Available online at www.sciencedirect.

com
Available online at www.sciencedirect.com

ScienceDirect
ScienceDirect
Available online atonline
Available www.sciencedirect.com
at www.sciencedirect.com
Procedia CIRP 00 (2019) 000–000
ScienceDirect
Procedia CIRP 00 (2019) 000–000
ScienceDirect
www.elsevier.com/locate/procedia
www.elsevier.com/locate/procedia

Procedia CIRP 00 (2017)

Procedia 000–000
CIRP 84 (2019) 434–441
www.elsevier.com/locate/procedia

29th CIRP Design 2019 (CIRP Design 2019)

29th CIRP Design 2019 (CIRP Design 2019)
Decision
Decision Analysis
Analysis for
28th CIRP
Optimal
for Design
Optimal Design
Design
Conference, May
Concept:
Concept:
2018, Nantes,
Hybridized
Hybridized
France
Fuzzified
Fuzzified
Weighted
Weighted Decision Matrix and Fuzzy TOPSIS Using Design for X
Decision Matrix and Fuzzy TOPSIS Using Design for X Tools
Tools
A new methodology to analyze the functional and physical architecture of
Olayinka Mohammed Olabanjia,b *, Khumbulani Mpofu a
existing products for an
Olayinka assembly
Mohammed oriented
Olabanji a,b
product
*, Khumbulani
Tshwane University of Technology Pretoria West, South Africa
a
family
Mpofua identification
a
Tshwane
b University
Federal of Technology
University Pretoria
of Technology West,
Akure, South Africa
Nigeria
Paul Stief *, Jean-Yves Dantan,
Federal University
b
Alain
of Technology Etienne,
Akure, Nigeria Ali Siadat
* Corresponding author. Tel.: +234 80 385 52 639; E-mail address: obayinclox@gmail.com; omolabanji@futa.edu.ng
* Corresponding author. Tel.: +234 80 385 52 639; E-mail address: obayinclox@gmail.com; omolabanji@futa.edu.ng
École Nationale Supérieure d’Arts et Métiers, Arts et Métiers ParisTech, LCFC EA 4495, 4 Rue Augustin Fresnel, Metz 57078, France

*Abstract
Corresponding author. Tel.: +33 3 87 37 54 30; E-mail address: paul.stief@ensam.eu
Abstract
The need for designs with extended capabilities and customization calls for a robust conceptual design phase of an equipment or a product. It
The needimportant
becomes for designs with extended
to appraise differentcapabilities
design conceptsand customization
using a holisticcalls for a robust
approach for theconceptual
comparison design phase
process of antoequipment
in order obtain an or a product.
optimal design.It
Abstract
becomes important to appraise different design concepts
In this article, a decision analysis for selecting optimalusing designa holistic approach for the
of a Reconfigurable comparison
Assembly process
Fixture (RAF)in order
fromtoa obtain
set of an optimal design.
alternative design
In this article,
concepts a decision
is presented analysis for
by developing selecting optimal
a Multi-Attribute designModel
Decision of a Reconfigurable
(MADM). The AssemblyMADM developedFixture (RAF)
in this from
articleahybridize
set of alternative design
the aggregating
Inconcepts
today’s isbusiness
presented environment,
by developing the atrend towards more
Multi-Attribute product
Decision variety
Model and customization
(MADM). The MADM isdeveloped
unbroken.inDue thistoarticle
this development,
hybridize the the need of
aggregating
strength of the fuzzified weighted decision matrix and computational strength of the fuzzy TOPSIS in terms of separation from positive and
agile and of
strength reconfigurable
the fuzzified production systems matrix
weightedcoefficient
decision emerged to computational
cope with various products theand product families. To design and optimize production
negative solutions and closeness to idealand solution. The design strength
for Xoftools fuzzy TOPSIS
(manufacture, in terms
assembly of
and separation frommaintenance,
disassembly, positive and
systems as well
negative solutions as to choose the optimal product matches, product analysis methods are needed. Indeed, most of the known methods aim to
reliability, life cycleand closeness
cost, coefficient
serviceability to ideal solution.
and environment) The to
are used design for Xthe
represent tools (manufacture,
design features andassembly
factors or and disassembly,
constraints maintenance,
for evaluating the
analyze a product
reliability, or one
lifeseveral
cycle product
cost, family on
serviceability theenvironment)
physical level. areDifferent product families, however, may differ largely in terms of theevaluating
number and
designs with sub-factors in orderand to have a robust decision usedprocess.
to represent
Two the
otherdesign
tools features
added toand the factors
X-toolsorare constraints
design forforreconfigurability the
nature
designs ofwith
components.
several This fact in
sub-factors impedes
order anhave
to efficient
a robustcomparison
decision and choice
process. Two ofother
appropriate
tools product
added to family
the X-tools combinations
are design forreconfigurability
for the production
and functionality. These two tools were added in order to consider the reconfigurable characteristic features for the design concepts of the RAFs.
system.
Results Aobtained
new methodology
and functionality. These
from thetwo istools
proposed
decision were to analyze
added
process existing
in order
proves that to products
consider
there theinreconfigurable
are slight view of their
differences functional
characteristic
in the andfeatures
final values physical forarchitecture.
of the design the concepts The
design conceptsaimcan
which is the
of to
becluster
RAFs.
linked
these products
Results obtained in new
from assembly
the decisionoriented product
process families for the
areoptimization of existing
theassembly lines
of and the creation of future reconfigurable
to the performance of each design concept inproves that there
the decision process slight
and thedifferences
weights or in priority
finalofvalues
the design thefeatures.
design concepts
Also, the which
closecan be linked
variations in
assembly
to the
the final systems.
values of Based
performance of each
the ondesign
design Datum Flow
concept
concepts mayChain,
inbethe
duethetophysical
decision structure
process
inclusion of and of the
the
weights products
weights
of the isattributes
analyzed.
or priority
design of theinFunctional
design subassemblies
features.
computation of Also, theare
close
the distances identified,
tovariations and
in
the positive
aand
functional
the final analysis
values of theis
negative ideal solutions. performed.
design Moreover,
concepts may be a
duehybrid
to functional
inclusion of and
weights physical
of the architecture
design graph
attributes in (HyFPAG)
computation is the
of theoutput which
distances to depicts
the the
positive
similarity
and negative between
ideal product
solutions.families by providing design support to both, production system planners and product designers. An illustrative
example of a nail-clipper is used to explain the proposed methodology. An industrial case study on two product families of steering columns of
© 2019
© 2019 TheThe Authors. Published
Published by by Elsevier
Elsevier B.V.
© 2019 The Authors.
thyssenkrupp
Peer-review
Presta France
Authors.
under
is then
Published
responsibility bycarried outB.V.
of Elsevier
the scientific
scientific
to give a first industrial evaluation of the proposed approach.
B.V. committee of of the
the CIRP
CIRP Design
Design Conference
Conference 2019. 2019
Peer-review
©Peer-review under
2017 The Authors. responsibility of the
Published byofElsevier B.V. committee
under responsibility the scientific committee of the CIRP Design Conference 2019
Peer-review under for
Keywords: Design responsibility of the scientific
X tools, Reconfigurable Assembly committee of theTOPSIS,
Fixture, Fuzzy 28th CIRP Design
Fuzzy Conference
Weighted Decision 2018.
Matrix
Keywords: Design for X tools, Reconfigurable Assembly Fixture, Fuzzy TOPSIS, Fuzzy Weighted Decision Matrix
Keywords: Assembly; Design method; Family identification

