Question
tudy of an innovative method based on complementarity between ARIZ, lean management and discrete event simulation for solving warehousing problems Fatima Zahra Ben Moussa a,
tudy of an innovative method based on complementarity between ARIZ,
lean management and discrete event simulation for solving warehousing
problems
Fatima Zahra Ben Moussa
a, , Roland De Guiob , Sbastien Duboisb , Ivana Rasovskac,
Rachid Benmoussa
a
a SyLPRO/National School of Applied Sciences of Marrakech, Avenue Abdelkrim Khattabi, 40000 Guliz, Marrakech, Morocco
b CSIP, ICube Laboratory, National Institute of Applied Sciences of Strasbourg, 24 Boulevard of Victory, 6700 Strasbourg, France
c ICUBE/University of Strasbourg, 4 Street Blaise Pascal, 67081 Strasbourg, France
ARTICLE INFO
Keywords:
Algorithm for Inventive Problem Solving
(ARIZ)
Lean management
Discrete event simulation
Witness
Design of experiments
Generalized system of contradictions
ABSTRACT
Warehousing challenges require not only effective management but also a new innovative strategy that can
significantly improve warehouse efficiency. This study presents an innovative method for solving warehouse
problems and achieving low-cost warehousing, based on complementarities between ARIZ (Algorithm for
Inventive Problem Solving), lean warehousing approach and discrete event simulation. In the proposed model,
the lean warehousing is first used to understand and analyze problems in order to propose solutions that improve
warehouse system efficiency. Then, a model of the improved system is developed and analyzed in a predictive
simulation software (Witness1 ), to evaluate its performance. The discrete event simulation allows for experi
ments on the system to be created and analyzed. Thus, an experimental design was developed to establish the
cause-and-effect relationships between the parameters of the system in order to formulate a generalized system
of contradictions. And finally, ARIZ is used to address related contradictions for searching for an innovative
solution, which must be subsequently implemented and evaluated in the discrete event simulation. The ap
plicability of the proposed method is demonstrated through a case study, involving an automotive company
specialized in the manufacturing of electronic modules that wants to reduce raw material warehousing costs.
Generally, the aim of the proposed method, which combined lean management, discrete event simulation and
the ARIZ algorithm, is to be able to propose solutions that go beyond the limits of continuous improvement and
optimization methods. Such an approach enables to propose a kind of continuum between traditional multi
objectives methods and inventive ones.
1. Introduction
In today's highly competitive industrial markets, warehouses in
creasingly play a significant role in supply chains. With the increasing
demand for good customer service, organizations are under pressure to
develop more efficient warehouses to satisfy customer requirements
and protect companies against variations in both productivity and de
mand rates (Drew, Mccallum, & Roggenhofer, 2008). As a result,
warehouses are now required to provide more flexibility, accessibility,
and efficient functionality to enhance organizational profits. The
warehouse plays a significant role in the supply chain, which delivers
the right quantity of the right product to the right customer, at the
optimal time, place, and price. A warehouse can be defined as a phy
sical storage system, from which products are received, transferred,
stored, picked up, sorted and accumulated, cross-docked, and shipped
(Cakmak, Gunay, Aybakan, & Tanyas, 2012).
Warehouse management, or warehousing, is the art of efficiently
operating a warehouse and is one of the key factors in supply chain
management (Lambert, Cooper, & Pagh, 1998). It is pertinent to men
tion that organizations face considerable challenges in managing
warehouses. Moreover, improving warehouse efficiency involves elim
inating any waste from a warehouse, streamlining its operations, and
improving the efficiency in every aspect of warehousing activities (Min,
2006). The successful application of continuous improvement
https://doi.org/10.1016/j.cie.2019.04.024
Received 24 March 2018; Received in revised form 22 March 2019; Accepted 15 April 2019
Corresponding author.
E-mail addresses: &..a@g(F.Z. Ben Moussa), r..o@insa-strasbourg.fr (R. De G..s@insa-strasbourg.fr (S. Dubois), i..a@unistra.fr (I. Rasovska).
1 https://www.lanner.com/en-gb/technology/witness-simulation-software.html.
Computers & Industrial Engineering 132 (2019) 124-140
Available online 16 April 2019
0360-8352/ 2019 Elsevier Ltd. All rights reserved.
Tmethodologies, especially "lean warehousing," would lead to sig
nificant improvements in a warehouse, such as the elimination of non
value-added processes, and an improvement in total lead-times and
customer satisfaction, with a zero-breakage rate for the warehouse.
In addition to continuous improvement strategies, another way to
improve warehouse efficiency is through the use of innovative strate
gies. Innovation can be recognized as a key success factor for increasing
competitiveness in a complex environment (Lambert et al., 1998). TRIZ
(The Theory of Inventive Problem Solving) is one of the most powerful
and widely accepted methods for enacting systematic innovation (Sheu
& Lee, 2011; Spreafico & Russo, 2016). TRIZ is a theory that has been
widely employed in multiple industries and domains to solve problems
and find inventive solutions. TRIZ provides a set of methods and tools
that are generally employed in preliminary technical product designs.
One of the main methods of TRIZ (the most complete one) is ARIZ
(Algorithm for Inventive Problem Solving). Historically, different ver
sions of ARIZ were proposed and tested through its use on many cases,
till ARIZ-85C, which could be recognized as the most mature meta
methods of TRIZ. ARIZ-85C clarifies the links between the different
tools and techniques of TRIZ, in order to apply them in a structured
manner. The TRIZ theory enables people to adopt a dialectical thinking
style, which guides them to understand problems as systems, con
ceptualize the ideal solution, and enhance the performance of the
system by solving contradictions, which are descriptions in TRIZ of
conflicts between problem objectives (Wang, Yeh, & Chu, 2016). TRIZ
was first proposed for solving technical problems related to product
engineering design. Subsequently, it was extended to resolve other
types of problems in other fields such as supply chains, service, edu
cation, and information processing.
