Towards Simulation of Business Processes
Transforming BPMN Models to Enterprise Dynamics Models
Ralf Schepers, Tobias Minning, Yannik Moog and Ingo J. Timm
Business Informatics I, University of Trier, Trier, Germany
Keywords:
Business Process Simulation, Material Flow Simulation, Enterprise Dynamics (ED), BPMN.
Abstract:
Due to the ISO 9001 certification and process oriented-organization, many business process models are avail-
able at enterprises and public institutions. They are e.g. used for documentation or for introducing processes
to new employees. As a de-facto standard notation BPMN “Business Process Model and Notation” is widely
used. These process models can support static analyses of business processes. Dynamic analysis, e.g., by
simulation is beneficial for in-depth analysis and optimization. However, only few approaches are available to
perform simulation on basis of BPMN. In the production engineering domain, process simulation for analysis
and optimization is a de-facto standard. In this domain, material flow simulation is a valid method of analysis,
planning and construction.
In this paper, we discuss the potentials and shortcomings of transforming BPMN models to a material flow
simulation model. On the basis of an analysis of BPMN and material flow simulation, we identified require-
ments for transforming. Four levels of transformation complexity are defined. Furthermore we developed
matching relation from BPMN elements to material flow elements. As proof of concept, we implemented the
transformation process using Enterprise Dynamics and evaluated its outcome. The benefits and limitations of
this approach are discussed in the paper in front of evaluation and related work.
1 INTRODUCTION
The ISO 9001 certification represents an authorita-
tive reference of quality management. Since 2009,
Switzerland’s public institutions have to be complied
with ISO 9001(eCH eGov, 2013). One requirement
of this certification is the documentation of business
processes. BPMN 2.0 (Business Process Model and
Notation) is a common standard for process nota-
tion(Freund and Ruecker, 2012), which has been cho-
sen by Switzerland. In consequence, for each cer-
tified institution (in Switzerland), the business pro-
cesses are available in BPMN. Those are used, e.g.,
for documentation and analysis matters. The model-
ing and maintenance of business process documents is
resource intensive. So an additional value should be
added. On basis of the process models, static analy-
sis are enabled, however, as underlying research ques-
tion of this paper, the question occurs, if it is possible
to derive a standard process for dynamic analysis of
business processes.
There exists already the approach to simulate busi-
ness processes by business process simulation. But it
lacks on a low leeway in decision-making (Shannon,
1998) and missing standards(Januszczak and Hook,
2011). To re-use existing models in simulation, a
transformation approach into executable models is
necessary. To improve support of decision-making,
the method of material flow simulation will be used,
because simulation is a well-defined process method
for analysis in engineering. Due to the relevance of
BPMN standards and the possibility of the conversion
within the standards, the transformation from BPMN
models to executable simulation models will be an-
alyzed. As a first step, our objective is to study the
transfer of BPMN into an existing simulation environ-
ment. This will be evaluated by the implementation of
the transformation function in a simulation system i.e.
Incontrol Enterprise Dynamics.
2 FUNDAMENTALS
Due to our research objective, BPMN and the mean
of material flow simulation will be introduced in the
following sections.
159
Schepers R., Minning T., Moog Y. and Timm I.
Towards Simulation of Business ProcessesTransforming BPMN Models to Enterprise Dynamics Models.
DOI: 10.5220/0005425501590165
In Proceedings of the Fourth International Symposium on Business Modeling and Software Design (BMSD 2014), pages 159-165
ISBN: 978-989-758-032-1
Copyright
c
 2014 by SCITEPRESS – Science and Technology Publications, Lda. All rights reserved

Administration
Ramp
Warehouse
Good s arrivin g Enough storage?
Refuse
delivery
accept
goods
Inventory updated
Goods stored
Good s refused
Figure 1: Example BPMN process.