1. Introduction each design criteria and group the design features under their
1. Introduction each design criteria
corresponding andThis
criteria. group is the design features
necessary because under
it providestheir
1. Introduction of corresponding
the product range and This
criteria. characteristics
is necessary manufactured
because it and/or
provides
The desire of manufacturers to satisfy the drive from mass a well-structured decision process. The selection of design
The desireofofequipment
manufacturers to satisfy the drive from mass assembled in this system.
a well-structured In this context,
decision theselection
main challenge in
production to users-oriented customization criteria is generally basedprocess.
on theThe of design
design requirements,
Due to of
production theequipment
fast development
to in thecustomization
users-oriented domain of modelling
criteria and
is analysis isbased
generally now on not the
onlydesign
to coperequirements,
with single
requires more tasks and design considerations to be product performance and decision of policy makers [3].
communication
requires more and an and ongoing trend of digitization and products,
producta limited product range or existing product families,
incorporated intotasks
the design phase design of a considerations
product. The inclusion to be Further, performance
the and
design features decision of policy
are grouped under makers
the criteria [3].
digitalization,
incorporated manufacturing
into the design enterprises
phase of a are facing
product. The important
inclusion butFurther,
also to betheable to analyze
design featuresandareto compare
grouped products
under the to criteria
define
of several considerations at the preliminary design stage considering the characteristics or attributes of the design or
challenges
of severalain today’s market
considerations at the environments:
preliminary a continuing
design stage newconsidering
product families. It can be observed
the characteristics that classical
or attributes existing
of the design or
produces robust design but requires the development of functional requirement [4].
tendency
produces towards
a robustreduction
design of product
but requires development
the developmenttimes of and product families
functional are regrouped
requirement [4]. in function of clients or features.
different design concepts. In situations where there are The design for X technique is an approach that provides a
shortened
different product lifecycles.
concepts. itIn
design concepts, Inisaddition,
situations there is anthere
where increasing
are However,
The assembly
design fororiented product
Xoftechnique families
is an are hardly
approach to find.a
different design necessary to decide on the holistic integration the relationship betweenthat provides
design factors
demand
different ofdesign
customization,
concepts, being
it is at the same
necessary to time in aonglobal
decide the On the product family level, productsbetween
differ mainly in two
criteria and design features that will be employed to evaluate in order to improve design flexibility, improved lifefactors
holistic integration of the relationship design cycle
competition
criteria with competitors
andconcepts
design features allwill
over the world.toThis trend, mainin characteristics: (i) the number of components and life(ii)cycle
the
the design in orderthat
to arrive be
at employed
an optimal design evaluate[1, order
cost, to improve
efficiency design
and productivity. flexibility, improved
It enhances the flexibility
which
the is inducing
design concepts the
in development
order to arrive atfrom
an macrodesign
optimal to micro
[1, type of components
cost, efficiency (e.g.productivity.
and mechanical,Itelectrical,
enhances electronical).
the flexibility
2]. In order to have a vigorous and unbiased evaluation of the design process, competitiveness measure, improved
markets, results
2]. In order to in
havediminished
a vigorous lot and
sizesunbiased
due to evaluation
augmenting Classical
of the designmethodologies
process, considering mainly
competitiveness single improved
measure, products
process, it is essential to consider the relative importance of decision making and enhanced operational efficiency due to
product
process,varieties (high-volume
it is essential to consider to low-volume
the relative production)
importance of [1]. or decision
solitary, making
alreadyand existing
enhanced product families
operational analyzeduetheto
efficiency
To cope with this augmenting variety as well as to be able to product structure on a physical level (components level) which
identify
2212-8271possible
© 2019 The optimization
Authors. Published potentials in the existing
by Elsevier B.V. causes difficulties regarding an efficient definition and
2212-8271
Peer-review©under
production 2019responsibility
system,TheitAuthors. Published
is important
of tobyhave
Elsevier
the scientific B.V.of the
a precise
committee knowledge comparison
CIRP Design Conference 2019 of different product families. Addressing this
Peer-review under responsibility of the scientific committee of the CIRP Design Conference 2019

