Fuzzy Logic Framework for Qualitative Evaluation of Supply Chain Responsiveness

The International Journal Of Engineering And Science (IJES)
|| Volume || 4 || Issue || 8 || Pages || PP -37-48|| 2015 ||
ISSN (e): 2319 – 1813 ISSN (p): 2319 – 1805
www.theijes.com The IJES Page 37
Fuzzy Logic Framework for Qualitative Evaluation of Supply
Chain Responsiveness
1
Amina S. Omar, 2
Prof. Mwangi Waweru, PhD, and 3
Dr. Richard Rimiru, PhD
1
Jomo Kenyatta University of Agriculture and Technology, Mombasa Campus,
Institute of Computer Science and Information Technology,
P. O. Box 94090–80107, Mombasa, Kenya,
2,3
Jomo Kenyatta University of Agriculture and Technology, Main Campus,
Institute of Computer Science and Information Technology,
P. O. Box Juja, Nairobi, Kenya,
--------------------------------------------------------ABSTRACT-----------------------------------------------------------
Fuzzy logic can be a powerful tool for managers to use instead of traditional mathematical models when
measuring the of supply chains responsivess. The flexibility of the model allows the decision maker to introduce
vagueness, uncertainty, and subjectivity into the evaluation system. Responsiveness measurement represents a
critically important decision that often involves subjective information. Fuzzy logic models provide a
reasonable solution to these common decision situations. After extensive exploration of the literature, we
recommend an outcome of developing a Fuzzy logic framework in measuring qualitative aspects of supply chain
responsiveness. In this paper, responsiveness as one of the important factors of measuring qualitative
performance is discussed and a fuzzy logic framework is developed to measure supply chain responsiveness.
Keywords - Supply Chain, Performance Measurement, qualitative measures and fuzzy logic
-------------------------------------------------------------------------------------------------------------------------------------------
Date of Submission: 13 July 2015 Date of Accepted: 27 August 2015
-------------------------------------------------------------------------------------------------------------------------------------------
I. INTRODUCTION
Conventional evaluation systems are representatives of structured systems that employ quantifiable and
non-quantifiable measures of evaluation. It is often difficult to quantify performance dimensions. For example,
“responsiveness” may be an important part of evaluating performance of supply chains. However, how exactly
does one measure “responsiveness”. Fuzzy approach can be effectively utilized to handle imprecision and
uncertainty [2]. This approach to performance evaluation allows the organization to exercise professional
judgment in evaluating its supply chains. In real problems, performance evaluation techniques engage in
handling cases like subjectivity, fuzziness and imprecise information.
It is often difficult to quantify performance dimensions because all critical parameters in a Supply Chain
Management are indicated subjectively by linguistic terms and are characterized by ambiguity [3] .Fuzzy set
theory is primarily concerned with quantifying and reasoning using natural language in which many words have
ambiguous meanings. Application of the fuzzy set theory in evaluation systems can improve evaluation results
[14].
The performance measurement process has evolved since the mid-eighties. Performance measures
provide the necessary feedback for management which assists in business decisions [14]. Models in the past have
only explored limited dimensions of supply chain performance such as cost [4], and flexibility [14]. Many
performance measures have been identified as appropriate for supply chain analysis, but have not yet been used
in supply chain modeling research, although these measures may be important characteristics of a supply chain,
their use in supply chain models is challenging, since the qualitative nature of such measures makes them
difficult to incorporate into quantitative models [15]. It is often difficult to quantify performance dimensions
because all critical parameters in a Supply Chain Management are indicated subjectively by linguistic terms and
are characterized by ambiguity [5] .Fuzzy set theory is primarily concerned with quantifying and reasoning using
natural language in which many words have ambiguous meanings. Supply chain performance extent can be
attributed as a function of multiple criteria/attributes. Most of the criterions/attributes being intangible in nature;
supply chain performance appraisement relies on the subjective judgment of the decision-makers [15].
Moreover, quantitative appraisement of supply chain performance appears very difficult due to involvement of
ill defined (vague) performance measures as well as metrics [15].