Till now, no global approach to apply TRIZ methods in supply chain
has been proposed. One can recognize that previous research proposes
the integration of some TRIZ tools into various stages of DMAIC (Jiang
& Nguyen, 2015; Muruganantham, Krishnan, & Arun, 2014; Wang
et al., 2016; Xie & Li, 2009; Zhao & Zhao, 2013). These proposals were
primarily focused on product or process design, but not in management
of supply chain. Otherwise, there are limited studies in the existing
literature that deal with the application of TRIZ for problems in supply
chain operations (Chechurin & Borgianni, 2016). A previous study (Ben
Moussa, Rasovska, Dubois, De Guio, & Benmoussa, 2017) investigated
the application of TRIZ to supply chain problems, particularly to in
ventory management problems. They demonstrated that there is no
obvious or straightforward reason preventing TRIZ from being applied
to all categories of supply chain problems, including warehousing
problems. Also, the authors in Ben Moussa et al. (2017) highlight the
limitation of using the most mature meta-methods of TRIZ (ARIZ-85C)
in solving unconventional generalized contradictions. Within this con
text, the aim of this paper is to study the use of the ARIZ algorithm in
solving problems formulated by generalized contradictions. However,
using ARIZ-85C for solving problems is sometimes difficult because
TRIZ theory lacks highly specific tools and methods to understand
complex problems and formulate a related system of contradictions. For
this reason, in this paper, the authors study the synergetic application
and complementarity of lean warehousing, discrete event simulation
and ARIZ 85C, and propose a framework based on these three methods.
In this method, lean warehousing and discrete event simulation are
used to provide the necessary data to formulate relevant systems of
contradictions that prevent the fulfillment of warehouse efficiency ob
jectives, whereas ARIZ-85C is adopted to address the formulated con
tradictions. Unlike the objectives of lean warehousing and discrete
event simulation, which are to improve warehouse efficiency and
search for best compromises between the problem constraints, TRIZ
allows to move beyond compromises by overcoming contradictions.
Then, the complementarity of lean warehousing, discrete event simu
lation and ARIZ-85C allows better results and performances to be ob
tained than solving the corresponding problems using only lean ware
housing or simulation software. Thus, this study contributes to existing
works by using the lean warehousing methodology in conjunction with
the inventive algorithm ARIZ-85C and highlights the application of
ARIZ-85C to solve problems formulated by generalized technical and
physical contradictions, as there is no existing literature that deals with
this subject.
The remainder of this paper is organized as follows. Section 2 in
troduces the materials and methods, including lean warehousing, ARIZ
algorithm, generalized system of contradictions and a literature review
of papers relating to the application of TRIZ tools with the lean-ware
housing and DMAIC. Section 3 presents the proposed innovative
method based on the complementarity of ARIZ, simulation and lean
warehousing. Section 4 presents the result of applying the proposed
method to a warehousing problem through a case study involving an
automotive supplier producing electronic modules. A discussion is
provided in Section 5, and conclusions and future research are dis
cussed in Section 6.
2. Materials and methods
2.1. Lean warehousing
Warehouses comprise a substantial component of logistic opera
tions, and they play a critical role in matching product demand with
supply across different echelons in a supply chain. Today, warehouses
represent centers for value-addition rather than centers for storage
(Johnson & McGinnis, 2010). Moreover, many warehouses continue to
suffer from significant inefficiencies, as forklift operators waste time
and resources hunting and digging because they lack adequate in
formation on the locations of items and optimal routes for storage, re
plenishment, and retrieval actions (Hackman, Frazelle, Griffin, Griffin,
& Vlasta, 2001). Warehousing activities/operations include receiving,
unpacking/sorting, storing, picking up, sorting, packing, and shipping
(Aminoff, Kettunen, & PajunenMuhonen, 2002). Furthermore, Bottani,
Cecconi, Vignali, and Montanari (2012) have shown that order picking
is the most labor-intensive and costly activity of most warehouses, with
approximately 55% of the total warehouse operating expenses relating
to order picking operations. To improve a warehouse's efficiency and
optimize its operations, many methods have been presented in the lit
erature, such as routing methods to determine the sequences and routes
of picking, storage assignment methods to assign items to storage lo
cations based on certain rules, order batching algorithms to group two
or more customer orders into one picking order, warehouse layout
design to minimize total travel distances, and methods of the lean
management philosophy to eliminate waste from warehousing activ
ities and non-value added processes, and improve the total lead-times to
achieve low-cost warehousing.
The lean management philosophy is the most widely known ap
proach to industrial continuous improvement, and comprises an op
erations management approach that refers to non-value adding activ
ities as waste. The key philosophy of lean management is that the
elimination of non-value adding activities, variability, and inflexibility
is imperative in order to deliver value to customers with the right time,
quantity, and quality and at the minimum cost (Drew et al., 2008). The
implementation of lean management leads to the removal of eight types
of waste: transport, inventory, motion, waiting, over-processing, over
production, defects, and talent (Kadarova & Demecko, 2016).
The subject of lean warehousing deals with the application of lean
concepts and practices in warehousing operations to improve the
warehouse efficiency. Furthermore, the benefits associated with lean
warehousing include an improved quality of operations, improved
profitability, and a leading edge in the market competition (Mustafa,
2015). To this end, lean warehousing includes numerous tools and
techniques, for instance VSM (value stream mapping), Mudas, root
cause analysis, and activity diagrams. The application and choice of
lean tools depends on the user's experience and knowledge. Further
more, when applying lean management or lean warehousing to
F.Z. Ben Moussa, et al.
Computers & Industrial Engineering 132 (2019) 124-140
125complex processes, the use of a structured method for problem solving,
such as DMAIC (Define, Measure, Analyze, Improve, Control), is
strongly recommended. DMAIC is a structured problem-solving method
and roadmap that can be used for any project, or with the goal of de
livering supply chain improvements (Bottani et al., 2012). The term
DMAIC stands for the five main steps in the process: identifying the
problem, measuring the current status, analyzing where the problem is,
improving the performance, and controlling the new process.
2.2. Algorithm for inventive problem solving (ARIZ)
TRIZ, which is the Russian acronym for the Theory of Inventive
Problem Solving, was developed by Genrich Altshuller starting in 1940.