2212-8271©©2017
2212-8271 2019The
The Authors.
Authors. Published
Published by Elsevier
by Elsevier B.V. B.V.
Peer-review
Peer-review under
under responsibility
responsibility of scientific
of the the scientific committee
committee of the of theCIRP
28th CIRP Design
Design Conference
Conference 2019.
2018.
10.1016/j.procir.2019.04.323
 Olayinka Mohammed Olabanji et al. / Procedia CIRP 84 (2019) 434–441 435
2 Author name / Procedia CIRP 00 (2019) 000–000

its concurrent design concepts and various tools used during numerous design criteria and sub features to be considered in
the process [5, 6]. Tools usually associated with the design the concept selection process [10]. In order to solve the
for X technique includes; design for manufacture, design for problem of ambiguity in terms of large data and bias nature
assembly and disassembly, design for reliability, design for of assigning crisp values to alternative design concepts, the
maintainability and lots more. These tools help to integrate MADM employs Triangular Fuzzy Number (TFN) and uses
the decision of the design engineer during the design process different methods to defuzzify the TFNs of the design
and they assist in predicting the life cycle cost and concepts after the selection process [11].
performance measure of a design. In practice, the design for In essence, this article proposes the hybridization of fuzzy
X tools have sub factors that contributes to their relevance in weighted decision matrix and fuzzy TOPSIS by using the
the optimal design. These sub-factors depend on the nature of design for X tools as design features for selecting optimal
the design, design features and criteria and performance design of a reconfigurable assembly fixture (RAF) from a set
requirements [7, 8]. of alternative design concepts. Design for reconfigurability
Several decision-making models have been applied to and design for functionality are added to other design for X
select optimal design from a group of conceptual designs. tools considering the RAF in the present study. The proposed
These models are generally classified as Multiobjetive method explores the computational strength of the fuzzified
decision making models (MODM) and multi attribute weighted decision matrix for aggregating the TFNs of the
decision making models (MADM) as shown in Figure 1. The numerous sub-factors grouped under various X tools and the
MODM such as Weighted Decision Matrix (WDM), fuzzy TOPSIS in terms of separation or distance from
Technique for Order Performance by Similarity to Ideal positive and negative solutions and closeness coefficient to
Solutions (TOPSIS), Analytic Hierarchy Process (AHP) and ideal solution. The sub-factors distributed under each tool are
lots more are usually suitable when little or no consideration expected characteristic performance of the design that can be
is given to the sub-features during the selection process [9]. rated using TFNs obtained from the parts analysis and
Also, the MADM plays an important role when there are morphology of the component parts.

Functional requirements, Users

customized features, management
Multi-Attribute constraints and design guidelines
Design Features/
Decision Making
Criteria
(MADM)
Alternative
Design concepts
Design Sub Multi-Objective
Factors Decision Making
(MODM) Optimal Design
Figure 1. Decision making process for selecting optimal design concept

2. Description of the Hybridized Model  1 l

a − l y − a − l y  l a  ,
In order to simplify the description of the model, a 
 1 y− u y  a u  ,
framework for the processes and flow involved is presented μa ( y ) = 
in Figure 2. Consider n number of design concept Dan that a − u a −u
are selected for decision making with a mixed scenario of 
0 Otherwise
design features, constraints and functional requirements  (1)
which can be expressed by i number of design criteria dci.
1
If each of the design features are further decomposed into (
d M N= ) 3
( x1 − y1 )2 + ( x2 − y2 )2 + ( x3 − y3 )2 (2)
sub-criteria dsub having weights Wdsub of different
dimensions ‘u’ according to the criteria under consideration, The identified criteria or design features can be rated
then in order to compensate for variable dimensions of the according to their level of importance as needed in the
sub-criteria, it is necessary to assign relative importance of optimal design [3]. In order to fuzzify the level of each design
the attributes with fuzzy number A using a triangular fuzzy feature and sub-criteria in the optimal design, membership
number (TFN) which membership function µ a(y) is
functions are allocated to the features to represent the
contained in [0 1] and defined by equation 1 [12, 13].
Where l  a  u and l, a and u represent the lower, modal and weights of the design alternatives. The distribution of TFNs
are tabulated in Table 1. The inclusion of sub-criteria
upper values of the fuzzy number A respectively [14]. Also,
membership functions for all the design alternatives or
if the lower, modal and upper values of two fuzzy numbers
concept will provide a robust selection process because an
m and n are represented as x1, x2 , x3 and y1, y2 , y3 , then the
analysis of the design characteristics of all the concepts
distance between these numbers can be obtained from the would have been considered [17, 18].
vertex method as described in equation 2 [15, 16].
436 Olayinka Mohammed Olabanji et al. / Procedia CIRP 84 (2019) 434–441
Author name / Procedia CIRP 00 (2019) 000–000 3

Start Stop

Fuzzified Weighted Decision Fuzzy TOPSIS

Matrix

✓ Identify all the functional requirements and features required ✓ Determine the closeness coefficient from the
from the optimal design. distances.
✓ Define the design features and criteria for the decision process. ✓ Rank the design concepts based on the magnitude
✓ Establish the basis for determining the level of important of of the calculated closeness coefficient.
the functional requirements in the optimal design.
✓ Determine the linguistic terms to be used for the level of
importance of the design features in the optimal design. ✓ Determine the fuzzy positive and negative ideal
solution.
✓ Estimate the separation of each design concept
✓ Identify all sub-criteria and characteristics features associated from the positive and negative ideal solution
with all the design features. ✓ Determine the cumulative distances for each
✓ Establish the basis for ranking the assessment of the sub- alternative design concept
criteria in all the design concepts.
✓ Determine the linguistic terms for assessment of sub-criteria
in all the design concepts ✓ Normalize the fuzzy decision matrix
✓ Establish TFNs for priority weights of design
features as required in the optimal design.
Determine the TFN to fuzzify the crisp values of the linguistic ✓ Determine the weighted normalized fuzzy
terms for the importance of design features and assessment of decision matrix
sub-criteria in the design concepts.

✓ Obtain aggregate TFNs for the design concepts

Develop comparison matrices for comparing the from the comparison matrices in order to obtain
assessment of the sub-criteria in each of the design the fuzzy decision matrix.
concepts or alternatives using the fuzzy weighted ✓ Establish TFNs for priority weights of design
model. features as required in the optimal design.