Fuzzy Logic Framework for Qualitative…
A feature typical of the natural language, to be in no way circumvented, is the vagueness of its
semantics. That is why a description delivered in the natural language cannot be translated directly into
mathematical formulas [6]. To be able to apply the classical mathematics, we have to have the task described
in precise figures. This method, however, can return unsatisfactory results, as precise figures often do not
properly reflect the reality. Fuzzy logic offers a solution to the problem, since it allows us to model the meanings
of words used in the natural language [7].
Fuzzy logic is, however, not fuzzy. Basically, fuzzy logic is a precise logic of imprecision and approximate
reasoning [6]. More specifically, fuzzy logic may be viewed as an attempt at formalization/mechanization of two
remarkable human capabilities. First, the capability to converse, reason and make rational decisions in an
environment of imprecision, uncertainty, incompleteness of information, conflicting information, partiality
of truth and partiality of possibility; in short, in an environment of imperfect information. And second, the
capability to perform a wide variety of physical and mental tasks without any measurements and any
computations [17].
Reality has almost always an aspect of randomness and an aspect of vagueness. The mathematical apparatus of
the theory of fuzzy sets provides a natural basis for the theory of possibility, playing a role which is similar to
that of measure theory in relation to the theory of probability [17]. Vagueness can be modeled using the theory
of fuzzy sets, while the randomness is modeled with reliance on the probability theory and possibly other
theories like the theory of possibility, different rates of veracity, etc. [18]. Viewed in this perspective, a fuzzy
restriction may be interpreted as a possibility distribution, with its membership function playing the role of a
possibility distribution function, and a fuzzy variable is associated with a possibility distribution in much the
same manner as a random variable is associated with a probability distribution [17].
Fuzzy provides a remarkably simple way to draw definite conclusions from vague, ambiguous or imprecise
information. In a sense, fuzzy logic resembles human decision making with its ability to work from approximate
data and find precise solutions. [18] set up the following steps which are necessary for successful application of
modeling through a general fuzzy system: Fuzzification of the input and output variable by considering
appropriate linguistic subsets, Construction of rules based on expert knowledge and/or the basis of available
literature, The result appears as a fuzzy subset and therefore, it is necessary to defuzzify the output and obtain a
crisp output.
I. THEORETICAL LITERATURE REVIEW
2.1 Qualitative Performance Measurement of Supply Chain Management
Performance measurement is the process of using a tool or a procedure to evaluate a concrete efficiency
parameter of the system The traditional performance measurement systems evaluate quantitative indicators
directly related to production parameters: throughput, number of delayed orders, WIP, manufacturing lead time,
etc. The problem is how to evaluate the performance of the systems in the presence of unexpected changes.
Here, performance indicators may be of a qualitative nature, since they usually reflect subjective views of the
expected behavior of the systems in those circumstances. In the field of performance measurement, inevitability
of subjectivity has to an extent been accepted. They recognized that elimination of judgmental criteria and their
associated subjectivity are unlikely. This suggests that many performance evaluation factors are subjective, and
hence qualitative in nature.
According to [15] SC performance measures are categorized in three main types; Resource, Output and
Flexibility and declared that output measures include customer responsiveness, quality and quantity of the final
product produced. Some of the output measures can be measured numerically such as number of items produced
but some of them such as customer satisfaction, responsiveness and product quality cannot be measured
numerically.
In his new classification, [15] introduced performance measures of SCM. They divided all the metrics into
quantitative and qualitative and then established sub factors for each category. Qualitative category is divided
into quality, flexibility, visibility, trust and innovativeness [15]. They claimed that quality factor of the
mentioned criteria have a full picture of the mentioned criteria which should be interpreted in the SCM
qualitative performance. Quality is categorized into customer satisfaction, customer response time, lead time, on
time delivery, fill rate, stock-out probability and accuracy.
Another performance measurement classification of agri-food was presented by [15] which include
efficiency, flexibility, responsiveness, and food quality. Responsiveness classified in fill rate, product lateness,
customer response time, lead time, customer complaints, shipping errors. Supply Chain Operations Reference
(SCOR) model presented the following five attributes of Supply Chain performance [8].