Unlike problem solving methods that accept randomness in the in
novation process, TRIZ is a knowledge-based systematic methodology,
which provides a logical approach to develop creativity for innovation
and inventive problem solving, while rejecting compromises between
objective restrictions and specific situation limits (Ilevbare, Probert, &
Phaal, 2013). TRIZ is based on the three fundamental concepts of re
sources, ideality, and contradictions. Ideality is one of the most pow
erful concepts of TRIZ. The meaning of this ideality is to maximize the
constraints of the considered problem in order to find the more robust
solutions. According to ideality, the ideal final result consists of solving
the problem without introducing new resources, and searching for
configurations of the system where the desired result is achieved by
itself. An ideal system may not be possible to achieve, but knowledge of
the ideal system helps in improving an existing system, by maintaining
the features of the ideal system as a goal to aim for. Contradictions can
be either technical or physical, and one hypothesis of TRIZ is that such
contradictions exist as soon as a problem exists without a known so
lution. Technical and Physical Contradictions are just two different
models to represent the same problem, but they always exist together.
In addition to the key concepts of TRIZ, Altshuller developed a set of
methods, tools, and a knowledge base for systematically generating
new ideas and solutions for problems that arise during the evolution of
technical systems. Subsequently, further studies based on TRIZ have
been conducted in order to extend TRIZ approaches to solve other kinds
of problems, not just those related to the evolution of technical systems
(Khomenko, De Guio, & Cavallucci, 2009).
ARIZ is the Russian acronym for the "Algorithm for Inventive
Problem Solving," developed by the TRIZ creator Genrich Altshuller.
ARIZ is a process that links the tools and techniques of TRIZ, in order to
apply them in a structured manner, to evolve from a complex, fuzzy,
problem to a point where it can be solved. Throughout its evolution,
several methods and tools were defined in TRIZ. To avoid considering
TRIZ as a tool box but to well structure the use of these methods and
tools in a systemic way, it was necessary to build a meta-method to
guide the designers to combine them in a relevant and robust way. ARIZ
has been developed through several versions, and the last one proposed
by Altshuller is ARIZ-85C (Fiorineschi, Frillici, & Rissone, 2015). The
framework of ARIZ-85C, illustrated in Fig. 1, consists of nine algo
rithmic parts (Altshuller, 1985). These parts can be selectively em
ployed according to the practical situation of solving a problem. Parts 1
to 4 are dedicated to analyzing the problem situation by converting the
initial problem into a formulated description, analyzing the problem
model by identifying the existing resources for solving the problem,
defining the ideal final result and the physical contradictions, and
mobilizing and using substance-field resources by increasing the
availability of resources by modifying the previously defined existing
resources, respectively. Scientific effects and a knowledge base are
applied in part 5. If the problem remains unsolved at the end of part 5,
we move to part 6 to change or substitute the problem. Otherwise, we
move to the last parts (parts from 7 to 9), which are dedicated to
analyzing the method of resolving the problem, applying the obtained
solutions, and analyzing the entire problem-solving process, respec
tively.
It is important to note that ARIZ-85C only addresses the re
formulation and resolution of the problem. Indeed its last version,
ARIZ-85C, the method has been developed to clearly state a problem
and to use cognitive tools to find a solution to this problem. But, con
trary to the previous versions, it lacks specific tools and methods to
analyze the problematic situation and to identify the prior problem to
consider. Contradictions are the pillar of ARIZ, and solving an inventive
problem means solving the related contradictions. Two steps of con
tradiction formulation are proposed. A Technical Contradiction is the
starting point of the whole problem-solving process using ARIZ.
Sometimes, the Technical Contradiction within a problem is clearly
evident, and at other times it seems that a problem does not contain any
Technical Contradiction, because it is hidden within the problem con
ditions (Davide & Montecchi Tiziano, 2012). This point reinforces the
necessity to well analyze the situation at the beginning of problem
solving process, to be sure to identify the good contradictions to con
sider. The second level of contradiction formulation is the Physical one,
stating the core of problem and explaining why the problem exists, and
why the system presents contradictory requirements.
2.3. Generalized system of contradictions
2.3.1. Classical TRIZ contradictions
One of the main ideas of TRIZ is to guide the evolution of systems by
identifying the contradictions that have to be overcome (Dubois,
Rasovska, & Guio, 2008). Contradictions can be either technical or
physical. Technical Contradictions (TC) appear when there are con
flicting requirements regarding two Evaluation Parameters (EP) of a
technical system. Physical Contradictions (PhC) appear when the same
Action Parameter (AP) should exhibit different properties at the same
time (Fresner, Jantschgi, Birkel, Brnthaler, & Krenn, 2010). Based on
the two contradiction models (technical and physical), a system of
contradictions was introduced by Khomenko et al. (2009) to represent
the causal relation between the AP and EP, as illustrated in Fig. 2. This
system of contradictions is based on the existence of a contradiction of
the parameter and a contradiction of the system, which justifies the
need for two different states of the parameter (Dubois, Rasovska, &
Guio, 2009).
Example. In an inventory management system managed by a Kanban
card system, a conflict between the stock size (capacity) and the
customer level service rate appears when the need for some products
increases (see Fig. 3). Indeed, the problem appears because there is a
lack of space to store an extra quantity in the stock. Moreover, if the
service concerned by this increase in demand does not receive the
necessary quantity, there will be a stock-out. To solve this problem, the
increased requested quantity must arrive to the stock without
increasing the storage space and without creating stock-out. To reach
this desired result, the reorder quantity of the Kanban card should be
big and small at same time, which is contradictory. This contradiction
where one parameter has to take two different values at the same time
is the so called physical contradiction in classical TRIZ.
2.3.2. Generalized system of contradictions model
The classical TRIZ contradiction is limited by the number of eva
luation parameters, which does not exceed two parameters. Indeed,
when dealing with real and complex problems, this model cannot be
used to represent the system of contradictions, owing to the problem
multidimensionality. In order to represent the system of contradictions
for a multidimensional problem, some improvements of the classical
contradictions models were proposed in the literature, such as the
Multiple-to-Multiple Parameter Contradictions (Sheu & Chen, 2011)
and the Generalized System of Contradictions (Dubois et al., 2009;
Dubois, De Guio, & Rasovska, 2011). Indeed, the Generalized System of
Contradictions is a generalization of the classical contradictions of
TRIZ, which links two Generalized Technical Contradictions (GTC) and
F.Z. Ben Moussa, et al.
