 1983; Ward, 1989; Salt, 1993; Chwif
on the same or another study, reusability would also be sub- et al, 2000; Lucas and McGunnigle, 2003; Thomas and
sumed within the concept of utility. The requirements for Charpentier, 2005):
utility are expressed through the general project objectives.
Within any context a range of conceptual models could * simple models can be developed faster,
be derived. The accuracy of these models would vary, but * simple models are more flexible,
some or all might be seen as sufficiently accurate and, hence, * simple models require less data,
under the definitions given above, they would be described * simple models run faster,
as valid and credible. This does not necessarily mean that * the results are easier to interpret since the structure of the
the models are useful. For instance, if a proposed model model is better understood.

is large and cumbersome, it may have limited utility due to

With more complex models these advantages are generally
reduced ease-of-use and flexibility. Indeed, a less accurate
lost. Indeed, at the centre of good modelling practice is the
(but still sufficiently accurate), more flexible model that runs
idea of resorting to simplest explanation possible. Occam's
faster may have greater utility by enabling a wider range of
razor puts this succinctly, 'plurality should not be posited
experimentation within a time-frame.
without necessity' (William of Occam) (quoted from Pidd,
Hodges (1991) provides an interesting discussion around
2003), as does Antoine de Saint-Exupery who reputedly said
model utility and suggests that a 'bad' model (one that is
that 'perfection is achieved, not when there is nothing more
not sufficiently accurate) can still be useful. He goes on
to add, but when there is nothing left to take away'.
to identify specific uses for such models. Bankes (1993)
The requirement for simple models does not negate the need
continues with this theme, discussing the idea of inaccurate
to build complex models on some occasions. Indeed, com-
models for exploratory use, while Robinson (2001) sees a
plex models are sometimes required to achieve the modelling
role for such models in facilitating learning about a problem
situation.
objectives. The requirement is to build the simplest model
possible, not simple models per se. What should be avoided,
The final requirement, feasibility, is defined as follows:
however, is the tendency to try and model every aspect of a
A perception, on behalf of the modeller and the clients, that system when a far simpler more focused model would suffice.
the conceptual model can be developed into a computer model The graph in Figure 3 illustrates the notional relationship
with the time, resource and data available. between model accuracy and complexity (Robinson, 1994).
Increasing levels of complexity (scope and level of detail)
A range of factors could make a model infeasible: it might improve the accuracy of the model, but with diminishing re-
not be possible to build the proposed model in the time avail- turns. Beyond point x there is little to be gained by adding
able, the data requirements may be too onerous, there may be to the complexity of the model. A 100% accurate model will
insufficient knowledge of the real system, and the modeller never be achieved because it is impossible to know everything
may have insufficient skill to code the model. Feasibility im- about the real system. The graph illustrates a further point.
plies that the time, resource and data are available to enable Increasing the complexity of the model too far, may lead to a
development of the computer model. less accurate model. This is because the data and information

This content downloaded from 194.47.65.106 on Tue, 21 Jun 2016 12:25:32 UTC
All use subject to http://about.jstor.org/terms
 S Robinson-Conceptual modelling for simulation 287

100% a modeller might go about determining what the conceptual

model should be in a simulation study. So what help is offered
in the simulation and modelling literature to guide modellers
in designing the conceptual model?
First, it is worth recognizing that conceptual modelling re-
quires creativity (Henriksen, 1989). Simulation modelling is
QU