2.1 BPMN
The BPMN 2.0 standard is defined by the OMG
Object Management Group(Chinosi and Trombetta,
2012). It includes seven groups of elements. There
are activities, gateways, events, swim lanes, data,
choreographies and conversation(Chinosi and Trom-
betta, 2012). Thus, the BPMN elements can be aggre-
gated to a sequence flow. An additional layer can be
constructed by message links between the elements.
An example of business process can be found in fig-
ure 1 and will be used as illustration within this paper.
Although syntax of BPMN is clearly defined, se-
mantics is not. The underlying logic can be annotated
or it can implicit. However, the logic have to be ex-
tractable. In our example (figure 1) the gateway is an-
notated with ”enough storage”. Alternatively, it could
be annotated with ”insufficient storage”. Also the def-
inition is implicit, which storage is meant. Since 2008
the OMG tries to solve this problem by the defini-
tion of ”Semantics of Business Vocabulary and Busi-
ness Rules” (SBVR)(OMG, 2008). This provides the
ability to map the context to a machine-interpretable
operands. It is also used to derive processes from tex-
tual descriptions(Bajwa et al., 2011). However the
problem of implicitly can only be solved by improv-
ing the modeling process.
Beside the semantics, the general transformation
and simulation potential of BPMN must be analyzed.
For this purpose, the technical basis of transformation
will be considered. In order to recondition BPMN
diagram information in a machine-interpretable way,
the OMG defines the ”BPMN Diagram Interchange”
schema. BPMN DI is part of the BPMN 2.0 standard.
The DI scheme includes all structural information of
the model. Meta-information, such as the element
type, its name or edge type, and the execution seman-
tics are included in the ”BPMN Execution Semantics
Schema” (BPMN-ES)(OMG, 2011). A BPMN model
consists of the Diagram Interchange and the Execu-
tion Semantics Information. It can be stored by a
XML file. This forms a reasonable technical basis,
which includes the information needed to transfer and
display BPMN processes between different modeling
platforms.
2.2 Material Flow Simulation
In a simulation, a model of a real world system,
will be run under experimental conditions (Shannon,
1998). The objective is to understand the behavior
of the system to find improvements, which can be
evaluated through experiments (Shannon, 1998). A
simulation project consists of analysis, modeling, ex-
periment and interpretation of results(Lattner et al.,
2011). Each simulation run is performed several
time due to its probabilistic parameters (Lattner et al.,
2011). Material flow simulation is one kind of simu-
lation with focus on production and logistic. An ab-
stract view on material flow simulation is given by
(Rittgen, 1998). A process consists of elements which
are atomic in nature. The process is define by its ele-
ments. Each element can perform an action, which is
hidden, only the effect of a process can be seen. But
this view is too abstract to be used as a transformation
basis.
The mean of material flow simulation is based on
experience and efforts of engineers, who are the main
user of those systems. Different commercial simula-
tion software, e.g. “Plant Simulation” by Siemens
1
,
“Enterprise Dynamics” (ED) by Incontrol Simulation
Solution
2
, and “Arena” by Rockwell Automation
3
are
available. There notations are based on graph models,
however different elements are obtainable. As an ex-
ample of different elements, Arena offers a “decide”
module comparable to a BPMN gateway. This func-
tion has to be implemented in ED and in Plant Simu-
lation due to elements properties or adapted elements.
As a representative of a material flow simulation soft-
ware, we will further use ED. It will be exemplarily
introduced.
ED offers an atomic library of elements. Each el-
ement is represented by an instance of an atom, with
its own variables and logic. So, ED combines both an
object orientation and an event implementation. As a
result, the atoms can be defined, reused and derived
(object-orientation). The logic, however, is imple-
mented inside the instance of an atom (event-based).
The tokens (e.g. products or orders) are created by
a source and leave the process via a sink. The status
of the atoms and global variables, like the set of to-
kens within the model or the content of a queue, are
1
www.plant-simulation.de
2
www.incontrolsim.com
3
www.arenasimulation.com
Fourth International Symposium on Business Modeling and Software Design
160

representing the state of the model. After creation,
the tokens are then processed according to the model
pattern.