Computers & Industrial Engineering 132 (2019) 124-140
126a Generalized Physical Contradiction (GPC) (see Fig. 4). In the model of
GPC, the single Action Parameter and its values (defined in the classical
PC) are replaced by a set of Action Parameters and two concepts (two
combinations of states of the related APs). The generalization of the
Technical Contradiction is built on two concepts, two sets of Evaluation
Parameters, which are satisfied or not. The model of GTC does not rely
Fig. 1. Framework of ARIZ (according to (Altshuller, 1985)).
F.Z. Ben Moussa, et al.
Computers & Industrial Engineering 132 (2019) 124-140
127only on two EPs, but generalized the definition, by emphasizing the
impossibility to satisfy simultaneously two set of EPs. Thus, the Gen
eralized System of Contradictions represents the generalization of the
classical TRIZ system of contradictions, where two concepts based on a
set of action parameters satisfy two sets of evaluation parameters. The
desired result is then the simultaneous satisfaction of the two sets of
evaluation parameters.
Example. Let's take the same inventory management system presented
in the previous section. A solution that focuses on increasing the
number of Kanban cards has been proposed. But, when solving the
contradiction presented in Fig. 3 by applying this solution, another
evaluation parameter is deteriorated (the transportation cost). Thus,
when the stock size and the stock-out meet the problem requirements,
the transportation cost does not meet the requirements. To solve this
system of contradictions (presented in Fig. 5), the stock size, the stock
out and the transportation cost should meet the problem requirements.
2.4. Lean-DMAIC and ARIZ
In the existing literature, one can find proposals of integration of
TRIZ with methods such as lean warehousing and DMAIC. In
Muruganantham, Krishnan, and Arun (2013) and Muruganantham et al.
(2014), the authors propose a synergetic approach of lean methods with
TRIZ to solve problems in a more effective way and to obtain better
results. The proposed approach was applied to reduce costs for manu
facturing components and to improve the productivity of industrial
processes. In this approach, lean methods are used to identify waste and
to identify and alleviate problems; TRIZ is employed to find the optimal
approach to reduce waste and to solve problems by resolving Technical
Contradictions through the use of one of the most widespread TRIZ tool,
the 40 inventive principles and its related matrix. These principles have
been extracted from the analysis of technical patents, and the matrix is
a statistical organization of the principles to propose the most used ones
for a given Technical Contradiction. The application of these principles
to non-technical problem, and more over the relevancy of the statistical
proposal, begs question. In Jiang and Nguyen (2015), the authors in
tegrate the TRIZ methodology with the "improve" phase of the lean
DMAIC approach, as a quality improvement strategy. Likewise, in the
proposed method, the authors employ only the tools dedicated to sol
ving Technical Contradictions, the contradiction matrix and the in
ventive principles. In Wang et al. (2016), the authors integrate TRIZ
methodology with the "analyze" phase of the DMAIC approach to re
volutionize the manner in which companies develop new products. In
the "analyze" phase, three basic TRIZ tools are used: (1) the engineering
parameters to describe technical conflicts and inventive principles to
solve technical contradictions, (2) a knowledge database of scientific
effects (physical, geometrical, and chemical) to solve problems, and (3)
the substance-field model for modeling technological problems. In
Tatjana Sibalija and Vidosav (2009); Xie and Li (2009); Zhao and Zhao
(2013), the authors propose incorporating TRIZ into various stages of
DMAIC to integrate the ideas of TRIZ with DMAIC. The objective of this
integration is to increase the effectiveness of DMAIC deployments.
From the existing literature, one can recognize that the proposed
methods in the cited papers share a common point, which is the in
tegration of some TRIZ tools into the DMAIC approach to improve the
overall potential of DMAIC for existing systems. In this article, the
authors aim at proposing a global approach, relying on the com
plementarities of ARIZ, lean warehousing and simulation tools, to solve
warehousing problems. This approach is proposed in the next section.
This proposal starts with a traditional resolution of problems with lean
warehousing, an optimization of the proposed solutions with simula
tion, and finally, if not satisfying, considering these optimized solutions
as the entry point to apply ARIZ.
2.5. Research methodology
Within the context of systematic innovation in the supply chain
areas, this paper aims at studying the use of the ARIZ-85C algorithm to
solve complex problems, formulated as a Generalized System of
Contradictions. This original proposal will also be illustrated through
an industrial case study, to validate the applicability of ARIZ-85C to
solve Generalized System of Contradictions. The chosen case study is
related to the warehouse area, thus distinguishing itself from the ma
jority of other studies (Ben Moussa et al., 2017).
3. Outline of the proposed methodology
Based on the lean warehousing methodology, following the DMAIC
approach, the discrete event simulation and the algorithm for inventive
problem solving (ARIZ) given above, the proposed model is illustrated
Fig. 2. Classical TRIZ system of contradictions.
Fig. 3. Example of a classical TRIZ system of contradictions.
Fig. 4. Generalized system of contradictions.
Fig. 5. Example of a generalized system of contradictions.
F.Z. Ben Moussa, et al.
Computers & Industrial Engineering 132 (2019) 124-140
128in Fig. 6. In this proposed model, lean warehousing is used to improve
the system, discrete event simulation software is used to optimize the
improved system and bring the problem to a point where the for
mulation of contradictions preventing the attainment of the problem
objectives can be achieved, and ARIZ is used to solve these contra
dictions. One of the main questions related to this proposed metho
dology is that ARIZ has been defined to treat classical TRIZ contra
dictions. But, as will be illustrated, there is no difficulty to apply this
generic algorithm with the Generalized System of Contradictions.
Step 1: Apply the methodology "lean warehousing- DMAIC"
The main purpose of step 1 is to improve the warehouse and its ef
ficiency by eliminating waste from the material handling activities and
non-value-added processes. Lean warehousing is a useful process im
provement methodology, which is applied to warehouses and described
as a systematic methodology for eliminating waste and reducing the
complexity of a process. It is employed in the first step of the proposed
methodology to optimize the performance of a warehouse. In general, the
DMAIC cycle is employed to solve problems via a lean approach by
following the different phases of DMAIC in order to frame and structure
any lean implementation project. However, the choice of lean manage
ment tools to apply always depends on the problem to be solved. The lean
management toolbox contains a wide range of tools and methods, to re
spond to any situation in a lean management project. The choice of ap
propriate tools and methods depends on the project requirements.