1. SC reliability. The performance of the SC in delivering the correct product to the correct place, at the correct
time, in the correct condition and packaging, in the correct quantity, with the correct documentation, to the
correct customer which is Perfect Order Fulfillment. Reliability is a customer-focused attribute.
2. Supply Chain responsiveness: The speed at which a Supply Chain provides products to the customer.
Responsiveness classified in fill rate, product lateness, customer response time, lead time, customer complaints,
shipping errors [15].
3. Supply Chain flexibility: The ability to respond to marketplace changes to gain or maintain competitive
advantage. Flexibility is presented into 4 categories: volume flexibility, delivery flexibility, mix flexibility and
new product flexibility [19]. Flexibility was considered to be a qualitative factor [15].
4. Supply Chain costs: The costs associated with operating the SC. Cost is one of the quantitative measures [15]
and it can be measured by distribution cost, manufacturing cost, inventory cost, warehouse cost, incentive cost
and subsidy, intangible cost, overhead cost and sensitivity to long-term cost. All these are quantitative measures.
5. SC asset management: The ability to efficiently utilize assets. Metrics include: inventory days of supply and
capacity utilization which is a quantitative measure.
Literature reveals that considerable amount of work has been carried out by pioneer researchers towards
performance measurement metrics, which are categorized into qualitative and quantitative. This research will
only look at qualitative aspects of performance.
III.FUZZY LOGIC FRAMEWORK DEVELOPMENT METHODOLOGY
Qualitative performance measurement evaluation of Supply Chain management for was done in five steps.
Following the process of methodology will be described.
The first step was designing the questionnaire for gathering true required information from various
supply chains in Kenya. The information was used to construct the proposed framework of performance
measurement in Supply Chain Management. The answers contain the most interesting criteria which were used
as input variable of the framework. Moreover, the questionnaire was designed in a way that it would help the
researcher to define the correct membership function and fuzzy rules based on the range of answers. Therefore,
role of the questionnaire in the research is to obtain the robust data of the Supply Chain Management which can
be applied in the proposed Fuzzy Logic Framework.
Questionnaire was designed in two sections. The first section gathered general respondents information
and information on what extent do they measure qualitative aspects of performance of supply chains based on
the predetermined factors. The second section gathered information on responsiveness and reliability of supply
chain performance. The correct range of data for each criterion was gathered through this questionnaire the
responses entered in the FLC design module.
The next step in the methodology was to fuse two variables at a time depending on their importance, in
order to reduce the rule base. Jamshidi [10] proposed to use sensory fusion to reduce a rule base size. This
method consists in combining variables before providing them to input of the fuzzy controller [11]. In this study
the variables were fused as shown. In the hierarchical fuzzy control structure from [12], the first-level rules are
those related to the most important variables and are gathered to form the first-level hierarchy. The second most
important variables, along with the outputs of the first-level, are chosen as inputs to the second level hierarchy,
and so on. For Supply Chain Responsiveness, the first fusion will be between on time delivery and response
time. On time delivery measures the percentage of all orders delivered by the requested delivery date, as
indicated in the PO/contract during a defined period of time. Logistics managers can use this indicator to
monitor supplier response time on shipments over a specified period of time [13].
The second fusion will be on lead time and probability out of stock. Lead time is the time of a supply
chain network to respond to customer demands. Furthermore, in the worst case lead time corresponds to the
response time when there are zero inventories. This was used as a measure of responsiveness in our previous
work [14]
The third fusion was on Fill rate and shipment errors. This indicator measures the ability of the supplier to fill
POs correctly. Shipments should always be checked against the shipping notice and the PO. What was shipped
may not be what was ordered.
The fourth fusion will be between accuracy and number of complaints. Accuracy and customer
complaints are more sensitive compare to fill rate and customer response time because when the company cannot
provide ordered products accurately, it cause to the dissatisfaction and reduce output performance. Mentioned
problem may not always cause to the complaints because customers prefer to change the company and provide
the requirement from the competitors [15]. This is shown in figure.3.1 to figure 3.7

Figure 3.1: Basic FIS Editor in Matlab showing accuracy and number of complaints parameters fused
linearly to produce Rep1 into a fuzzy model.
Figure 3.2: Basic FIS Editor in Matlab showing lead time and probability out of stock parameters fused
linearly to produce Resp2 into a fuzzy model.
Figure 3.3: Basic FIS Editor in Matlab showing on time delivery and response time parameters fused linearly
to produce Resp3 into a fuzzy model.
Figure 3.4: Basic FIS Editor in Matlab showing fill rate and shipping errors parameters fused linearly to
produce Reps4 into a fuzzy model.