both art and science (Shannon, 1975) with conceptual mod-

elling lying more at the artistic end! As Schmeiser (2001)
points out: 'While abstracting a model from the real world is
x very much an art, with many ways to err as well as to be cor-
Scope and level of detail (complexity)
rect, analysis of the model is more of a science, and therefore
easier, both to teach and to do'. The need for creativity does
Figure 3 Simulation model complexity and accuracy (based on
Robinson, 1994). not, however, excuse the need for guidelines on how to model
(Evans, 1992). Ferguson et al (1997), writing about software
development, point out that in 'most professions, competent
are not available to support such a detailed model. For in-
stance, it is unlikely that we could accurately model the exact
work requires the disciplined use of established practices. It
is not a matter of creativity versus discipline, but one of bring-
behaviour of individuals in a queue, and attempts to do so,
ing discipline to the work so creativity can happen'.
beyond very simple rules, may lead to a less accurate result.
In searching the modelling literature for advice from sim-
Ward (1989) provides a lucid account on the simplicity
of models. In doing so, he makes a useful distinction be- ulation modellers and operational researchers on how to
tween constructive simplicity and transparency. Transparency
develop models, three basic approaches can be found: prin-
is an attribute of the client (how well he/she understands the ciples of modelling, methods of simplification and modelling
frameworks.
model), while constructive simplicity is an attribute of the
model itself (the simplicity of the model). Because trans-
Principles of modelling
parency is an attribute of the client, it depends on his/her
level of knowledge and skill. A model that is transparent to Providing a set of guiding principles for modelling is one
one client may not be transparent to another. In developing approach to advising simulation modellers on how to develop
a conceptual model, the modeller must consider transparency (conceptual) models. For instance, Pidd (1999) describes six
as well as simplicity, designing the model with the particular principles of modelling:
needs of the client in mind. The need for transparency is, of
course, confounded by the presence of multiple clients (as is * model simple; think complicated,
the case in many simulation studies), all of whom must be * be parsimonious; start small and add,
satisfied with the model. These ideas closely link to the re- * divide and conquer; avoid megamodels,
quirement for credibility, as discussed above, since a model * use metaphors, analogies, and similarities,
that is not transparent is unlikely to have credibility.
* do not fall in love with data,

Having emphasized the importance of simplicity, there are * modelling may feel like muddling through.
those that warn against taking this to an extreme. Pritsker The central theme is one of aiming for simple models through
(1986) reflects on his experience of developing models of evolutionary development. Others have produced similar sets
differing complexity of the same system. He concludes that of principles (or guidelines), for instance, Morris (1967),
the simplest model is not always best because models need to Musselman (1992), Powell (1995), Pritsker (1998) and Law
be able to evolve as the requirements change. The simplest and Kelton (2000). The specific idea of evolutionary model
model is not always the easiest to embellish. Schruben and development is further explored by Nydick et al (2002).
Yticesan (1993) make a similar point, stating that simpler These principles provide some useful guidance for those
models are not always as easy to understand, code and debug. developing conceptual models. It is useful to encourage mod-
Davies et al (2003) point out that simpler models require ellers to start with small models and to gradually add scope
more extensive assumptions about how a system works and and detail. What such principles do not do, however, is to
that there is a danger in setting the system boundary (scope) guide a modeller through the conceptual modelling process.
too narrow in case an important facet is missed. When should more detail be added? When should elabora-

tion stop? There is a difference between giving some general

Guidance on conceptual modelling principles and guiding someone through a process.

Exhortations to develop simple models highlight an impor-

Methods of simplification
tant consideration in designing a conceptual model. Mod-
elling requirements provide a guide as to whether a concep- Simplification entails removing scope and detail from a model
tual model is appropriate. Neither, however, describes how or representing components more simply while maintaining

This content downloaded from 194.47.65.106 on Tue, 21 Jun 2016 12:25:32 UTC
All use subject to http://about.jstor.org/terms
288 Journal of the Operational Research Society Vol. 59, No. 3

a sufficient level of accuracy. In Zeigler's (1976) terms this pose; specification of the model's components; specification
could be described as further lumping of the lumped model. of the parameters and variables associated with the compo-
This is the opposite of the start small and add principle. nents; and specification of the relationships between the com-
There are quite a number of discussions on simplifica- ponents, parameters and variables.
tion, both in the simulation and the wider modelling context. Both Nance and Pace have devised frameworks which re-
Morris (1967) identifies some methods for simplifying mod- late primarily to the development of large-scale models in the
els: making variables into constants, eliminating variables, military domain. Nance (1994) outlines the conical method-
using linear relations, strengthening the assumptions and re- ology. This is an object oriented, hierarchical specification
strictions, and reducing randomness. Ward (1989) provides a language which develops the model definition (scope) top-
similar list of ideas for simplification. Meanwhile, Courtois down and the model specification (level of detail) bottom-up.
(1985) identifies criteria for the successful decomposition of A series of modelling steps are outlined. Balci and Nance
models in engineering and science. (1985) focus specifically on a procedure for problem formu-
For simulation modelling, Zeigler (1976) suggests four lation. Meanwhile, Nance and Arthur (2006) identify the po-
methods of simplification: dropping unimportant components tential to adopt software requirements engineering (SRE) ap-
of the model, using random variables to depict parts of the proaches for simulation model development. They also note
model, coarsening the range of variables in the model, and that there is little evidence of SRE actually being adopted by
grouping components of the model. There is an apparent con- simulation modellers.