2.3 Discussion
From a methodological point of view, business pro-
cesses are similar to material flow models witch re-
spect to its graph structure as well as its atomic el-
ements. However its perspective can be different.
Business processes, which are modelled for docu-
mentation matters, should map all relevant elements
of its business objectives. In contrast a material flow
model, is a problem-oriented homomorphous map-
ping of a real world system. This could be congruent
to business processes, but it is not compulsorily. In
case of the usage of an existing BPMN model, the fo-
cus of the model possibly have to changed in order to
meet the underlying question of the analysis.
3 REQUIREMENTS ON
SIMULATION SOFTWARE
The simulation environment must be defined, before
raises the possibility of a transformation, the overall
requirements to.
1. Availability of basic concepts of nodes and edges
2. Decision functions within nodes
3. Model import via XML format
4. Graphical representation of BPMN model
5. Interface for input/ output data
6. Validated random generator
Starting with the sixth requirement, a validated
random generator is necessary to receive statistically
valid results(L’Ecuyer, 1997). Random values are
used as an abstraction of variance of the real world
to a model. A validated random generator should be
included in a commercial and established simulation
software. The basis of an analysis in simulation lies
in the analysis of data. However, it is not a typical
job to a business process modeler. In contrast, data
handling, is a core task of the method of material flow
simulation. Well studied solutions for data analysis
are available, e.g. by (Bogon et al., 2012).
In the case of transforming existing models into a
material flow simulation software, a graphical repre-
sentation, close to BPMN would reduce training. The
graphical representation leads to an increased accep-
tance of the method of simulation, because the partic-
ipants understand the model. This is a crucial point
in simulation projects(Wenzel et al., 2007). To per-
form simulation experiments, the gateways have to be
negotiable. So there is a need of decision functions
inside the available nodes. Close related to the graph-
ical representation, is the ability of the model import.
This is not yet possible to the above mentioned soft-
ware solutions.
The first four requirements could be covered by
common BPMN simulators such as Signavio
4
. How-
ever, the last requirements, necessitate the use of more
extensive simulation environments, such as material
flow simulation software solutions. The use of mate-
rial flow simulations demands a high variance in mod-
eling, due to corresponding production. Therefore it
provides expandability of the used elements. Manu-
facturing processes models follows similar to BPMN
models, the pattern of change between puffer, pro-
cessing and transportation.
4 CONCEPT
At a high level of abstraction, the generation of a
transfer process of business process to material flow
simulation model, can be seen in figure 2. At the be-
ginning the business model has to be readable, e.g.
via a XML file. Subsequently transfer rules from
a meta-model of business process or specific a meta
model (e.g. BPMN meta-model) to a meta-model of
material flow simulation have to be defined
5
. After-
wards, the rules can be deployed on the business pro-
cess model. As a proof of concept, whether a trans-
import of business
model
definition of
transformation
rules
use of
transformation
rules
(domain specific)
meta-model
business process
(domain specific)
meta-model
material-flow simulation
Figure 2: Generation of a transformation process.
formation is possible, direct transformation rules will
be specified in the following. The abstraction steps
will be avoided.
4.1 Matching Concept
As mentioned in section 2.2, there are differences be-
4
www.signavio.com
5
So far no meta-model of material flow simulation is
known to the authors.
Towards Simulation of Business Processes - Transforming BPMN Models to Enterprise Dynamics Models
161

Table 1: ED atom overview.
Atom Function Input / Output Attribute
source source of product product / n inter-arrival time
sink leaving of product 1 / - -
server cycle time n / n cycle time,in- & output strategy
splitter sequence flow split 1 / n cycle time & quantity matrix
assembler sequence flow merge n / 1 cycle time & quantity matrix
composition container element group - / - -
tween available simulation solutions, and there ex-
ists no consensus of notation. To prove the trans-
ferability, a direct transformation to ED will be dis-
cussed. In the following, destination elements in ED
will be searched, to map BPMN elements. Thus stan-
dard properties and behaviors according to the BPMN
standards could be defined in advance. Alternatively,
the properties of existing ED elements can be used.