Moreover, the control step of DMAIC is not applied in the proposed
methodology, as the objective of the proposed "ARIZ based lean ware
housing and computer aided simulation" method is to look for more
opportunities for problem improvement. Furthermore, the evaluation of
the proposed concept of solutions is performed with the aid of simulation.
Step 2: Evaluate the solutions using discrete event simulators
The main objective of this step is to implement the concept solutions
and provide the decision variables for the simulation program to con
firm that there is a measurable improvement. To do so, a model of the
future warehouse system of the studied problem is implemented on a
discrete event simulator (Witness 14 is used here). The simulation
program will run the model and provide results for the objective
Fig. 6. ARIZ based lean warehousing and computer aided simulation.
F.Z. Ben Moussa, et al.
Computers & Industrial Engineering 132 (2019) 124-140
129function, to evaluate the future system efficiency. The chosen simula
tion software (Witness 14) provides tools for random experiments,
called 'experimenter' to find the best solutions (the Pareto front) for the
problem in the solution space.
Step 3: Use the design of experiments method
The Design of Experiments (DoE) is a systematic method for de
termining the relation between the input and output parameters of a
process. This information is required to manage process inputs in order
to optimize the output.
In this step, we propose the realization of two designs of experiments:
The design of experiments of the initial situation (DoE 1) and design of
experiments of the improved situation (DoE 2). Each DoE is characterized
by (1) a set of action parameters X = (X1, ..., Xk), (2) a set of evaluation
parameters Y = (Y1, ..., Yn), and (3) a set of experiments E = (E1, ..., Ep).
Each experiment Ei is characterized by a set of values (Vi1, ..., Vik) at
tributed to the set of action parameters, and by a set of values
(Zi1, ..., Zin) taken by the evaluation parameters, as listed in Table 1.
The realization of DoE1 and DoE2 will help to clearly visualize the
difference between the two situations, and compare their performances.
Step 4: Extract and formulate the generalized technical contra
dictions of the problem
The extraction and the formulation of a generalized system of
contradictions is one of the most important elements in the proposed
methodology. The generalized system of contradictions represents the
result of the analysis performed in the previous steps, and is the entry
point for the next step, which deals with the application of ARIZ.
After realizing two designs of experiments, one for the initial si
tuation (DoE1) and the other for the improved situation (DoE2), the
next step is to transform each DoE response matrix (Zij) into a binary
form, in which all the variables are restricted to binary values. Thus, Zij
takes the value of 1 if the evaluation parameter Yj is satisfied by the
experiment Ei and Zij takes the value of 0 otherwise. If no experiment Ei
enables the satisfaction of all the evaluation parameters Yj, this means
that no solution can be found in Table 1 for the performed DoE. Con
sequently, a Pareto set with at least two points exists for this problem,
and a classical TRIZ system of contradictions, or a generalized system of
contradictions, can be formulated. First, we start by creating the rela
tions of dominance between the different experiments Ei, in order to
find the non-dominated solutions from DoE1 and DoE2 (Pareto front).
Following the well-known concept of Pareto dominance, one solution
(S1) is better than another (S2), or S1 dominates S2, if the set of values
taken by the evaluation parameters for S1 is better than the set of values
taken by the evaluation parameters for S2. If the Pareto front is com
posed of solutions from DoE1 and DoE2, then a generalized system of
contradictions exists. If not, only a classical TRIZ system of contra
dictions exists. Finding the best generalized technical contradictions
involves finding the existing conflicts between the evaluation para
meters from the non-dominated solutions. To complete the description
of the system of contradictions, it is necessary to find the generalized
physical contradiction (GPC) that leads to the generalized technical
contradictions (GTCs). The GPC is composed of two concepts, based on
a set of action parameters that characterize the experiments of the non
dominated solutions (see Fig. 4).
Step 5: Apply ARIZ to address the contradictions
The main objective of this step is to propose concepts of solutions
for the studied problem, through the resolution of the physical con
tradiction or the generalized physical contradiction identified in Step 4.
The physical contradictions are conflicts among the design variables
(action parameters of the DoE) explaining (causing) the Pareto frontier
(effect). This model is different from the classical lean cause-effect
models, in that it analyzes conflicts of causes resulting in conflicts of
effects. In the previous step, we proposed formulating the system of
contradictions. This information allows the transition from an initial
problem situation to a clearly formulated and simplified description of
the problem. The identification of available resources may be useful for
solving the problem. The Ideal Final Result (IFR) and the Physical
Contradiction (PhC) that prevents the achievement of the IFR should be
formulated. The existing TRIZ principles should be used to solve the
formulated physical contradiction.
4. Results from the application of ARIZ based lean warehousing
and computer aided simulation method
In this section, we present a case study to demonstrate the applic
ability of the proposed method (ARIZ based lean warehousing and
computer aided simulation). The case study focuses on the reduction of
the raw material warehousing costs of an automotive company, spe
cializing in the manufacturing of electronic modules, by reducing the
number of operators working in the warehouse. The plant is composed
of two production areas called the SMT area and power area. The fac
tory operates for 24 h a day, six days a week. To ensure the production
continuity of the two areas, 48 operators (called feeders) work every
day to supply the various production lines with the necessary raw
materials to manufacture the electronic modules. The 48 feeders are
divided into three shifts (16 operators per shift, working eight hours a
day). Depending on the nature of their tasks, the feeder operators are
divided into four teams:
- team for reception and storage of raw material (two operators per
shift)
- team for reception of the raw material used in the SMT production
area (three operators per shift)
- team for raw material preparation for the SMT production area
(seven per shift)
- team for raw material preparation team for the power production
area (four operators per shift)
The storage warehouse for raw materials is composed of three main
areas: a reception area for raw material, a storage area for raw material
intended for the "SMT" production area, and a storage area for raw
material intended for the "power" production area. The preparation of
raw material is carried out in trolleys.
The company is implementing a downsizing policy. Thus, the supply
chain manager wishes to reduce the number of operators working in the
Table 1
Design of experiments matrix.
X1
X2
...
Xk-1
Xk
Y1
Y2
...