Figure 3.5: Basic FIS Editor in Matlab showing RESP1 and RESP2 outputs fused linearly to produce
RESPFUSIONONE into a fuzzy model.
Figure 3.6: Basic FIS Editor in Matlab showing RESP3 and RESP4 outputs fused linearly to produce
Respfusiontwo into a fuzzy model.
Figure 3.7: Basic FIS Editor in Matlab showing RESFUS1and REFUSION2 outputs fused linearly to
produce RESPONSIVENESS into a fuzzy model.
The next step in the methodology was FLC design which includes membership function design and
fuzzy rules design. In membership function design, factors divided into two groups; numerically and
proportionally, according to the factors definitions. Using natural language, commonsense linguistic labels and
traditional crisp set values, the numerical group will be measured as “high” or “low”. While the proportional
group will be measured as “poor” and “good”. A fuzzy set is an extension of a crisp set which allows partial
membership. The permissiveness of fuzziness in the human thought process suggests that much of the logic
behind thought processing is not traditional two valued logic or even multivalued logic, but logic with fuzzy
truths. Therefore a partial membership label “Medium” will be added in the linguistic term set for the numerical
group and “Average “ for the proportional group of the performance measurement variables.
Certain operators may be included to slightly change the meaning of the linguistic labels involved in a
specific linguistic fuzzy rule. It a way to do so with a minor description loss is to use linguistic hedges [6]. A
linguistic hedge (also known as linguistic modifier) is a function that alters the membership function of the fuzzy
set associated to the linguistic label, obtaining a definition with a higher or lower precision depending on the
case. Two of the most well-known modifiers are the concentration linguistic hedge “very” and the dilation
linguistic hedge “more-or-less.” and their effects on a triangular membership function are shown in Fig. 3.11
Figure 3.11.Effects of the linguistic Hedges “very” and “more-or-less”

The linguistic terms will then be quantified on a numerical scale based on inputted numerical data
collected from various Supply Chain Management system and membership functions will then be defined to
determine their degree of membership. The triangular membership function shape will be chosen because it is
most popular in the performance measurement [3*]. The above linguistic variables and terms will then be
matched and fuzzy rules generated, and expressed in a fuzzy associative matrix and output results obtained for
each parameter which will then be aggregated into one crisp value using a defuzzication technique.
Range of the membership function depended on the alternatives that were designed for each criterion in
the questionnaire. For instance alternatives which were designed for lead time has four alternatives; less than 1
day, 1 to 3 day, 3 days , 3 to 5 days and more than 5 days. First alternative in triangular membership function
should be covered range of data which are less than 1 day. Therefore, four alternatives cover four types of
membership function. Figure 3.12 shows the membership functions plots for the mentioned factors.
Figure 3.12: Responsiveness Membership Functions plots in Performance Measurement of SCM
The rule base of the proposed Fuzzy Logic Framework was defined in three steps; firstly, total numbers
of interactions between the input variables of the FLC were defined. The numbers of rules were established
based on the permutation with number of membership function and numbers of criteria. The numbers of factors
are eight for responsiveness and numbers of membership functions are eight. Using Hierarchical Fusion method,
two variables were fused at a time. The total amount of rules are 5^2 equal to 25 rules for every fusion. In
second step, some illogical relationships between the rules found and were omitted from the rule base. These
rules showed states that had illogical interaction between two or more criteria. For example if the “accuracy”
variable be “high” or “very high” then the “response time” would not be “very low” or “low”. It’s because when
the speed of production is not short enough, the company cannot response to the order of customer before due
date. Finally, in last fuzzy rule design step, remaining
rules were contracted in 104 general rules and the output for each rule was determined by the researcher.
The contraction was done for better decision making for each output state and easy understanding for
find out the relationships between the factors. In the last step of the methodology, the fuzzy logic toolbox of
MATLAB was applied to entering the membership functions and fuzzy rules. Furthermore, the software helped
to show the result and analyze the output performance of supply chains. The following section shows the result
of the research and discuss about the output analysis in the qualitative performance measurement in Supply
Chain Management. Fig. 3.13, 3.14, 3.15, 3.16, 3.17 and 3.18 respectively shows fuzzy rules and Fig. 3.19
shows an output membership function for responsiveness and reliability.