To model the production process flow-logic and
the presentation of the processing activities, ED pro-
vides six basic elements (atoms): source, sink, server,
assembler, transport, and composition container. The
products are distributed via directed channels, con-
nected by in- and output channels. An overview of
atoms, their functions, and their properties can be
seen in table 1. The processing is provided by the
server, the assembler and the splitting atoms. Here,
the choice of the output channel (send to), as well as
the processing time can be fixed by stochastic distri-
bution or by an absolute term. An example of process-
ing atoms, are packing and unpacking activities. The
split atoms or the assembler atoms are able to split
or to join the process flow. A splitter atom can be
entered via one input channel and n output channels.
Vice versa to the assembler atom. The determined in-
and output ratio of products can be set by a table. Also
a processing time can be set.
For visual grouping of the atoms, ED uses a com-
position container atom. It could be also used for a
logical grouping. The atom provides a rectangle that
is drawn around atoms. According to the association
of data objects, ED uses an ActiveX or a database con-
nection. Thus, the simulation can be based for exam-
ple on an Excel spreadsheet, containing input data. Of
course data elements in the BPMN perspective, like
standard forms are not needed for simulation.
The prescribed ED atomic structure potentially of-
fers both the modeling options to map the BPMN flow
elements, as well as the possibility to represent the
process. A mapping of BPMN to ED elements can be
seen in table 2. On this basis a transformation possi-
bility can be derived to infer the suitability of ED as
a target simulation environment. In addition, a logic
detection is needed. In the following, four different
levels of complexity of transformation can be differ-
entiated.
4.2 1
st
Degree: Straight Mapping
The straight mapping includes presentation of the ba-
sic elements, shown in table 2. A 1:1 transfer of ele-
ments to ED takes place. Simple tasks can be mapped
to server atom with appropriate properties. From the
group of gateways, the parallel and the exclusive gate-
ways can be transferred. The parallel gateway can
be matched to the splitter atom. The process flow
can be merged by an assembler atom. In case of an
exclusive gateway, a server atom with specific “sent
to” ratio will be set. This can be set by conditions,
as well as a percent ratio. The merging gateway can
be set to a server atom. Such takes the product from
any input channel and supplies it to one output chan-
nel. Informations about average processing time of
tasks is not given by a BPMN model. Consequently
it would have to be inserted while the transformation
process. To get plausible simulation results, data have
to be collected. By default, this follows a negative
Table 2: Mapping complexity degrees.
BPMN Enterprise Dynamics
1
st
degree
start-event source
end-event sink
timer-events time based release
parallel gateway splitter/ assembler
exclusive gateway output strategy server
sequence flow relations
task server
(swim) lane composition atom
2
nd
degree
start message event adapted source
message event adapted atom
message task adapted token
sub process composition atom
3
rd
degree
looping task adapted server
event based gateway adapted server
Fourth International Symposium on Business Modeling and Software Design
162

exponential function, set by ED.
4.3 2
nd
Degree: Local Logic Detection
The 2
nd
complexity degree requires to implement the
rules based on the local recognition of logic. It is
based on the defined elements to represent the atoms.
In this case, these rules primarily support message
events. This requires the adaption of ED atoms, be-
cause communication actions are not a part of the ma-
terial flow simulation method yet. There are possibil-
ity to send messages like parameters, but a delibera-
tion is not yet possible. We did some work, to imple-
ment communication actions according to FIPA stan-
dard(Bellifemine et al., 1999).