Yn-1
Yn
E1
V1,1
V1,2
V1,k-1
V1,k
Z1,1
Z1,2
Z1,n-1
Z1,n
E2
V2,1
V2,2
V2, k-1
V2, k
Z 2,1
Z 2,2
Z 2, n-1
Z 2, n
E3
V3,1
V3,2
V3, k-1
V3, k
Z 3,1
Z 3,2
Z 3, n-1
Z 3, n
Ep-2
Vp-2,1
Vp-2,2
Vp-2, k-1
Vp-2, k
Z p-2,1
Z p-2,2
Z p-2, n-1
Z p-2, n
Ep-1
Vp-1,1
Vp-1,2
Vp-1, k-1
Vp-1, k
Z p-1,1
Z p-1,2
Z p-1, n-1
Z p-1, n
Ep
Vp,1
Vp,2
Vp, k-1
Vp, k
Z p,1
Z p,2
Z p, n-1
Z p, n
F.Z. Ben Moussa, et al.
Computers & Industrial Engineering 132 (2019) 124-140
130warehouse, as he considers that the work necessary for reception and
preparation does not require such a large number of operators. On the
contrary, the team leaders of feeder operators ask to increase their
numbers, because they fail to complete their tasks in time without the
intervention of external resources.
According to the ARIZ based lean warehousing and computer
aided simulation model, the implementation results are presented
below.
4.1. Step 1: Apply the methodology "lean warehousing- DMAIC"
4.1.1. Step 1.1: Define the problem statement (define)
In this step, the five W's and one H "5w1h" method (Table 2) is first
applied to define the problem. Then, in order to define the process/
needs of the process customer, the SIPOC "Supplier Input Process
Output Customer" diagram (Fig. 7) is employed.
- 5w1h method:
- SIPOC diagram:
4.1.2. Step 1.2: Measure the current state of the problem (measure)
The main objectives of this section are documenting the current
situation of the activities of the operators working in the raw material
warehouse and identifying their non-value-added activities (Mudas).
- The Business Process Model (BPM) diagram
The reception of the raw material is carried out in the reception area
by the reception and storage team, and then the raw material is con
trolled by the quality control team. Then, raw material for power is
addressed in the storage area for raw material intended for the "power"
production area (PA), and the raw material for SMT is addressed in its
temporary location. Next, operators of the reception team of the RM
used in the SMT PA transmit the boxes of RM, unpack them, check the
references (check the labels of each piece or set of RM to ensure that the
arrived product corresponds to the one noted on the box), and then
store them in the racks.
In the storage warehouse intended for the SMT PA, there are 19
trolleys to fill. The trolleys are localized in the trolley zone, and the raw
material is stocked in several racks. The trolleys are composed of a large
number of empty slots (up to 280 places), which are characterized by
the reference of the RM to be put in place. To fill the trolleys, operators
make round trips from the trolley zone to the storage racks to pick up
the necessary RM. In a round trip, an operator picks up only one re
ference. However, in the storage warehouse intended for the power PA,
the trolleys are mobile, and operators push them during their picking
up operations.
All the operations performed by operators are managed by a system
with barcode labels (with reader guns), linked with the ERP of the
company. Stock status is known throughout the system. Thus, any
product, its quantities, and its locations are recorded on the system.
By using the BPM diagram, we describe the activities of operators
working in the warehouse described above in detail (Fig. 8).
- The identification of waste/Mudas:
The identification of Mudas (non-value-added activities) is the pri
mary goal of this step. Mudas can be classified into several categories,
which include transportation, waiting time, unnecessary movement,
inventory, over-processing, and defects. A field analysis for the opera
tors' activities was carried out to identify the Mudas. Then, a time
measurement was performed to calculate the time spent in each Muda.
The following table classifies the identified Mudas by category.
Table 2
5w1h to define the problem.
what What is the problem?
- Reduce the number of operators (feeders) working in the warehouse.
who Who has the problem?
- The supply chain manager.
where Where is it happening?
- In the storage warehouse for raw material.
when When is it happening?
- 24 h a day, seven days a week, when the factory is in production
activity.
how How can this problem be overcome?
- Use the lean management approach to solve the problem and improve
the main process.
why Why solve this problem? What are the quantified issues?
- To eliminate waste from the material handling activities.
- To organize the warehouse's operations.
- To achieve low-cost warehousing.
Suppliers of the raw
material
Raw material
Delivery form
Feeding trolleys
filled by the raw
material
Production lines
Reception and
control of the
raw material
Storage
of the raw
material
Preparation of
the raw
material in the
feeding
trolleys (order
picking)
bring the
feeding
trolleys to the
production
site
Inputs
Suppliers
Process
Outputs
Customers
Fig. 7. The SIPOC diagram.
F.Z. Ben Moussa, et al.
Computers & Industrial Engineering 132 (2019) 124-140
131From Table 3, the most expensive non-value-added activity is the
"unnecessary travel/round trips" in the storage area intended for the
SMT during the picking-up operations. This waste costs 37 h a day,
divided between the 21 feeders (seven operators per shift) for the SMT
production area. The next step will focus on the elimination of the se
lected Muda.
Fig. 8. The BPM diagram describing the activities of operators.
F.Z. Ben Moussa, et al.
Computers & Industrial Engineering 132 (2019) 124-140
1324.1.3. Step 1.3: Analyze the root causes of the problem (analyze)
In this step, a cause and effect analyses is performed using the
Ishikawa diagram to determine the principal root causes of the most
expensive Muda, "unnecessary travels/round trip in the storage area in
tended for the SMT production area," identified in the previous step
(Fig. 9). The Ishikawa diagram identifies and maps out many possible
causes for a problem, related to several categories, from a brain
storming session.
4.1.4. Step 1.4: Identify possible improvement actions to eliminate the root
causes of the problem (improve)
After identifying the root causes of the problem, the problem-sol
ving step to propose improvement solutions follows. To generate ideas
for improvement or concept solutions, tools such as brainstorming,
creativity techniques, lean management tools, and specific warehouse
optimization methods (see Section 2.1) can be employed. Indeed, there
exist four methods for order picking optimization to reduce the travel
distance of an order-picker: (1) storage location assignment, (2) ware
house layout design, (3) order batching algorithms and (4) pick-routing
methods. To solve the present problematic situation, the authors focus
on the pick routing and the layout design methods. Thus, it is proposed
to change the order picking strategy from a round trip strategy to a
traversal strategy, and make some changes to the warehouse layout to
minimize the total travel distance. In the initial situation, an operator
fills the trolleys of the raw material one by one, by moving back and
forth from the location of the trolleys ("trolley zone") to the storage
racks where the product to pick up is stored. In this situation, the op
erators adopt a random picking sequence. Indeed, the order in which
the products are picked up has no influence on the total workload of the
operators. In the improved situation, it is proposed that operators
Table 3
Identification of Mudas.