Fig 3.13: Fuzzy logic rules for responsiveness one which is a fusion between accuracy and no of
complaints
Fig 3.14: Fuzzy logic rules for responsiveness two which is a fusion between lead time and probability out
of stock

Fig 3.15: Fuzzy logic rules for responsiveness three which is a fusion between on time delivery and
response time
Fig 3.16: Fuzzy logic rules for responsiveness four which is a fusion between fill rate and shipping errors

Fig 3.17: Fuzzy logic rules for resfusion1 which is a fusion between Resp1 and Resp2
Fig 3.18: Fuzzy logic rules for Respfusiontwo which is a fusion between Resp3 and Resp4
The triangular membership function shown in Equation (1) is then used in MATLAB to obtain the degree of
fuzziness for each input variable.
(1)
(Retrieved from http://www.academia.edu/5542014/Fuzzy _Logic)
For instance if number of complaints are 40 in a specified time period, 40 has an intersection between
average and high then membership degree can be calculated using Eq.(2) as
(2)
F=60-40 =0.67
60-30

Similarly if the value is greater than 30 and less than 45, then this value lies between both high and average thus
the degree of membership shows to what extent the input presents a value represented by a the specified fuzzy
set. E.g. 34 lie more on the average side. Then the degree of membership can be calculated using Eq.(2) as
F=60-34 =0.866
60-30
This implies it has a degree of 0.87 average and 0.13 high. So this will be considered as average. For percentage
of orders delivered on time, e.g. if 67% of them are delivered on time. This lies between average and good but
more on the good side, then the degree of membership will be calculated using .Eq. (2) as
F=100-67 =0.74
100-50
0.74 good and 0.26 average, so this will be considered as good. After obtaining the degree of membership for
each input variable. The fuzzy rules that match the inputs degree of membership are fired and the output is
mapped into the output membership functions.
The final step in developing the fuzzy logic framework is to input data collected for all the defined supply chain
responsiveness parameters in the rule viewer in Matlab and the output which is a fusion between two variables at
a time will be generated as shown in fig 3.20, 3.21 and 3.22 respectively
.
Fig 3.20: Rule Viewers for fusion of supply chain responsivess parameters and their intermediate outputs
Fig 3.21: Rule Viewers for fusion of supply chain responsivess parameters and their intermediate outputs