4.4 3
rd
Degree: Global Recognition
In order to map more complex control structures like
anonymous processes and pools, a composition of ED
atoms is required. Also sub processes and black boxes
can be represented. For this purpose, incoming and
outgoing messages as well as sequence flows of con-
tainers have to be identified and grouped. They could
lead to a source or a sink atom in a composition con-
tainer.
For this level of complexity, pattern recognition is
necessary. It should check the kind of parent and son
element. The event- based gateways can be viewed
in a global context with their downstream events.
The event gateway can be mapped to a server with
its own processing strategy. This can be influenced
by specifying a script in the ED scripting language
4DSCRIPT. The in- and output channels of the group
are then aggregated into a server atom. Likewise, the
mapping of loop tasks can be classified under this
level of complexity.
4.5 4
th
Degree: SBVR Logic and
Complex Mapping
The 4
th
level includes the mapping of the remain-
ing elements. Although these affect the process flow,
the underlying logic implemented is difficult to rec-
ognize, and consequently more difficult for a ma-
chine due to lack of semantic interpretability. The im-
plementation requires an extended semantic context,
such as it is used for example in the SBVR research.
To implement the logic correctly, additional recogni-
tion of complex patterns as well as dynamic transfer
rules will be necessary.
5 EVALUATION
Starting to evaluate this approach, a transformation
application based on 1
st
degree, has been imple-
mented. The transformation fulfills the requirements
three, by acting as an interface between a BPMN
model and ED. It includes importing of the BPMN
XML documents, as well as conversion and storage in
the file format used by ED. Due to the application, ED
is able to fully implement all transferable elements.
As shown in figure 1, the application were able to
transform all elements. The atom properties, like cy-
cle time were predefined. In comparison to figure 1,
which were used as origin, requirements one (avail-
ability of basic concepts of nodes and edges) as well
as the “graphical representation of BPMN model” can
be proven. Due to the predefined parameters, output
data could be generated by ED (requirements five)
but there were no further transformation as 1
st
de-
gree complexity. In addition, no data collections were
made to set parameters. Those were set random. As
a result, generated data are not valid. For each con-
trol structure of the 1
st
and 2
nd
complexity degrees a
mapping could be found. In addition, a corresponding
control structure has been developed to map the 3
rd
level of complexity. For a transformation of higher
degree than 1
st
degree, a pattern and logic recognition
is necessary. However, the author belief, that the in-
formation base should be adequate up to 3
rd
degree,
but a composition of atoms is required. The mapping
of further elements and control structures of the 4
th
degree of complexity underlie the mentioned prob-
lems. They cannot be transpose without an extension
of the information base. The requirement for a correct
representation of the process flow logic is given only
for the first three levels of complexity.
6 RELATED WORK
An approach to transform BPMN to a simulation en-
vironment was proposed by (Cetinkaya et al., 2012).
They developed an executable meta-model of the
BPMN standard. But they weren’t able to establish an
executable model. In addition, no reproduce of the di-
agrams was possible. Another possibility to simulate
business processes are proposed by (Mueller, 2012).
They are proposing to use EPC (event-driven process
chain) notation to model simulation models. Those
models were then compiled to an executable program.
This forms an advantage in contrast of our approach,
because no standard notation or software are used to
run the process model. However shortcoming were
identified due to EPC extension for simulation like
Towards Simulation of Business Processes - Transforming BPMN Models to Enterprise Dynamics Models
163

Figure 3: Process in ED: 1 swim lane, 2 start-event, 3 parallel gateway, 4 task, 5 exclusive gateway, 6 end-event.
data. Also simulation experts should be necessary. In
case the use of our approach, also simulation specific
data are needed. The advantage of our approach lies
in the reuse of existing models as well as the use of
material flow simulation software. In contrast to gen-
erated code, adjustment can be done with respect to
modeling and configuration actions. This can be done
and understood by BPMN modeler too after training.