Mudas
Processes
Feeders for the SMT production area
Feeders for the Power production area
Transportation
Waiting time
Unnecessary movement
In the storage area for SMT, operators do not move the trolleys during
the picking operations. They make several round trips from the trolley
area to the storage racks to collect the raw material needed to fill each
trolley. The number of round trips is equal to the number of product
references to put on the trolley.
35 h a day
In the storage area for power, operators move a lot between the
storage racks to prepare the RM. operators can go around the store
every time to find the component to be picked up.
13 h a day
Inventory
Over-processing, unnecessary
treatments, additional
control
A weekly inventory is done for the SMT storage area. 27 operators
participate in this inventory. Each operator spends an average of
45 min on the inventory.
(27 op * 45 min) = 20.25 h (per week)
An inventory is done three times a week for the storage area for
power. 15 operators participate in this inventory. Each operator
spends an average of 25 min in the inventory.
(15 op * 25 min) = 18.75 h (per week)
Defect
The operators sometimes forget to scan the barcode of a component to
declare its new location during the operations of picking up. This
generates a bad image of the stock on the information system.
The operators sometimes forget to scan the barcode of a component to
declare its new location during the operations of picking up. This
generates a bad image of the stock on the information system.
Unnecessary
travels in the
storage area
intended for the
"SMT"
Method
Workers
Materials
Environment
Part numbers are pasted on the trolley
A type of RM is always stored in different locations.
The operator does not move the trolleys
Operator does not have pick list
Random storage of the RM
The space between the storage racks does
not allow more than two trolleys to pass at
the same time
existence of a
great variety of
types of the
raw material
several
operators work
at the same
time on filling
trolleys with
raw material
Fig. 9. Ishikawa diagram.
F.Z. Ben Moussa, et al.
Computers & Industrial Engineering 132 (2019) 124-140
133perform a warehouse tour between the storage racks by pushing the
trolley to fill and following the shortest path through the storage racks
(see Fig. 10). Each trolley must be filled in one journey between the
racks, while the starting and ending point of this journey is the trolley
area. To achieve this, operators must have a pick-up list that indicates
the order picking to follow for each trolley. As the raw material is
randomly stored in the warehouse, the picking sequence will be defined
based on the locations of products in the storage racks according to the
information system. This strategy leads to a route in which the racks,
that are to be visited, are totally traversed. In this improved situation,
the trolley becomes mobile, and the picking sequence becomes pre
defined. As part of the same improvement action, we make minor
changes to the warehouse layout.
4.2. Step 2: Evaluate the solutions with simulators
This step presents the results from the implementation of the pro
posed solution. The model of the future warehouse system for the pro
blem has been implemented on a discrete event simulator (Witness 14),
as shown in Fig. 10. In order to simulate the system, the identification of
the different parameters of the problem must be performed at this stage,
based on the problem objectives. Thus, the evaluation parameters of the
problem and the action parameters that influence these are:
PE1: delay to fill the 19 trolleys must be less than or equal to five
hours, to create a 3-hour availability to operators, in order to assign
them new tasks to perform.
PE2: total load for operators be less than or equal to 20 h, at a rate of
one hour per trolley.
PE3: medium capacity reserve (which represents the available time of
an operator after completing his tasks of filling the trolleys with raw
material) must be greater than or equal to three hours, to assign to
operators new tasks to perform.
PE4: total waiting time must be equal to zero minutes.
PE5: total travel time of operators must be less than two hours, at a
rate of two full warehouse tours for each trolley
PA1: picking sequence,
PA2: number of available operators,
PA3: type of trolleys.
The objective of simulating the new system is to compute its per
formance, in order to examine whether there are measurable im
provements or not compared with the initial system. Furthermore, this
simulation will allow the optimal solution (the Pareto optimum) for the
problem to be determined, by mean of the Witness experimenter tool.
As stated previously, the goal of the present problem is to reduce the
number of operators. After simulating the impact of the proposed so
lution on the studied picking process, the results indicate that there is a
very interesting improvement in the evaluation parameters (see
Table 4). The travel time decreases from 12.37 h to 1 h, reflecting a
91.6% improvement in the time lost in round trips between the "trolley
zone" and the storage racks. Furthermore, the delay to fill the 19
H
Trolleys zone
1 2 3 4 5 6 7 8 9 10
1
1
1
2
1
3
1
4
1
5
1
6
1
7
1
8
1
9
Rack
Trolley
Fig. 10. Witness model of the warehouse problem.
Table 4
Values of the evaluation parameters for the different situations.
Number of
operators
Delay to fill the 19
trolleys (h)
Total load for
operators (h)
Medium capacity
reserve (h)
Total waiting
time (h)
Total travel time of
operators (h)
Initial situation
7
4.44
30.25
3.68
0
12.37
Improved situation
7
3.22
20.73
5.04
1.95
1.00
Optimal solution obtained with optimization 4
4.98
19.42
3.14
0.64
1.00
F.Z. Ben Moussa, et al.
Computers & Industrial Engineering 132 (2019) 124-140
134trolleys decreases from 4.44 h to 3.22 h. Thus, the number of operators
can be decreased since there is a margin of 1.78 h to reach the limit of
"the delay to fill the 19 trolleys" (which must not exceed 5 h). There
fore, by using the Witness 14 "experimenter" tool, the new system
configuration reduces the number of operators to 4 operators per shift.
Thus, the system gains 3 operators per shift, which equates to 9 op
erators for the 3 shifts.
4.3. Step 3: Use the design of experiments (DoE) method
In this step, two designs of experiments are realized in the Witness
experimenter tool, the design of experiments of the initial situation
(DoE1) (see Table 5) and design of experiments of the improved si
tuation (DoE2) (see Table 6), to determine the relation between the
input and output parameters of each system and to visualize the dif
ference between the two situations.