tradiction between Morris' and Zeigler's advice in that the Pace (1999, 2000a) explores a four-stage approach to con-
former suggests reducing randomness, while the latter sug- ceptual model development, similar to that of Shannon: col-
gests increasing it by representing sections of the model with lect authoritative information on the problem domain; iden-
random variables. This difference in opinion can be recon- tify entities and processes that need to be represented; iden-
ciled by recognizing that simplification methods are sensitive tify simulation elements; and identify relationships between
to the modelling approach that is being applied. Morris is the simulation elements. He also identifies six criteria for de-

concentrating more on mathematical algorithms where the in- termining which elements to include in the conceptual model.
clusion of randomness is less convenient. Zeigler is writing These criteria focus on the correspondence between real world
about simulation specifically, where complex behaviours can items and simulation objects (Pace, 2000a, p. 8).
sometimes be reduced to a single random variable. Within our domain of interest, simulation for modelling
Yin and Zhou (1989) build upon Zeigler's ideas, discussing operations systems, there is quite limited work on conceptual
six simplification techniques and presenting a case study. modelling frameworks. Brooks and Tobias (1996b) briefly
Sevinc (1990) provides a semiautomatic procedure based on propose a framework for conceptual modelling, but go no fur-
Zeigler's ideas. Innis and Rexstad (1983) enter into a detailed ther in expanding upon the idea. Recent papers by Guru and
discussion about how an existing model might be simplified. Savory (2004) and van der Zee and van der Vorst (2005) pro-
They provide a list of 17 such methods, although they do not pose conceptual modelling frameworks in some more detail.
claim that these are exhaustive. They conclude by suggesting Guru and Savory propose a set of modelling templates (tables)
that managers should be provided with both a full and a sim- useful for modelling physical security systems. Meanwhile,
plified simulation model. There is a sense in which the Ford van der Zee and van der Vorst propose a framework for supply
example followed this approach, with one model being more chain simulation. Both are aimed at an object-oriented imple-
detailed than the other, although neither could be described as mentation of the computer-based simulation model. Mean-
a 'full' model. Robinson (1994) also lists some methods for while, Kotiadis (2006) looks to the ideas of Soft Operational
simplifying simulation models. Finally, Webster et al (1984) Research, and specifically soft systems methodology (SSM)
describe how they selected an appropriate level of detail for (Checkland, 1981), for aiding the conceptual modelling pro-
generating samples in a timber harvesting simulation model. cess. She uses SSM to help understand a complex health
Such ideas are useful for simplifying an existing (concep- care system and then derives the simulation conceptual model
tual) model, but they do not guide the modeller over how to from the SSM 'purposeful activity model'.
bring a model into existence. Model simplification acts pri-
marily as a redesign tool and not a design tool.
Conclusion

There is, in large measure, a vacuum of research in the area

Modelling frameworks
of conceptual modelling for discrete-event simulation. Albeit
A modelling framework goes beyond the idea of guiding prin- that many simulation researchers consider effective concep-
ciples and methods of model simplification by providing a tual modelling to be vital to the success of a simulation study,
specific set of steps that guide a modeller through develop- there have been few attempts to develop definitions and ap-
ment of a conceptual model. There have been some attempts proaches that are helpful to the development of conceptual
to provide such frameworks going back to Shannon (1975) models. The discussion above attempts to redress this balance
who describes four steps: specification of the model's pur- by offering a definition of a conceptual model and outlining

This content downloaded from 194.47.65.106 on Tue, 21 Jun 2016 12:25:32 UTC
All use subject to http://about.jstor.org/terms
 S Robinson--Conceptual modelling for simulation 289