Figure 2: Framework of the Hybridized model

Table 1. TFN for Rating and Ranking Design criteria and Sub-criteria
respectively
1
 bsf bsf2 1 .......... bsfs 1 
 1 
Linguistic Terms for Linguistic Terms for Crisp Value Triangular
 b1 bsf2 2 s 
.......... bsf 2
Rating of Relative Ranking of availability of Ranking Fuzzy Scale
Significance of of Sub-criteria features or Rating Membership Bsf =  sf 2  (4)
Design Criteria in the Design concepts Function
 
2.5 3 3.5
 
Highly Important Very High 5 1
 bsfk 2 s 
2 2.5 3  bsfk ...........bsfk 
Important High 4
Where bsf is a TFN that can be represented by (lsf asf usf )
Very Necessary Medium 3 1.5 2 2.5
Necessary Low 2 1 1.5 2
Not Necessary Very Low 1 0.5 1 1.5 as presented in equation 1. For i=1, 2, 3…. k, j=1, 2, 3……s.
Assigning TFNs to the relative significance of the sub- A weighted aggregate of all the sub features under each
features and assessing the availability of the sub-features in design attribute is necessary to provide a basis for
the alternative design concepts based on the parts analysis of comparison using the weights of the relative importance of
each concept will produce a comparison matrix which the design attribute. A matrix of all the weighted aggregate
aggregate will form a basis for the analysis of the relative
importance of the design features. This will produce the
 
of all the sub features W = Wsfij for i number of design

aggregate of each design criteria ( ndci ) and a weight criteria can be represented by;
function of the significance of the sub-features as described  B1 Bsf2 1 .......... Bsfj 1 
in equation 3. In essence, the rating of the sub-features  sf 1 
 1 j 
considering a particular design attribute will be of the form B Bsf2 2 .......... Bsf 2 
of a triangular fuzzy matrix whose judgment matrix Wsf n =  sf 2 (5)
 
s
 
B = bsfk of i define set of design criteria is presented in 
 B1 2 j 

Bsfi ...........Bsfi
equation 4;  sfi 
u =u
ndci =  (Wd sub * bsfu ) (3)
u =1
 4 Olayinka Mohammed
Author Olabanji
name / Procedia et al.
CIRP 00/ (2019)
Procedia CIRP 84 (2019) 434–441
000–000 437

n
Where Bsf is a TFN that is equal to the cumulative aggregate
of all the sub features in a design attribute for a particular
1
i =1 3
 (
di− =   sin , sn− 
 ) (12)
design alternative. In order to normalize the triangular fuzzy
The closeness coefficient ( CCi ) represents the distances of
matrix, consider a fuzzy number yij = (lij aij uij ) for
the design alternatives to the fuzzy positive ideal solution
=(i 1.....
= n j 1.....m ) the normalization process can be
(F+) and fuzzy negative ideal solution (F - ) simultaneously.
represented as; [19 - 21]. This can be obtained from;

( yij )N = (lij )N ( aij )N (uij )N  (6) di−

CCi = (13)
Min Min Min 
di− + di+
 lij − l j aij − l j cij − l j
( y=
ij ) N 
 Min Max
,
Max
,
Max
= , i 1,..n; j  Ωb (7)
In essence, the design alternative with highest closeness
 Min Min 
coefficient represents the optimal design and is closest to the
fuzzy positive ideal solution and farthest from fuzzy negative
 uij − u j Max aij − u j Max lij − u j Max  ideal solution [24].
( ) ij N 
y=
 Max
,
 Max
,
 Max 
=  , i 1,..n; j  Ωc (8)
 Min Min Min  3. Application to the Design of RAF

Where
= l Min
j Min l= Max
ij and u j Max
= uij for i 1,......n ; Four design concept of a reconfigurable assembly fixture
is considered in this article using design for X tools as design
 Max Max
Min = u j − l Min . Also, Ωb and Ωc are sets of benefit and
j features with various sub factors as presented in Figure 3.
cost criteria respectively. In order to simplify the analysis, let The weights of the design features and sub features in the
the normalized weights of each design criteria be (WdA ) . This optimal design are considered from the parts analysis,
can be expressed as a normalized TFN of the form predefined functional requirements of the RAF and expected
performance of the optimal design [22]. In view of this, the
( )
(WdAl ) N , (WdAa )N , (WdAu )N . This will represent the TFNs for each design alternative based on the weights of the
priority weights of the design criteria under consideration. sub features in each design attributes are presented in Tables
Also, the weighted normalized performance value of the nth 2 to 10. The fuzzy decision matrix is obtained from the
alternative in terms of the ith design criteria can be obtained aggregate TFNs of the sub features from each design
in the form of TFNs from the product of the priority weight attributes. Further, the fuzzy decision matrix is normalized
and the normalized weighted aggregate. The fuzzy positive ( applying equations 6-8 in order to arrive at the normalized
F + ) and negative ( F − ) ideal solutions for the design fuzzy decision matrix as presented in Table 11.
alternatives can be obtained from the weighted normalized Considering the normalized performance value of the nth
performance value of the nth alternative in terms of the ith alternative in terms of the ith design feature as presented in
design criteria as presented in equations 9 and 10; Table 11, multiplying the weights of each attributes with the

F + = s1+ , s +2 ,.......sn+  (9) normalized TFNs of the decision matrix will provide the
weighted normalized fuzzy decision matrix as presented in
F− = s −
1
, s −2 ,.......s 
−
n (10) Table 12.
Also, from the weighted normalized fuzzy decision
Where ( sn+ ) and ( sn+ ) are vector TFNs obtained from matrix, the fuzzy positive ( F + ) and negative ( F − ) ideal
solutions for the design alternatives can be obtained from
sn+ = ( e, e, e ) and sn− = ( f , f , f ) respectively. Such that equations 9 and 10 as described in equation 14. Also, the
e = Max
i
''
E=  
ik (for i 1,......
= ''
n and k 1,......j ). Eik is the distances of each design alternatives from the positive and
negative ideal solutions can be derived from equations 2, 11
upper value TFN in the column of the weighted normalized and 12 as shown in Table 13. The closeness coefficients of
performance matrix. Also, f = Min
i
''
Fik'' . Fik is the lower   the design alternatives are obtained from these distances
applying equation 13 as presented in Table 14.
value TFN in the column of the weighted normalized
performance matrix [23]. The distance of each design F − =  0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
(14)
alternative from the positive ideal ( d i+ ) and negative ideal ( F + = 3.50 3.00 3.50 2.50 2.50 2.50 3.50 3.00 3.00
di− ) solution is needed for computation of the relative
closeness of the design alternatives to the optimal design. 4. Discussion of Results
This distance can be obtained by combining equation 2 with
the ideal solutions of equations 9 and 10; Considering the final values of the closeness coefficients
n for the design alternatives, it is evident that design concept
1
di+ =   sin , sn+ 
i =1 3
  ( ) (11) three appears to be optimal design of the RAF. However, it
can be observed that the differences between these values is
not large. The slight difference is due to significant number
of design features used to compare the design alternatives
438 Olayinka Mohammed Olabanji et al. / Procedia CIRP 84 (2019) 434–441
Author name / Procedia CIRP 00 (2019) 000–000 5

with various dimensionless weighted sub features. This is features in the design criteria which reduces the significant
possible because the aggregating method employed by the margin that would have been created by virtue of
fuzzified weighted decision matrix apportioned TFNs to the apportioning TFNs based on the availability of sub features
design alternative based on the importance of each sub in the alternative design concepts.

