SIMPLIFICATION AND AGGREGATION STRATEGIES APPLIED
FOR FACTORY ANALYSIS IN CONCEPTUAL PHASE USING SIMULATION
Matías Urenda Moris
Amos H.C. Ng
Jacob Svensson
Virtual Systems Research Centre
University of Skövde
Skövde, SE-541 28, SWEDEN
ABSTRACT
Despite that simulation possesses an establish background
and offers tremendous promise for designing and analyzing
complex production systems, manufacturing industry has
been less successful in using it as a decision support tool,
especially in the conceptual phase of factory design. This
paper presents how simplification and aggregation strate-
gies are incorporated in a modeling, simulation and analy-
sis tool, with the aim of supporting decision making in
conceptual phase. Conceptual modeling is guided by a
framework using an object library with generic drag and
drop system components and system control objects. Data
inputs are simplified by the use of Effective Process Time
distributions and a novel aggregation method for product
mix cycle time differences. The out coming specification is
through a Web Service interface handle by modeling sys-
tem architecture, automatically generating a simulation
model and analysis. Case studies confirm a breakthrough
in project time reduction without appreciable effects on the
model’s fidelity.
1 INTRODUCTION
Today, industry has very limited support from working
procedures, methods, and tools for analysis of complete
plants in early program stages. This leads to difficulties in
predicting the consequences from the early decisions that
are basic for robust, flexible and cost effective production.
Despite that simulation possesses an establish background
and offers tremendous promise for designing and analyzing
complex production systems, manufacturing industry has
been less successful in using it as a decision support tool,
especially in the early conceptual phase of factory design
(McNally and Heavey 2004). Unfortunately, simulation
analysis for system design is used in later phases when cost
and system performance are more or less locked.
There are several reasons to the poor use of simulation
at this phase. Industrial experience shows that decisions are
many times based on heuristic and historical approaches. If
tools are used they are mainly limited to analytical tools.
The arguments suggesting an analytical solution, instead of
a discrete event simulation (DES) approach, are among
others a more effective use of time and the lack of detail
process data (Jägstam and Klingstam 2002; Patchong et al.
2003). Remember that many conceptual models have a
very short life span, due to the nature of the task which is
evaluating different production concepts and control
strategies towards each other. The production manager is
therefore dependent on the help of a simulation expert and
needs to order models for the different concepts. If the
company has in-house experts there is still a time-
consuming dialog and process between the manager and
the expert. That particular dialog is somewhat intricate.
The simulation specialist, used to work with detail models
for planning and scheduling of existing production lines,
gets into an “ethical dilemma”. Not used to build simple
models he has his own lack of credibility to the models an
the use of estimated data. The troubles with models at con-
ceptual phase are therefore summarized in lack of data,
time and knowledge. Can these problems be overcome or
should production managers continue to design production
lines based on guesses, “this is the way we have always
done” attitude or be limited to basic queuing network mod-
els which will most certain result in bigger assumptions
than a DES model?
The objective of this research is to overcome these
problems and “frontload” the use of virtual methods to
analyze complete factories. The project targets were to in-
crease the use of simulation analysis, make them faster and
more accurate in the early conceptual phase. This is
through developing a new work method and a correspond-
ing toolset that tightly integrates model abstraction, input
data management and simulation-based optimization under
an innovative framework that is specifically designed for
production system’s designer/managers in the conceptual
design phase. As a part of the research work within the
Factory Analyses in ConcepTual phase using Simulation
(FACTS) project, which is funded by VINNOVA in Swe-
den (see Acknowledgement) and took place between Janu-
1913 978-1-4244-2708-6/08/$25.00 ©2008 IEEE
Proceedings of the 2008 Winter Simulation Conference
S. J. Mason, R. R. Hill, L. Mönch, O. Rose, T. Jefferson, J. W. Fowler eds.