Information Technology Reference
In-Depth Information
connection between activities can be represented by
conditional relations
with
specific characteristics. Depending on certain conditions or probabilities one or the
other path is used. The model itself is marked as a
ServiceAspect
and the elements
representing characteristics are marked as
ServiceDescriptionElements
. So these
marked elements represent the connection to other models [31].
Now, the question arises of where to get the information for building a simulation
model. As already mentioned, modeling business processes including different partial
logistics services using BPMN is part of the 4PL's planning process. These process
models can be regarded as given in a service repository [31]. So we can use the
structure (start, end, tasks, relations, gateways) of the process to derive the structure
(source, sink, activities and relations) for the simulation model. Gateways, for
example, in process models are represented by conditional relations. Further
information is available in service profiles (see Section 6). So these profiles contain
the specific information required to characterize the activities in the simulation model,
e.g. time, quality or capacities. The method of collecting this information is described
in Section 5 in more detail.
To combine the process model with the provider information, the process editor is
extended by a provider selection. Based on the process model for each partial logistics
services represented as BPMN tasks a suitable provider and the required information
are selected. With this information and the underlying simulation metamodel (see Fig.
6), a 4PL can automatically generate a simulation model for a specific simulation tool.
This requires that for each simulation tool transformation rules have to be defined
only once. This approach enables a 4PL to make use of the advantages of simulation
for securing the planning process and to improve decision-making without any special
training and special experience in the creation of simulation models. Simulation
models can be created in an easy and efficient way and the effort for comparing a set
of different logistics network configurations is reduced. Furthermore, the simulation
results serve to improve the planned process in form of a planning cycle (see Fig. 3).
LSP1.LS2
LSP1.LS1
Revise planning
Provider selection
LSP2.LS1
LSP2.LS1
Process
variation
+
Service
profiles
Process
model
Simulation
model
Planning completed
Fig. 3.
Automated planning cycle within the design phase
