Proceedings of the Operational Research Society Simulation Workshop 2012 (SW12)
B.Tjahjono, C. Heavey, S. Onggo, and D-J. van der Zee, eds.
OPTIMISATION OF A PRODUCTION LINE USING
SIMULATION AND LEAN TECHNIQUES
Dr. Enrique Ares Dr. Gustavo Pelaez
University of Vigo University of Vigo
School of Industrial Engineering School of Industrial Engineering
Spain Spain
[email protected] [email protected] Dr. Luis Pinto Ferreira Mr. Daniel Prieto
Instituto Politécnico do Porto University of Vigo
Escola Superior de Estudos Industriais e de Gestão School of Industrial Engineering
Portugal Spain
[email protected] [email protected] Mr. Antonio Chao
University of Vigo
School of Industrial Engineering
Spain
[email protected]
ABSTRACT
This paper presents the use of simulation to assist the decision-making process involved in implementing lean
manufacturing principles at a carton box die factory. The paper describes the application of discrete event
simulation to improve and optimise the performance of the carton box die assembly line. Simulation
experiments measure each system’s resource requirements and performance, quantifying benefits to be derived
from applying the principles of lean manufacturing. In this study, Simio is used to model and simulate different
experimental scenarios in order to quantify the impact of selected input parameters on objective functions such
as lead time. Results show that changes in the layout can reduce workers’ movements and increase productivity.
Keywords: Lean Manufacturing, Industrial Simulation, Carton Box Die Assembly Line
1 INTRODUCTION AND LITERATURE REVIEW
Simulation is one of the most valuable decision-support tools used by practitioners to solve
production issues. The objective of simulation can be that of quantifying performance improvements
which can be expected from effecting changes. Simulation is able to demonstrate the benefits of
making some decisions throughout the entire manufacturing system. Finally, simulation through
animation can provide a visual and dynamic illustration to management of how the new system would
work (Ferreira et al. 2011a; Ferreira et al. 2011b). Computer simulation is particularly useful in
analysing systems which incorporate randomness in their parameters or when there are no exact
analytical solutions (Lucko et al. 2008). Basically, simulation constitutes a technique which allows for
the transfer of reality into the computer, through the construction of a model which accurately mirrors
the behaviour of a process (in existence or in its project phase) (Ferreira et al. 2010; Ferreira et al.
Ares, Pelaez, Ferreira, Prieto and Chao
2011c). By using this technique, it is possible to proof the real performance of several models, simply
by varying a parameter, without unnecessary expenditure. In this manner, it is thus possible to
maximise productivity in a real-life scenario.
Lean manufacturing is a philosophy for structuring, operating, controlling, managing and
continuously improving industrial production systems. The key points of this philosophy include
process stability, standardized work, balanced production, visual control, continuous flow production,
time inventory management, etc. For additional information, readers are referred to Shingo (1989),
Womack et al. (1990), Monden (1993) and (Julien and Tjahjono, 2009).
A review of the literature reveals that not much research has been explicitly undertaken with the
purpose of applying simulation techniques to studies that specifically demonstrate the application of
the principles of lean thinking and execution in production systems. This could be due to the lack of a
comprehensive mechanism to identify the most significant lean drivers for the purpose of system
processing optimisation. The works which show a narrow relationship between the production-
simulation environment and the lean manufacturing philosophy can be grouped into two general
categories:
The use of simulation to establish specific parameters of a lean manufacturing system. For
example, Lummus (1995) used simulation to study three production-sequencing strategies
(mixed-model, minimum setup, and demand pull).
The use of simulation to design, test and improve lean systems. For example,
(Carlson and Yao, 1992) used simulation to pre-test various flow layouts for a low-volume,
mixed-model JIT assembly system.
In this study, one resorted to Simio, a rather innovative simulation tool, to model and simulate
two different scenarios in order to understand and quantify the impact of selected input key lean-
variables on objective production parameters, such as lead time. Simio was developed in 2007 and
represents a new approach in simulation object orientation. Among other characteristics, one should
highlight that Simio supports (Pegden, 2007) (www.simio.com, accessed 1 January 2011):
Creation of 3D animation in one step, importing 3D objects from the Google 3D Warehouse.
Importation of data from Excel worksheets.
Its own logic function (e.g. priority rules) in many languages (C++, Visual Basic, etc.).
Creation of its own intelligent objects and libraries.
2 METHODOLOGY
In the first phase, useful information for the study was collected in order to adequately represent the
current process; a detailed analysis of the production processes was then carried out to produce the
various types of carton box dies.
The data used to feed the simulation program were obtained from an existing ERP, where
important variables of each production module are updated manually. After a preliminary analysis, a
simulation model representing the behaviour of the entire factory process was constructed. The model
was fed with the data from the ERP system. In the second step, the model was validated by comparing
actual results with those obtained in the simulation model.