(Januszczak and Hook, 2011) discuss a simulation
standard for business process managements in gen-
eral. In this approach, also the usage of a standard no-
tation (e.g. BPMN) is proposed to reduce the needed
training for simulation. The discussion remained at
a meta-level and proposed no transformation but an
extension of business process modeling notations. In
consequence, BPMS (business process management
suite) can be used for simulation. In our approach,
the use of material flow simulation software is pro-
posed, to use existing method(s), as well as existing
software solutions.
Another approach is represented by (Dijkman
et al., 2008). They map BPMN to petri net. In dif-
ference to material flow simulation and BPMN, petri
net have got a well-studied formal language as well as
“efficient static analysis techniques”(Dijkman et al.,
2008). As already mentioned, semantics are identi-
fied as a challenge by (Dijkman et al., 2008). A sim-
ilar approach is presented by (Raedts et al., 2007). A
lot of BPMN primitives are successfully mapped into
a petri net. The subjacent question was to be able
to find specification inconsistency located within the
BPMN diagrams like deadlocks or loops.
In summary, the tendency to generate an addi-
tional advantage of the models gets obvious. Also the
ability to convert between the notation and methods
are met. In case of transformation, the problem of
machine-interpretability demand of an ontology for
business process(Cabral et al., 2009).
7 CONCLUSION
BPMN as a modeling language is well established in
business. One benefit of this approach is, to improve
the communication between the principal and agent
due to a joint basis of communication in a simulation
project. Thus, as an additional application of our ap-
proach, BPMN models could be specified in work-
shops between customers and service providers of
simulation technology. Doing so, a first scratch, i.e.,
a first simulation model could be derived from these
models. As a missing link, semantics, especially the
quantitative information of the models, are required
for sophisticated simulation. Here more research is
needed to identify the best point of time, when to ac-
quire these information in the simulation process. De-
pending on this result the acquisition should extend
the BPMN or the material flow model.
As there is an extensive usage of BPMN models in
companies, adding additional value to these process
models is of high relevance. In this paper we ana-
lyzed, how BPMN process can be transferred into ED
material flow simulation to enable a dynamic analysis
of the underlying process structure. The feasibility
of transferring a BPMN process model into a simu-
lation model of commercial material flow simulation
software has been shown at a defined (low) level. At
higher level, the relevant tasks have been discussed.
As a result of the analysis, there is a high coverage of
concepts in BPMN and in material flow simulation.
We proposed a general transformation process
(see fig. 2), by which a transformation will be run
through a meta-level. By the employment of BPMN
meta-model or the Business Process Definition Meta-
Model, a transformation to a meta-model of simula-
tion is needed. In contrast, the presented approach
discuss a direct transformation. Based on section
2.2 argumentation, that different simulation software
Fourth International Symposium on Business Modeling and Software Design
164

have got different natured, but similar notations, the
presented approach is limited to following aspects:
1. static specific rules (BPMN to ED)
2. portability of approach to other material flow sim-
ulation is not ensured
3. portability of approach to other business process
modeling notations like EPC is not ensured
The negotiability to other material flow simula-
tion software systems as well as other modeling no-
tations is disputable. Difference between modeling
methods (e.g. focus) of other software and notations,
lead to different used elements. As a result, there is
no standardized meta-model of material flow simu-
lation. However, relevant tasks, e.g. logic detection
remain relevant to other modeling notations as wells
as to other material flow simulation software. Further
research steps will include the analysis of common
material flow simulation software. In comparison to
BPMN notation, an ontology based approach will be
developed. Goal is the interoperability of BPMN to
material flow simulation software.
REFERENCES
Bajwa, I. S., Lee, M. G., and Bordbar, B. (2011). Sbvr
business rules generation from natural language spec-
ification. In AAAI Spring Symposium: AI for Business
Agility. AAAI.
Bellifemine, F., Poggi, A., and Rimassa, G. (1999). JADE
- A FIPA-compliant agent framework, pages 97–108.
The Practical Application Company Ltd.