the requirements for a conceptual model. The conceptual Brooks RJ and Tobias AM (1996b). A framework for choosing the
model definition is useful for providing a sense of direction best model structure in mathematical and computer modeling.
Proceedings of the 6th Annual Conference AI, Simulation, and
to simulation modellers during a simulation study. If they do
Planning in High Autonomy Systems, University of Arizona,
not know what they are heading for, how can they head for pp 53-60.
it? The requirements provide a means for determining the Carson JS (1986). Convincing users of model's validity is challenging
appropriateness of a conceptual model both during and af- aspect of modeler's job. Ind Eng 18(6): 74-85.
ter development. For researchers, the definition and require- Checkland PB (1981). Systems Thinking, Systems Practice. Wiley:
Chichester, UK.
ments provide a common foundation for further research in
Chwif L, Barretto MRP and Paul RJ (2000). On simulation model
conceptual modelling. complexity. In: Joines JA, Barton RR, Kang K and Fishwick
What the definition and requirements do not provide is PA (eds). Proceedings of the 2000 Winter Simulation Conference.
a sense of how to develop a conceptual model. Three ap- IEEE: Piscataway: NJ, pp 449-455.
proaches have been used in this respect: principles of mod- Cochran JK, Mackulak GT and Savory PA (1995). Simulation project
characteristics in industrial settings. Interfaces 25: 104-113.
elling, methods of simplification and modelling frameworks.
Courtois PJ (1985). On time and space decomposition of complex
The latter has potential to provide the most specific guid- structures. Commun ACM 28: 590-603.
ance on how to develop a conceptual model. It is also the Davies R, Roderick P and Raftery J (2003). The evaluation of disease
area that has seen the least development of the three, partic- prevention and treatment using simulation models. Eur J Opl Res
150: 53-66.
ularly in simulation for operations systems. It is to a frame-
DMSO (2005). Defence modeling and simulation office, HLA Page.
work for conceptual modelling that our attention turns next. Available online.
In the paper that follows (Robinson, 2007), a framework is Evans JR (1992). Creativity in MS/OR: Improving problem solving
described that is built upon the foundations laid out here. The through creative thinking. Interfaces 22(2): 87-91.
framework is illustrated by applying it to the Ford engine Ferguson P et al (1997). Results of applying the personal software
assembly plant model. process. Computer 5: 24-31.
Fishwick PA (1995). Simulation Model Design and Execution:
Building Digital Worlds. Prentice-Hall: Upper Saddle River, NJ.
Acknowledgements-- Some sections of this paper are based on: Gass SI and Joel LS (1981). Concepts of model confidence. Comput
Opns Res 8: 341-346.
* Robinson S (2004). Simulation: The Practice of Model Development
and Use. Wiley: Chichester, UK Guru A and Savory P (2004). A template-based conceptual modeling
* Robinson S (2004). Designing the conceptual model in simulation infrastructure for simulation of physical security systems. In:
studies. In: Brailsford SC, Oakshott L, Robinson S and Taylor SJE Ingalls RG, Rossetti MD, Smith JS and Peters BA (eds).
(eds). Proceedings of the 2004 Operational Research Society Simulation Proceedings of the 2004 Winter Simulation Conference. IEEE:
Workshop (SW04). Operational Research Society: Birmingham, UK, Piscataway: NJ, pp 866-873.
pp 259-266. Haddix F (2001). Conceptual modeling revisited: A developmental
* Robinson S (2006). Issues in conceptual modelling for simulation: Set- model approach for modeling and simulation. Proceedings of the
ting a research agenda. In: Garnett J, Brailsford S, Robinson S and
2001 Fall Simulation Interoperability Workshop. Available online.
Taylor S (eds). Proceedings of the Third Operational Research Soci-
Henriksen JO (1988). One system, several perspectives, many models.
etv Simulation Workshop (SW06). The Operational Research Society:
Birmingham, UK, pp 165-174. In: Abrams M, Haigh P and Comfort J (eds). Proceedings of
* Robinson S (2006). Conceptual modeling for simulation: Issues and the 1988 Winter Simulation Conference. IEEE: Piscataway: NJ,
research requirements. In: Perrone LE Wieland FP. Liu J, Lawson BG, pp 352-356.
Nicol DM and Fujimoto RM (eds). Proceedings of the 2006 Winter Henriksen JO (1989). Alternative modeling perspectives: Finding the
Simulation Conference. IEEE: Piscataway, NJ, pp 792-800. creative spark. In: MacNair EA, Musselman KJ and Heidelberger
P (eds). Proceedings of the 1989 Winter Simulation Conference.
The Ford engine plant example is used with the permission of John
IEE: Piscataway: NJ, pp 648-652.
Ladbrook, Ford Motor Company.
Hodges JS (1991). Six (or so) things you can do with a bad model.
Opns Res 39: 355-365.
References Innis G and Rexstad E (1983). Simulation model simplification
techniques. Simulation 41(1): 7-15.
Balci O (1994). Validation, verification, and testing techniques Kotiadis K (2006). Extracting a conceptual model for a complex
throughout the life cycle of a simulation study. Ann Opns Res 53: integrated system in healthcare. In: Garnett J, Brailsford S,
121-173. Robinson S and Taylor S (eds). Proceedings of the Third
Balci O (2001). A methodology for certification of modeling and Operational Research Society Simulation Workshop (SW06). The
simulation applications. ACM Trans Model Comput Simul 11: Operational Research Society: Birmingham: UK, pp 235-245.
352-377. Lacy LW et al (2001). Developing a consensus perspective on
Balci O and Nance RE (1985). Formulated problem verification as an conceptual models for simulation systems. Proceedings of the 2001
explicit requirement of model credibility. Simulation 45(2): 76-86. Spring Simulation Interoperabilitf Workshop. Available online.
Bankes S (1993). Exploratory modeling for policy analysis. Opns Res Law AM (1991). Simulation model's level of detail determines
41: 435-449. effectiveness. Ind Eng 23(10): 16-18.
Borah JJ (2002). Conceptual modeling-The missing link of Law AM and Kelton WD (2000). Simulation Modeling and Analysis,
simulation development. Proceedings of the 2002 Spring Simulation 3rd edition. McGraw-Hill: New York.