Design for the Environment (RF) Design for

(EN)
Concept one Optimal Reconfigurability
❖ Safety and Health SH Design ❖ Modularity M
❖ Energy Consumption EC ❖ Integrability I
❖ Ease of use EU ❖ Customization CU
❖ Parts disposal PD ❖ Convertibility CO
❖ Hydraulic effluents HE ❖ Scalability S
❖ Diagnosability D

Design for Reliability (RE) Concept two

❖ Repair frequency and occurrence (MA) Design for Manufacture
RF ❖ Availability of parts AP
❖ Usage Limits UL ❖ Overall cost of manufacturing
❖ Design complexity DC OM
❖ Number off-the-shelf parts NP
❖ Redundancy RD ❖ Manufacturing time MT
❖ Robustness RS ❖ Interchangeability of
component parts IP
Concept three
❖ Parts intricacy PI
Design for Maintainability (MN) ❖ Part material PM
❖ Required Routine maintenance
RM
❖ Downtime maintenance DM Design for Assembly and
❖ Maintenance cost MC Disassembly (AD)
❖ Logistics of replacement parts ❖ Number of connectors NC
LP ❖ Accessibility of locators and
❖ Maintenance frequency and connectors AL
Concept four ❖ Intricacy in arrangement of
occurrence MF
❖ Maintenance safety MS hydraulic components AC
❖ Accessibility of actuators AA
❖ Total assembly and
disassembly time AD
Design for functionality (FU)
❖ Direction of clamping force to
workpiece CW Design for Life Cycle Cost
❖ Stability of locators to (LC)
workpiece SW ❖ Device acquisition and
❖ Number of actuators providing installation costs DA
clamping force AF Design for Serviceability (SE) ❖ System replacement costs
❖ Damage to workpiece DW ❖ Customer service attributes CS SR
❖ Morphology of locators ML ❖ Size of parts SP ❖ Long term repair costs RC
❖ Spatial arrangement of ❖ Labor time LT ❖ Operation cost OC
actuators and locators SL ❖ Service simplification SS ❖ Salvage and disposal costs
❖ Position synchronization of ❖ Special tools for servicing TS SC
actuators PA ❖ Failures caused by the service procedures FP
Figure 3. Application of the model to RAF design using X tools
Table 2. Rating of RAF designs based on Reconfigurability
Design for Reconfigurability (2.5 3 3.5)
Design
M I CU CO S D Aggregate
Concepts
2.5 3 3.5 2 2.5 3 1.5 2 2.5 1.5 2 2.5 2.5 3 3.5 1 1.5 2
Concept 1 2.5 3 3.5 2 2.5 3 1.5 2 2.5 1.5 2 2.5 1.5 2 2.5 1.5 2 2.5 3.3 5.4 7.9
Concept 2 1 1.5 2 2 2.5 3 2 2.5 3 1.5 2 2.5 1.5 2 2.5 2 2.5 3 2.9 4.9 7.4
Concept 3 2 2.5 3 2 2.5 3 2 2.5 3 2 2.5 3 2 2.5 3 2.5 3 3.5 3.9 6.0 8.4
Concept 4 1 1.5 2 2 2.5 3 2.5 3 3.5 2 2.5 3 1.5 2 2.5 1.5 2 2.5 3.3 5.4 8.0

Table 3. Rating of RAF designs based on Functionality

Design for Functionality (2.5 3 3.5)
Design Aggregate
CW SW AF DW ML SL PA
Concepts
2 2.5 3 2 2.5 3 2.5 3 3.5 2.5 3 3.5 2 2.5 3 2.5 3 3.5 2.5 3 3.5
Concept 1 1.5 2 2.5 1.5 2 2.5 1.5 2 2.5 1.5 2 2.5 1.5 2 2.5 1.5 2 2.5 2 2.5 3 3.8 6.2 8.7
Concept 2 2 2.5 3 1.5 2 2.5 1 1.5 2 1 1.5 2 1.5 2 2.5 1 1.5 2 1.5 2 2.5 3.0 5.1 7.7
Concept 3 2 2.5 3 2 2.5 3 2 2.5 3 2 2.5 3 2.5 3 3.5 2 2.5 3 2.5 3 3.5 4.8 7.1 9.8
Concept 4 2 2.5 3 1.5 2 2.5 1.5 2 2.5 2 2.5 3 2 2.5 3 2 2.5 3 2 2.5 3 4.3 6.6 9.4

Table 4. Rating of RAF designs based on Assembly and Disassembly

Design for Assembly and Disassembly (2 2.5 3)
Design
NC AL AC AA AD Aggregate
Concepts
1.5 2 2.5 2 2.5 3 1.5 2 2.5 2.5 3 3.5 2.5 3 3.5
Concept 1 2.5 3 3.5 1.5 2 2.5 2 2.5 3 1.5 2 2.5 2 2.5 3 3.7 5.9 8.6
Concept 2 0.5 1 1.5 1 1.5 2 2 2.5 3 1.5 2 2.5 1.5 2 2.5 2.7 4.6 7.0
Concept 3 2 2.5 3 2 2.5 3 2.5 3 3.5 1.5 2 2.5 2 2.5 3 3.9 6.2 8.9
Concept 4 1.5 2 2.5 1.5 2 2.5 2 2.5 3 2 2.5 3 2 2.5 3 3.7 5.8 8.5
 6 Author
Olayinka name / Procedia
Mohammed CIRP
Olabanji 00/ (2019)
et al. 000–000
Procedia CIRP 84 (2019) 434–441 439