Bogon, T., Jessen, U., Lattner, A. D., Paraskevopoulos, D.,
Schmitz, M., Spieckermann, S., Timm, I. J., and Wen-
zel, S. (2012). Towards assisted input and output data
analysis in manufacturing simulation: The EDASim
approach. In Laroque, C., Himmelspach, J., Pasupa-
thy, R., Rose, O., and Uhrmacher, A. M., editors, Win-
ter Simulation Conference (WSC 2012).
Cabral, L., Norton, B., and Domingue, J. (2009). The busi-
ness process modelling ontology. In Proceedings of
the 4th International Workshop on Semantic Business
Process Management, SBPM ’09, pages 9–16, New
York, NY, USA. ACM.
Cetinkaya, D., Verbraeck, A., and Seck, M. D. (2012).
Model transformation from bpmn to devs in the
mdd4ms framework. In Wainer, G. A. and Moster-
man, P. J., editors, SpringSim (TMS-DEVS), page 28.
SCS/ACM.
Chinosi, M. and Trombetta, A. (2012). Bpmn: An intro-
duction to the standard. Computer Standards & Inter-
faces, 34(1):124–134.
Dijkman, R. M., Dumas, M., and Ouyang, C. (2008). Se-
mantics and analysis of business process models in
bpmn. Inf. Softw. Technol., 50(12):1281–1294.
eCH eGov (2013). ech-0073 dokumentation.
Freund, J. and Ruecker, B. (2012). Praxishandbuch BPMN
2.0. Hanser, Mnchen; Wien, dritte edition.
Januszczak, J. and Hook, G. (2011). Simulation standard
for business process management. In Simulation Con-
ference (WSC), Proceedings of the 2011 Winter, pages
741–751.
Lattner, A. D., Pitsch, H., Timm, I. J., Spieckermann, S.,
and Wenzel, S. (2011). Assistsim–towards automation
of simulation studies in logistics. SNE, page 119.
L’Ecuyer, P. (1997). Uniform random number generators:
A review. In Proceedings of the 29th Conference on
Winter Simulation, WSC ’97, pages 127–134, Wash-
ington, DC, USA. IEEE Computer Society.
Mueller, C. (2012). Generation of epc based simulation
models. In Troitzsch, K. G., M
¨
ohring, M., and Lotz-
mann, U., editors, ECMS, pages 301–305. European
Council for Modeling and Simulation.
OMG (2008). Semantics of business vocabulary and busi-
ness rules (sbvr), v1.0. Technical report, Object Man-
agement Group.
OMG (2011). Business Process Model and Notation
(BPMN), Version 2.0.
Raedts, I., Petkovic, M., Usenko, Y. S., van der Werf, J. M.
E. M., Groote, J. F., and Somers, L. J. (2007). Trans-
formation of bpmn models for behaviour analysis. In
Augusto, J. C., Barjis, J., and Ultes-Nitsche, U., edi-
tors, MSVVEIS, pages 126–137. INSTICC PRESS.
Rittgen, P. (1998). Prozeßtheorie der Ablaufplanung.: Alge-
braische Modellierung von Prozessen, Ressourcenre-
striktionen und Zeit. Teubner-Reihe Wirtschaftsinfor-
matik. Teubner B.G. GmbH.
Shannon, R. E. (1998). Introduction to the art and science of
simulation. In Proceedings of the 30th Conference on
Winter Simulation, WSC ’98, pages 7–14, Los Alami-
tos, CA, USA. IEEE Computer Society Press.
Wenzel, S., Collisi-B
¨
ohmer, S., Weiß, M., Rose, O., and
Pitsch, H. (2007). Qualit
¨
atskriterien f
¨
ur die Sim-
ulation in Produktion und Logistik: Planung und
Durchf
¨
uhrung von Simulationsstudien. (VDI-Buch)
Chemische Technik Verfahrenstechnik. Springer.
Towards Simulation of Business Processes - Transforming BPMN Models to Enterprise Dynamics Models
165