The application of lean manufacturing principles has led to the study of some parameters, thus
allowing for improvements to be carried out. This will help the company to make the most effective
decisions in order to reach its objectives. Some lean-manufacturing improvement proposals were:
Changes in the plant layout could result in a reduction of lead times, by simplifying the product
flow.
The redistribution of workers in the factory could streamline production to meet customers’
orders.
3 DEVELOPMENT OF THE MODEL
This work comprises the analysis, modelling and simulation of a carton box die factory. The system
manufactures two types of carton box dies (rotary and flat). These two forms of dies for cutting are
178
Ares, Pelaez, Ferreira, Prieto and Chao
shown in Figure 1. The assembly line was modelled using Simio, a discrete-event simulation tool
(see Figure 2).
Figure 1 Rotary and Flat Carton Box Die
Figure 2 Screenshot of the Simulation Model
In the model, any manufacturing order could be represented. The orders arrive at the system with
an arrival pattern that is based on an arrival table. The processing times of stations were taken from
the ERP data of recent years (see Figure 3). These times, are naturally assumed to possess a tendency
to follow a normal distribution in this type of manufacturing process. Therefore, statistics software
was used to contrast the normal distribution considered to be intrinsic to the ERP, although this
sometimes presents a very high standard deviation. This study revealed that normal distribution
appears to match these data.
In the modelling process attention must be drawn to the different processing times for each type
of product. Apart from these different processing times, orders must follow the predefined paths set in
the model. Each order is a completely different entity and is launched into the system through a
schedule provided by the ERP.
179
Ares, Pelaez, Ferreira, Prieto and Chao
Figure 3 Statistical Distribution of Process times
The aim of using simulation as a decision-making support tool is that of evaluating the
performance of a system during a specific period of time. Most simulation models start off with an
empty system and with free resources (Majan and Ingalls, 2004). In fact, the observation gathered
during the initial period of simulation will affect the accuracy of the model’s performance
measurements, leading to the formulation of erroneous conclusions regarding its performance
(Ferreira et al. 2011b). In accordance with this, a warm-up period of 1 hour was considered for this
study. During this period, the data collected were not considered for statistical purposes. The model
was finally validated by comparing the simulation results of twelve orders programmed over a 24-
hour period (see Table 1) and confronted with real data taken also taken from the ERP, which stores
all the times for each order during that period.
Table 1 Data for Validation
Time in Simio (Hours) Real Time in ERP (Hours)
DIE Nº 1 12.1 11.9
DIE Nº 2 15.5 15.0
DIE Nº 3 4.3 3.8
DIE Nº 4 3.2 3.1
DIE Nº 5 3.1 2.8
DIE Nº 6 15.6 15.2
DIE Nº 7 7.4 7.1
DIE Nº 8 12.5 12.9
DIE Nº 9 16.5 16.1
DIE Nº10 13.2 13.7
DIE Nº11 5 4.0
DIE Nº12 9.9 10.3
180
Ares, Pelaez, Ferreira, Prieto and Chao
4 RESULTS
In order to apply the previously-presented methodology, one ran simulation experiments to test some
of the proposals of the lean manufacturing concepts. One of these involves changing the plant layout
and consists of comparing the actual layout with a new layout proposal, both using the same
resources. The simulation experiments are fed with the same schedule data used for validation
(see Figure 3). After running simulation, the results presented in Figures 4 and 5 show that one has
managed to increase production from 9 to 11 in the carton box dies cut per day. This therefore implies
an improvement of more than 20% in throughput. The simulation related with the lean manufacturing
proposals of the application of reallocation, an increase or reduction of workers are, as far as tests
seem to reveal to this date, not conclusive in their results.
Figure 4 Current Layout and Results
Figure 5 Improved Lay Out and Results
181
Ares, Pelaez, Ferreira, Prieto and Chao
5 CONCLUSIONS
Simulation can be used to support decision-making and to evaluate the impact of various
opportunities for improvement. Ideally the simulation model can be used to evaluate the alternatives.
Discrete-event simulation has proved to be robust as a tool to help quantify the benefits of lean
manufacturing. The use of the simulation described in this paper can provide credible estimates for
the savings in shop-floor resources and the improvements in the time-based performance statistics to
be attained with lean manufacturing. In the case study presented in this paper, the simulation model
demonstrated the impact of lean principles in terms of reducing transportation time and floor space,
which results in significant savings in the company’s running costs.
Future work in this project would include the use of real-time data and the development of more
realistic 3D animations in the model. Another line of investigation is that of obtaining a more
realistic simulation model and more helpful decision-making elements by analysing the input
data through fuzzy logic, as well as the application of new lean manufacturing concepts as
rearrangements in workstations based on 6S techniques so as to generate different processing times in
any workstation where these approaches could be applied.