DoE1 and DoE2 are characterized by:
(1) The set of action parameters X = (X1 = sequence of picking,
X2 = number of available operators, X3 = type of trolleys).
(2) The set of evaluation parameters Y= (Y1 = delay to fill the 19
trolleys, Y2 = total load for operators, Y3 = medium capacity re
serve, Y4 = total waiting time, Y5 = total travel time of operators).
(3) The set of experiments E = (E1, ..., E19) for (DoE1) and
E = (E20, ..., E38) for (DoE2).
4.4. Step 4: Extract and formulate the generalized technical contradictions
of the problem
In this step, a step-by-step approach is followed to extract the best
system of contradictions corresponding to the studied problem. The
system of contradictions represents the entry point for the next step.
The evaluation parameters of the two designs of experiments should
be transformed into binary values, in order to clearly identify the ob
jectives achieved in each experiment. Table 7 combines the results of
DoE1 and DoE2, where an evaluation parameter takes the value "1" if it
is satisfied by the experiment Ei, and takes the value "0" otherwise.
From Table 7, it is seen that no solution can be found, because no
Table 5
Design of experiments of the initial situation (DoE 1).
Experiments X1: sequence of
picking
X2: number of
available operators
X3: type of
trolleys
Y1: delay to fill the
19 trolleys (h)
Y2: total load for
operators (h)
Y3: medium
capacity reserve
(h)
Y4: total
waiting time (h)
Y5: total travel time
of operators (h)
E1
random
1
fixed 30.15
30.15
0.00
0.00
12.37
E2
random
2
fixed 15.13
30.17
0.00
0.00
12.37
E3
random
3
fixed 10.15
30.19
0.00
0.00
12.37
E4
random
4
fixed 7.65
30.20
0.45
0.00
12.37
E5
random
5
fixed 6.15
30.21
1.96
0.00
12.37
E6
random
6
fixed 5.14
30.23
2.96
0.00
12.37
E7
random
7
fixed 4.44
30.25
3.68
0.00
12.37
E8
random
8
fixed 3.97
30.27
4.22
0.00
12.37
E9
random
9
fixed 3.61
30.26
4.64
0.00
12.36
E10 random
10
fixed 3.30
30.31
4.90
0.00
12.39
E11 random
11
fixed 3.09
30.33
5.24
0.00
12.37
E12 random
12
fixed 2.91
30.32
5.47
0.00
12.37
E13 random
13
fixed 2.76
30.34
5.67
0.00
12.37
E14 random
14
fixed 2.60
30.37
5.83
0.00
12.37
E15 random
15
fixed 2.51
30.38
5.97
0.00
12.37
E16 random
16
fixed 2.52
30.40
6.10
0.00
12.37
E17 random
17
fixed 2.50
30.41
6.21
0.00
12.37
E18 random
18
fixed 2.50
30.42
6.31
0.00
12.37
E19 random
19
fixed 2.51
30.41
6.40
0.00
12.37
Table 6
Design of experiments of the improved system (DoE 2).
Experiments X1: sequence of
picking
X2: number of
available operators
X3: type of
trolleys
Y1: delay to fill the
19 trolleys (h)
Y2: total load for
operators (h)
Y3: medium
capacity reserve
(h)
Y4: total
waiting time
(h)
Y5: total travel time
of operators (h)
E20 predefined 1
mobile 18.78
18.78
0.00
0.00
1.00
E21 predefined 2
mobile 9.53
18.94
0.00
0.16
1.00
E22 predefined 3
mobile 6.56
19.20
1.60
0.42
1.00
E23 predefined 4
mobile 4.98
19.42
3.14
0.64
1.00
E24 predefined 5
mobile 4.23
20.04
3.98
1.27
1.00
E25 predefined 6
mobile 3.61
20.48
4.57
1.70
1.00
E26 predefined 7
mobile 3.22
20.73
5.04
1.95
1.00
E27 predefined 8
mobile 2.88
21.15
5.36
2.37
1.00
E28 predefined 9
mobile 2.66
21.72
5.58
2.94
1.00
E29 predefined 10
mobile 2.48
21.85
5.82
3.07
1.00
E30 predefined 11
mobile 2.35
21.89
6.01
3.12
1.00
E31 predefined 12
mobile 2.20
22.09
6.15
3.32
1.00
E32 predefined 13
mobile 2.04
22.41
6.26
3.63
1.00
E33 predefined 14
mobile 1.91
23.04
6.35
4.27
1.00
E34 predefined 15
mobile 1.83
24.02
6.39
5.24
1.00
E35 predefined 16
mobile 1.82
24.92
6.44
6.14
1.00
E36 predefined 17
mobile 1.81
25.40
6.51
6.63
1.00
E37 predefined 18
mobile 1.81
25.96
6.56
7.18
1.00
E38 predefined 19
mobile 1.81
26.38
6.61
7.61
1.00
F.Z. Ben Moussa, et al.
Computers & Industrial Engineering 132 (2019) 124-140
135experiment Ei enables the satisfaction of all the evaluation parameters
Yj. To formulate a generalized system of contradictions, two sets of
evaluation parameters that are in contradiction should be determined.
Finding the best system of contradictions involves finding the existing
conflicts between non-dominated solutions. Thus, a dominance analysis
between the different experiments is performed to find the best non
dominated solutions. Indeed, based on the principle of Pareto dom
inance, one solution is better than another if it has at least as good
performance on all objectives and is more successful on at least one
objective. The calculation result of the dominance between the different
experiments is presented in the following table (Table 8). The binary
rank of a solution (or a set of solutions) is equal to the number
Step by Step Solution
There are 3 Steps involved in it
Step: 1
![blur-text-image](https://dsd5zvtm8ll6.cloudfront.net/includes/images/document_product_info/blur-text-image.webp)
Get Instant Access with AI-Powered Solutions
See step-by-step solutions with expert insights and AI powered tools for academic success
Step: 2
![blur-text-image](https://dsd5zvtm8ll6.cloudfront.net/includes/images/document_product_info/blur-subtext-image.webp)
Step: 3
![blur-text-image](https://dsd5zvtm8ll6.cloudfront.net/includes/images/document_product_info/blur-subtext-image.webp)
Ace Your Homework with AI
Get the answers you need in no time with our AI-driven, step-by-step assistance
Get Started