Table 5. Rating of RAF designs based on Manufacturing

Design for Manufacture (2.5 3 3.5)
Design
AP OM MT IP PI PM Aggregate
Concepts
2.5 3 3.5 2 2.5 3 2 2.5 3 1.5 2 2.5 2 2.5 3 2.5 3 3.5
Concept 1 2.5 3 3.5 1.5 2 2.5 1.5 2 2.5 1 1.5 2 2 2.5 3 2 2.5 3 3.8 6.0 8.6
Concept 2 2.5 3 3.5 2 2.5 3 2.5 3 3.5 2.5 3 3.5 2.5 3 3.5 2 2.5 3 4.8 7.3 10.3
Concept 3 2 2.5 3 2 2.5 3 1 1.5 2 1 1.5 2 1.5 2 2.5 1.5 2 2.5 3.3 5.1 7.3
Concept 4 2.5 3 3.5 1.5 2 2.5 2 2.5 3 1 1.5 2 2 2.5 3 2 2.5 3 4.0 6.2 8.9

Table 6. Rating of RAF designs based on Maintainability

Design for Maintainability (1.5 2 2.5)
Design Aggregate
RM DM MC LP MF MS
Concepts
2 2.5 3 2 2.5 3 2.5 3 3.5 2.5 3 3.5 2 2.5 3 1.5 2 2.5
Concept 1 1.5 2 2.5 1.5 2 2.5 2 2.5 3 1 1.5 2 2 2.5 3 1 1.5 2 3.2 5.2 7.8
Concept 2 1 1.5 2 2 2.5 3 1.5 2 2.5 1.5 2 2.5 1.5 2 2.5 1 1.5 2 3.0 5.0 7.5
Concept 3 2 2.5 3 1 1.5 2 2 2.5 3 1 1.5 2 2 2.5 3 2 2.5 3 3.5 5.5 7.8
Concept 4 1 1.5 2 1.5 2 2.5 1 1.5 2 2 2.5 3 2 2.5 3 1 1.5 2 3.0 5.0 7.5

Table 7. Rating of RAF designs based on Reliability

Design for Reliability (1.5 2 2.5)
Design Aggregate
RF UL DC NP RD RS
Concepts
2.5 3 3.5 2 2.5 3 2.5 3 3.5 1.5 2 2.5 1 1.5 2 1.5 2 2.5
Concept 1 1 1.5 2 2 2.5 3 1.5 2 2.5 1.5 2 2.5 2.5 3 3.5 1.5 2 2.5 2.9 4.9 7.4
Concept 2 1 1.5 2 1.5 2 2.5 1.5 2 2.5 2 2.5 3 1 1.5 2 1 1.5 2 2.5 4.5 6.9
Concept 3 1.5 2 2.5 2.5 3 3.5 2 2.5 3 1.5 2 2.5 2 2.5 3 1.5 2 2.5 3.5 5.5 7.8
Concept 4 1.5 2 2.5 2 2.5 3 1 1.5 2 1.5 2 2.5 1.5 2 2.5 2.5 3 3.5 3.0 5.0 7.5

Table 8. Rating of RAF designs based on Serviceability

Design for Serviceability (1.5 2 2.5)
Design Aggregate
CS SP LT SS TS FP
Concepts
1.5 2 2.5 2 2.5 3 1 1.5 2 2 2.5 3 1.5 2 2.5 1 1.5 2
Concept 1 1.5 2 2.5 1.5 2 2.5 2 2.5 3 2 2.5 3 1 1.5 2 2.5 3 3.5 2.5 4.4 6.8
Concept 2 1.5 2 2.5 2 2.5 3 2.5 3 3.5 1 1.5 2 1 1.5 2 1.5 2 2.5 2.3 4.1 6.4
Concept 3 1.5 2 2.5 1 1.5 2 1 1.5 2 2.5 3 3.5 1 1.5 2 2 2.5 3 2.5 4.1 6.1
Concept 4 1.5 2 2.5 1.5 2 2.5 2 2.5 3 2 2.5 3 1 1.5 2 2.5 3 3.5 2.5 4.4 6.8

Table 9. Rating of RAF designs based on Environment

Design for Environment (2 2.5 3)
Design Aggregate
SH EC EU PD HE
Concepts
2.5 3 3.5 2 2.5 3 2.5 3 3.5 1 1.5 2 1.5 2 2.5
Concept 1 2.5 3 3.5 2 2.5 3 1.5 2 2.5 2 2.5 3 1.5 2 2.5 3.7 5.8 8.5
Concept 2 1 1.5 2 2.5 3 3.5 1 1.5 2 2 2.5 3 2 2.5 3 3.0 5.1 7.6
Concept 3 2.5 3 3.5 1 1.5 2 2.5 3 3.5 2 2.5 3 1 1.5 2 3.6 5.7 8.3
Concept 4 1.5 2 2.5 2 2.5 3 1.5 2 2.5 2 2.5 3 2.5 3 3.5 3.5 5.6 8.3

Table 10. Rating of RAF designs based on Life Cycle Cost

Design for Life Cycle Cost (2 2.5 3)
Design Aggregate
DA SR RC OC SC
Concepts
2.5 3 3.5 1.5 2 2.5 2 2.5 3 1.5 2 2.5 2.5 3 3.5
Concept 1 1.5 2 2.5 1 1.5 2 2.5 3 3.5 2 2.5 3 2 2.5 3 3.5 5.6 8.2
Concept 2 2.5 3 3.5 1 1.5 2 1.5 2 2.5 1.5 2 2.5 1.5 2 2.5 3.1 5.1 7.6
Concept 3 1.5 2 2.5 1.5 2 2.5 1.5 2 2.5 2.5 3 3.5 2.5 3 3.5 3.7 5.8 8.5
Concept 4 2 2.5 3 1.5 2 2.5 2.5 3 3.5 2 2.5 3 1.5 2 2.5 3.6 5.8 8.4