Interoperability Workshop. Available online. Lucas TW and McGunnigle JE (2003). When is model complexity
Brooks RJ and Tobias AM (1996a). Choosing the best model: Level too much? Illustrating the benefits of simple models with Hughes'
of detail, complexity and model performance. Math Comput Model salvo equations. Naval Res Logist 50: 197-217.
24(4): 1-14. Morris WT (1967). On the art of modeling. Mngt Sci 13: B707-B717.

This content downloaded from 194.47.65.106 on Tue, 21 Jun 2016 12:25:32 UTC
All use subject to http://about.jstor.org/terms
290 Journal of the Operational Research Society Vol. 59, No. 3

Musselman KJ (1992). Conducting a successful simulation project. Robinson S (2004). Simulation: The Practice of Model Development
In: Swain JJ, Goldsman D, Crain RC and Wilson JR (eds). and Use. Wiley: Chichester, UK.
Proceedings of the 1992 Winter Simulation Conference. IEEE: Robinson S (2005). Distributed simulation and simulation practice.
Piscataway: NJ, pp 115-121. Simulation: Trans Soc Model Comput Simul 81: 5-13.
Nance RE (1994). The conical methodology and the evolution of Robinson S (2007). Conceptual modelling for simulation Part
simulation model development. Ann Opns Res 53: 1-45. II: A framework for conceptual modelling. J Opl Res Soc,
Nance RE and Arthur JD (2006). Software requirements engineering: doi: 10.1057/palgrave.jors.2602369.
Exploring the role in simulation model development. In: Garnett Robinson S and Pidd M (1998). Provider and customer expectations
J, Brailsford S, Robinson S and Taylor S (eds). Proceedings of the of successful simulation projects. J Opl Res Soc 49: 200-209.
Third Operational Research Society Simulation Workshop (SW06). Salt J (1993). Simulation should be easy and fun. In: Evans GW,
The Operational Research Society: Birmingham: UK, pp 117-127. Mollaghasemi M, Russell EC and Biles WE (eds). Proceedings
Nydick RL, Liberatore MJ and Chung QB (2002). Modeling by of the 1993 Winter Simulation Conference. IEEE: Piscataway: NJ,
elaboration: An application to visual process simulation. INFOR pp 1-5.
40: 347-361. Schmeiser BW (2001). Some myths and common errors in simulation
Oren TI (1981). Concepts and criteria to assess acceptability of experiments. In: Peters BA, Smith JS, Medeiros DJ and Rohrer
simulation studies: A frame of reference. Commun ACM 28: MW (eds). Proceedings of the 2001 Winter Simulation Conference.
190-201. IEEE: Piscataway: NJ, pp 39-46.
Oren TI (1984). Quality assurance in modeling and simulation: Schruben L and Yiicesan E (1993). Complexity of simulation
A taxonomy. In: Oren TI, Zeigler BP and Elzas MS (eds). models: A graph theoretic approach. In: Evans GW, Mollaghasemi
Simulation and Model-based Methodologies: An Integrative M, Russell EC and Biles WE (eds). Proceedings of the
Approach. Springer-Verlag: Heidelberg: Germany, pp 477-517. 1993 Winter Simulation Conference. IEEE: Piscataway: NJ,
Pace DK (1999). Development and documentation of a simulation pp 641-649.
conceptual model. Proceedings of the 1999 Fall Simulation Sevinc S (1990). Automation of simplification in discrete event
Interoperability Workshop. Available online. modeling and simulation. Int J Gen Syst 18: 125-142.