REFERENCES
Luis Pinto Ferreira, E Ares Gómez, G Peláez Lourido, A Resano Lázaro, C J Luis Pérez and
Benny Tjahjono (2011a). Evaluation of the working limits changes in an automobile assembly
line using simulation, Proceedings of 4th Manufacturing Engineering Society International
Conference, Cadiz, Spain, pp 21-23.
Luis Pinto Ferreira, E Ares Gómez, GC Peláez Lourido, José Diéguez Quintas and Benny Tjahjono
(2011b). Analysis and Optimisation of a Network of Closed-loop
Automobile Assembly Line using Simulation, The International Journal of Advanced
Manufacturing Technology, published online: 12 July 2011, Springer.
Gunnar Lucko, Perakath C Benjamin and Michael G Madden (2008). Harnessing the Power of
Simulation in the Project Management / Decision Support Aspects of the Construction Industry,
Proceedings of 2008 Winter Simulation Conference, pp 2479-2487.
Luis Pinto Ferreira, E Ares Gómez, G C Peláez Lourido and Marina Salgado (2010). Analysis of the
Influence of Conveyor Speed on the Behavior of an Automobile Assembly Line, DAAAM
International Scientific Book 2010, chapter 41, pp 463-470, B. Katalinic (Ed.), published by
DAAAM International, ISBN 978-3-901509-74-2.
Luis Pinto Ferreira, Enrique Ares Gómez, Gustavo Peláez Lourido and Benny Tjahjono (2011c).
Optimization of a Multiphase Multiproduct Production Line based on Virtual Cells, Lecture Notes
in Engineering and Computer Science: Proceedings of The World Congress on Engineering 2011,
The 2011 International Conference of Manufacturing Engineering and Engineering Management,
Imperial College London, London, U.K., pp 616-621.
Shingo S (1989). A Study of the Toyota Production System from an Industrial Engineering Viewpoint,
OR: Productivity Press: Portland.
Womack JP Jones DT and Roos D (1990). The Machine That Changed the World. Harper Perennial:
New York.
Monden, Y. (1993). Toyota Production System, an Integrated Approach to Just-in-Time. 2nd edition
Engineering and Management Press: Norcross, GA, US.
Julien D & Tjahjono B (2009). Lean Thinking Implementation at a Safari Park, Business Process
Management Journal. Vol.15, Issue: 3, page 321 - 335.
Lummus R (1995). A simulation analysis of sequencing alternatives for JIT lines using kanbans.
Journal of Operations Management, Volume 13, Number 3, pp 183-191.
Carlson J G and Yao A C (1992). Mixed-model assembly simulation. International Journal of
Production Economics, Volume 26, pp 161-167.
Pegden D (2007). Simio: A new simulation system based on intelligent objects. Proceedings of the
2007 Winter Simulation Conference, pp 2293-2300.
182
Ares, Pelaez, Ferreira, Prieto and Chao
Prasad S Mahajan and Ricki G Ingalls (2004). Evaluation of Methods Used to Detect Warm-Up
Period in Steady State Simulation. Proceedings of the 2004 Winter Simulation Conference.
AUTHOR BIOGRAPHIES
ENRIQUE ARES is a Full Professor in Manufacturing Processes Engineering at Vigo University in
Spain. He has served as a member of the several National and International Scientific Committees
(XVth CNIM 2002, MESIC’05, MESIC’07 , MESIC'09, MESIC`11, XXIIIth CNIM 2010).
Reviewer in International Journals related to Manufacturing Technologies like: JIMSP,CAD, IJMTM.
He is President of the Spanish Manufacturing Engineering Society (http://www.sif-mes.org/). His
research interests include manufacturing technology optimisation, control and simulation, operation
decision analysis and optimisation in manufacturing processes, planning and scheduling of multistage
multiproduct manufacturing systems, as well as sustainable manufacturing.
GUSTAVO PELAEZ is a Senior Lecturer in the Manufacturing Processes Engineering area at the
University of Vigo (Spain). He was awarded he best doctoral thesis in 1999 at the University of Vigo,
where he is also the Director of the Master’s Program in Design and Manufacturing Processes. His
current research interests include holonic systems, simulation and its relationship with lean
manufacturing and manufacturing technology for machining and forming processes.
LUIS PINTO FERREIRA is a Senior Lecturer at the Technical-Scientific Unit of Industrial
Engineering and Production at the School of Management and Industrial Studies (Porto Polytechnic,
Portugal). He received his Ph.D. degree in Engineering Design and Manufacture from Vigo
University (Spain) in 2011. His current research interests include systems engineering methodologies,
decision sciences, modelling and simulation, logistics, distribution and transportation, planning and
scheduling of multistage multiproduct manufacturing systems.
DANIEL PRIETO is a Lecturer in the Manufacturing Processes Engineering area at the University
of Vigo (Spain). His current research interests include simulation and lean manufacturing.
ANTONIO CHAO is a Master’s student in Engineering Design and Manufacturing at the University
of Vigo (Spain). His Master’s thesis is on simulation and lean techniques.
183