Table 11. Normalized decision matrix considering the sub-features of design for X tools
Alternative Design Concepts
Design for X Tools
Concept 1 Concept 2 Concept 3 Concept 4
Reconfigurability 0.07 0.46 0.91 0.00 0.36 0.82 0.18 0.56 1.00 0.07 0.46 0.93
Assembly and Disassembly 0.16 0.52 0.95 0.00 0.31 0.69 0.19 0.57 1.00 0.16 0.50 0.94
Manufacturing 0.07 0.39 0.76 0.21 0.57 1.00 0.00 0.26 0.57 0.10 0.41 0.80
Maintainability 0.04 0.46 1.00 0.00 0.42 0.94 0.10 0.52 1.00 0.00 0.42 0.94
Reliability 0.08 0.45 0.93 0.00 0.38 0.83 0.19 0.57 1.00 0.09 0.47 0.94
Serviceability 0.04 0.47 1.00 0.00 0.40 0.91 0.04 0.40 0.84 0.04 0.47 1.00
Functionality 0.12 0.47 0.84 0.00 0.31 0.69 0.27 0.60 1.00 0.19 0.53 0.94
Environment 0.13 0.51 1.00 0.00 0.38 0.84 0.11 0.49 0.96 0.09 0.47 0.96
Life Cycle Cost 0.07 0.46 0.94 0.00 0.37 0.83 0.11 0.50 1.00 0.09 0.50 0.98

Additionally, a further consideration of the relative stated that the application of hybridized multicriteria
importance of the design criteria in the optimal design will decision making models to selection of optimal design will
also try to nullify the effect of inappropriate apportioning that reduce the appearance of inappropriate apportioning to a
may have occurred during the sub-feature’s comparison with particular alternative design that may have occurred due to
the design alternatives. In essence, it can be hypothetically availability of the desirable sub-feature in the concept. The
440 Olayinka Mohammed Olabanji et al. / Procedia CIRP 84 (2019) 434–441
Author name / Procedia CIRP 00 (2019) 000–000 7

separation distances of the design concepts from the positive consideration. Hence, this relative comparison tends to
and negative ideal solutions computed by the fuzzy TOPSIS reduce or eliminate the possibility of having design concepts
cannot be ruled out as a reason for the closeness of the with splendid values.
coefficients. This is possible because these distances
compare a design concept based on the availability of design
features or criteria present in it relative to another concept
that is optimally identified for the design feature under
Table 12. Weighted Normalized decision matrix considering relevance of design for X tools
Alternative Design Concepts
Design for X Tools
Concept 1 Concept 2 Concept 3 Concept 4
Reconfigurability 0.18 1.38 3.19 0.00 1.08 2.87 0.45 1.68 3.50 0.18 1.38 3.26
Assembly and Disassembly 0.32 1.30 2.85 0.00 0.78 2.07 0.38 1.43 3.00 0.32 1.25 2.82
Manufacturing 0.18 1.17 2.66 0.53 1.71 3.50 0.00 0.78 2.00 0.25 1.23 2.80
Maintainability 0.06 0.92 2.50 0.00 0.84 2.35 0.15 1.04 2.50 0.00 0.84 2.35
Reliability 0.12 0.90 2.33 0.00 0.76 2.08 0.29 1.14 2.50 0.14 0.94 2.35
Serviceability 0.06 0.94 2.50 0.00 0.80 2.28 0.06 0.80 2.10 0.06 0.94 2.50
Functionality 0.30 1.41 2.94 0.00 0.93 2.42 0.68 1.80 3.50 0.48 1.59 3.29
Environment 0.26 1.28 3.00 0.00 0.95 2.52 0.22 1.23 2.88 0.18 1.18 2.88
Life Cycle Cost 0.14 1.15 2.82 0.00 0.93 2.49 0.22 1.25 3.00 0.18 1.25 2.94

Table 13. Distances of the design alternatives to the positive and negative ideal solutions
Design Design for X Tools Cumulative
Alternatives RF AD MA MN RE SE FU EN LC Distances
d + (Concept 1, F +) 2.28 1.83 2.39 1.68 1.66 1.67 2.23 1.87 1.97 17.58
d + (Concept 2, F +) 2.48 2.22 2.00 1.73 1.78 1.75 2.58 2.12 2.13 18.79
d + (Concept 3, F +) 2.05 1.76 2.70 1.60 1.50 1.73 1.90 1.91 1.90 17.05
d + (Concept 4, F +) 2.28 1.85 2.33 1.74 1.64 1.67 2.07 1.94 1.94 17.46
d - (Concept 1, F -) 2.01 1.82 1.68 1.54 1.44 1.54 1.89 1.89 1.76 15.57
d - (Concept 2, F -) 1.77 1.28 2.27 1.44 1.28 1.39 1.49 1.55 1.53 14.00
d - (Concept 3, F -) 2.26 1.93 1.24 1.57 1.59 1.30 2.31 1.81 1.88 15.89
d - (Concept 4, F -) 2.04 1.79 1.77 1.41 1.46 1.54 2.11 1.80 1.82 15.74

Table 14. Closeness Coefficient CCi and Ranking of Design Alternatives

Design
d +
d− CCi Ranking
Alternatives
Concept 1 17.58 15.57 0.470 3rd
Concept 2 18.79 14.00 0.427 4th
Concept 3 17.05 15.89 0.482 1st
Concept 4 17.46 15.74 0.474 2nd