Pace DK (2000a). Simulation conceptual model development. Shannon RE (1975). Systems simulation: The Art and Science.
Proceedings of the 2000 Spring Simulation Interoperability Prentice-Hall: Englewood Cliffs, NJ.
Workshop. Available online. Teeuw WB and van den Berg H (1997). On the quality of conceptual
Pace DK (2000b). Ideas about simulation conceptual model models. In: Liddle SW (ed). Proceedings of the ER'97 Workshop
development. Johns Hopkins APL Tech Dig 21: 327-336. on Behavioral Models and Design Transformations: Issues and
Pace DK (2002). The value of a quality simulation conceptual model. Opportunities in Conceptual Modeling. Available online.
Model Simul Mag 1(1): 9-10. Thomas A and Charpentier P (2005). Reducing simulation models
Pace DK (2003). Thoughts about the simulation conceptual model. for scheduling manufacturing facilities. Eur J Opl Res 161:
Proceedings of the 2003 Spring Simulation Interoperability 111-125.
Workshop. Available online. van der Zee DJ and van der Vorst JGAJ (2005). A modeling framework
Pidd M (1999). Just modeling through: A rough guide to modeling. for supply chain simulation: Opportunities for improved decision
Interfaces 29: 118-132. making. Decis Sci 36: 65-95.
Pidd M (2003). Tools for Thinking: Modelling in Management Ward SC (1989). Arguments for constructively simple models. J Opl
Science, 2nd edition. Wiley: Chichester, UK. Res Soc 40: 141-153.
Pidd M (2004). Computer Simulation in Management Science, 5th Webster DB, Padgett ML, Hines GS and Sirois DL (1984).
edition. Wiley: Chichester, UK. Determining the level of detail in a simulation model-A case
Powell SG (1995). Six key modeling heuristics. Interfaces 25: study. Comput Ind Eng 8: 215-225.
114-125. Wild R (2002). Operations Management, 6th edition. Continuum:
Pritsker AAB (1986). Model evolution: A rotary table case history. London.

In: Wilson J, Henriksen J and Roberts S (eds). Proceedings of Willemain TR (1994). Insights on modeling from a dozen experts.
the 1986 Winter Simulation Conference. IEEE: Piscataway: NJ, Opns Res 42: 213-222.
pp 703-707. Willemain TR (1995). Model formulation: What experts think about
Pritsker AAB (1987). Model evolution II: An FMS design problem. and when. Opns Res 43: 916-932.
In: Thesen A, Grant H and Kelton WD (eds). Proceedings of Yin HY and Zhou ZN (1989). Simplification techniques of simulation
the 1987 Winter Simulation Conference. IEEE: Piscataway: NJ, models. Proceedings of Beijing International Conference on System
pp 567-574. Simulation and Scientific Computing, IEEE: Piscataway: NJ,
Pritsker AAB (1998). Principles of simulation modeling. In: Banks J pp 782-786.
(ed). Handbook of Simulation. Wiley: New York, pp 31-51. Zeigler BP (1976). Theory of Modeling and Simulation. Wiley:
Robinson S (1994). Simulation projects: Building the right conceptual New York.

model. Ind Eng 26(9): 34-36.

Robinson S (2001). Soft with a hard centre: Discrete-event simulation
in facilitation. J Opl Res Soc 52: 905-915.
Robinson S (2002). Modes of simulation practice: Approaches to
Received April 2006;
business and military simulation. Simul Pract Theory 10: 513-523.
accepted November 2006 after one revision

This content downloaded from 194.47.65.106 on Tue, 21 Jun 2016 12:25:32 UTC
All use subject to http://about.jstor.org/terms