5. Conclusion

Selection of optimal design concept from a set of further carry out a comparison process by evaluating the
alternative design concepts is necessary before carrying out distances and closeness coefficients of each design
a detail design because it assists in ascertaining that desirable alternative from an ideal optimal design. In essence, it can be
characteristics features are considered in the optimal design. concluded that the developed hybridized model can be
Design for X tools such as design for manufacture and design applied to decision analysis for determining optimal design
for assembly and disassembly, will assist the designer to concepts from a set of alternative design concepts.
predict probable manufacturing obstacles that may be
encountered and solutions would have been produce from the Acknowledgements
design stage. The design for reconfigurability and design for
functionality introduced in this article are characteristics The authors of this article wish to appreciate the Tshwane
features that can be linked to the performance of the RAF. University of Technology Pretoria West South Africa and the
Other design tools such as serviceability, maintainability, Gibela Research Chair in Manufacturing and skills
reliability, environment and life cycle cost usually assist in Development in the Faculty of Engineering and the Built
predicting the performance of the design in order to ascertain Environment.
its economic viability. The fuzzified weighted decision
matrix have provided a means of solving the hierarchical References
nature of the problem by aggregating the dimensionless sub-
features that can be linked to the design criteria. The fuzzy [1] Olabanji, O.M. ‘Reconnoitering the suitability of fuzzified weighted
TOPSIS model normalized the aggregated TFNs obtained decision matrix for design process of a reconfigurable assembly fixture’, Int.
J. Design Engineering 2018; 8(1); p. 38–56.
from the fuzzified weighted decision matrix. It considered
[2] Nikander, Jan B, Liikkanen, Lassi A and Laakso, Miko. The preference
the relevance of the design criteria in the optimal design and effect in design concept evaluation. Design studies 2014; 35(5); p. 473-499
 Olayinka Mohammed Olabanji et al. / Procedia CIRP 84 (2019) 434–441 441
8 Author name / Procedia CIRP 00 (2019) 000–000

[3] Olabanji, O.M. and Mpofu, K. Comparison of weighted decision matrix, in a supply chain’, International Journal of Production Economics. 2009;
and analytical hierarchy process for CAD design of reconfigurable assembly 121(2) p.323–332
fixture’, Conference on Assembly Systems and Technologies - CIRP [15] Triantaphyllou, E. and Lin, C-T. Development and evaluation of five
Procedia 23, 2014; p.264–269. fuzzy multiattribute decision-making methods, International Journal of
[4] Cash P, Stankovic T and Storga M. Evaluation of Empirical Design Approximate Reasoning. 1996; 14(1); p.281–310.
Studies and Metrics in Experimental Design Research Approaches, [16] Chen, C-T., Lin, C-T. and Huang, S-F. A fuzzy approach for supplier
Perspectives, applications (Eds.). Springer International Publishing evaluation and selection in supply chain management, International Journal
Switzerland. 2016; p. 13-39. ISBN 978-3-319-33779-1. of Production Economics. 2006; 102(2); p.289–301.
[5] Ming C. C and Gul E O. K. Investigation of the applicability of Design [17] Wang, J. Ranking engineering design concepts using a fuzzy
for X tools during design concept evolution: a literature review. outranking preference model, Fuzzy Sets and Systems. 2001; 119(1); p.161–
International Journal of Product Development. 2011; 13(2); p. 132-167 170.
[6] Uwe Dombrowski, Stefan Schmidt, Kai Schmidtchen 2014. Analysis [18] Wang, J. Improved engineering design concept selection using fuzzy
and integration of Design for X approaches in Lean Design as basis for a life sets’, International Journal of Computer Integrated Manufacturing. 2002;
cycle optimized product design. 21st CIRP Conference on Life Cycle 15(1); p.18–27.
Engineering Elsevier. 385-390
[19] Mokhtarian M.N. A new fuzzy weighted average (FWA) method
[7] Jari Lehto, Janne Harkonen, Harri Haapasalo, Pekka Belt, Matti based on left and right scores: An application for determining a suitable
Mottonen and Pasi Kuvaja. Benefits of DfX in Requirements Engineering. location for a gas oil station; Computers and Mathematics with Applications
Technology and Investment Scientific Research. 2011; 2(1); p. 27-37 2011; 61; p. 3136–3145.
[8] Tsai C. Kuo, Samuel H. Huang and Hong C. Zhang. Design for [20] Mir. B. Aryanezhad, M.J. Tarokh, M.N. Mokhtarian and F. Zaheri. A
manufacture and design for ‘X’: Concepts, applications and perspectives. Fuzzy TOPSIS Method Based on Left and Right Scores; International
Computers and Industrial Engineering Elsevier. 2001; 41(1); p. 241-260 Journal of Industrial Engineering & Production Research. 2011; 22(1); p.
[9] Hambali, A, Sapuan, SM, Rahim, AS, Ismail, N and Nukman, Y. 51-62.
Concurrent decisions on design concept and material using analytical [21] Mokhtarian, M.N. and Vencheh A. Hadi. A new fuzzy TOPSIS
hierarchy process at the conceptual design stage. Concurrent Engineering, method based on left and right scores: An application for determining an
2011; 19(2); p. 111-121. industrial zone for dairy products factory; Applied Soft Computing Elsevier.
[10] Ho, William, Xu, Xiaowei and Dey, Prasanta K. Multi-criteria 2012; 12; p. 2496–2505
decision making approaches for supplier evaluation and selection: A [22] Olabanji, O., Mpofu, K. and Battaïa, O. Design, simulation and
literature review. European Journal of operational research, 2010; 202(1); p. experimental investigation of a novel reconfigurable assembly fixture for
16-24. press brakes, The International Journal of Advanced Manufacturing
[11] Okudan, Gul E and Tauhid, Shafin. Concept selection methods–a Technology, Springer. 2016; 82(1); p. 663–679.
literature review from 1980 to 2008. International Journal of Design [23] Ayda E. D, Prisa B and Mohammad F. An Integrated Fuzzy AHP and
Engineering 2008; 1(3); p. 243-277. Fuzzy TOPSIS Approach for Ranking and Selecting the Chief Inspectors of
[12] Jaini, N.I. and Utyuzhnikov, S.V. A fuzzy trade-off ranking method Bank: A Case Study. J. Appl. Res. Ind. Eng. 2017; 4(1); p. 8–23
for multi-criteria decision-making, Axioms, 2017; 7(1); p.1-21 [24] Baral. A, and Behera. S. Application of Fuzzy TOPSIS to Agricultural
[13] Shameli-Sendi, A., Shajari, M., Hassanabadi, M., Jabbarifar, M. and Farm for Optimum Allocation of Different Crops’, International Journal of
Dagenais, M. Fuzzy multi-criteria decision-making for information security Science and Research 2015; 6(2) p. 2059-2064
risk assessment, The Open Cybernetics & Systemics Journal, 2012; 6(1); p.
26-37.
[14] Amid, A., Ghodsypour, S. and O’brien, C. A weighted additive fuzzy
multiobjective model for the supplier selection problem under price breaks