Fig 3.22: Rule Viewers for final supply chain responsivess performance measurement
To obtain the degree of membership the triangular membership function in Eq.(1) will be used to map the overall
responsiveness into the output membership function as shown in fig. 3.19
Figure 3.19: Output Membership functions for Responsiveness.
IV. CONCLUSION
This study has addressed the questions of how to measure supply chain. Supply responsiveness has
become an important avenue in recent times. Many organizations around the world have been attempting to
implement performance concepts in their supply chain. The evaluation of supply chain responsiveness gains vital
importance in modern scenario. Responsivess in nature is associated with complexity and ambiguity; therefore
conventional evaluations are inappropriate and incompetent. However fuzzy logic is a very powerful tool to
compensate this limitation and deal with vague and complex situations. Responsivess of any supply chain can be
effectively evaluated using fuzzy inference system. Exploration of fuzzy logic helps in dealing with decision-
makers’ linguistic evaluation information efficiently, thereby eliminating ambiguity, imprecision and vagueness
arising from subjective human judgment. Also for any industries to be survive in today competitive market they
should periodically evaluate their supply chain responsivess. Well performing supply chain result in improving
the response and service to the customer, therefore increasing the supply chain profitability.
REFERENCE
[1] Hamidreza P., Rosnah Y., Tang S. H., & Seyed M. H. (2010, 12). Qualitative performance measurement of supply chain management
using fuzzy logic controller. Paper presented at 11th Asia Pacific Industrial Engineering and Management System. Malaysia.
[2] Sánchez, A., & M. Pérez, 2005: Supply Chain Flexibility and Firm Performance: A Conceptual Model and Empirical Study in the
Automotive Industry. International Journal of Operations & Production Management, 25, 7-8
[3] Sirigiri, P., & Gangadhar, P.V., & Kajal, K.G.(2012). Evaluation of teacher’s performance using Fuzzy Logic Techniques.
International Journal of Computer Trends and Technology, 3(2), 200-210
[4] Pyke, D.F. & Cohen, M.A. (1994). Multi-product integrated production-distribution systems. European Journal of Operational
Research, 74: 18-49
[5] Smolová, J., & Pech, M.( 2011). Dynamic Supply chain modeling using a new fuzzy hybrid negotiation mechanism," International
Journal of Production Economics, 59(1-3) ,443-453
[6] Zadeh, L.A.(1976). Fuzzy set as a basis for theory of possibility. Fuzzy Sets Systems, 1(2), 3-28. DOI: 10.1016/0165-0114(78)90029-
5
[7] Novak, V., Perfilieva, I., and Mockor, J. (2000), Mathematical principles of fuzzy logic, Dordrecht: Kluwer.
[8] Supply Chain Council, Inc. (2010). Supply Chain Operations Reference (SCOR) mode. Overview - Version 10.0.
[10] Jamshidi, M. (1997). Fuzzy Control Systems. Springer-Verlag, chapter of Soft Computing, pp.42–56.
[11] Ledeneva, Y. (2006a). Automatic Estimation of Parameters to Reduce Rule Base of Fuzzy Control Complex Systems. Master thesis,
INAOE Mexico.
[12] Ledeneva, Y., & Reyes García, C. A. (2006b). Automatic Estimation of Fusion Method Parameters to Reduce Rule Base of Fuzzy
Control Complex Systems. In: Gelbukh A.,et al. (Eds.): MICAI 2006, Springer-Verlag Berlin Heidelberg, LNAI 4293, pp. 146-155.
Ledeneva, Y.,

[13] Measuring Supply Chain Performance Guide to Key Performance Indicators for Public Health Managers(2010).Retrieved from
http://deliver.jsi.com/dlvr_content/resources/allpubs/guidelines/MeasSCPerf.pdf.
[14] Ezutah, U.O., & Kuan, Y.W. (2010). Supply Chain Performance Evaluation and Challenges. American Journal of Engineering and
Applied Sciences: 2 (1), 202-211
[15] Nomesh, B., Pranav, S., & Jalaj, B. (2012) .Quantification of agility of a Supply Chain using Fuzzy Logic. American Journal of
Engineering and Applied Sciences: 2 (2), 100-106.
.[16] Shirouyehzad, H., Panjehfouladgaran, R, Badakhshian, M. (2011).Vendor performance management using Fuzzy logic controller.
The Journal of Mathematics and Computer Science,.2(12), 311-318.
[17] Zadeh, L.A.(1976). Fuzzy set as a basis for theory of possibility. Fuzzy Sets Systems, 1(2), 3-28. DOI: 10.1016/0165-0114(78)90029-
5.
[18] Marsili-Libelli, S., 2004. Fuzzy prediction of the algal blooms in the Orbetellolagoon. Environmental Modelling and Software 19,
799–808.
[19] Beamon, B.M. (1999). Measuring Supply Chain Performance, International Journal of Operations and Production Management,
19(3), 275-292

Fuzzy Logic Framework for Qualitative Evaluation of Supply Chain Responsiveness

More Related Content

What's hot (18)

Viewers also liked (19)

Similar to Fuzzy Logic Framework for Qualitative Evaluation of Supply Chain Responsiveness (20)

Recently uploaded (20)

Fuzzy Logic Framework for Qualitative Evaluation of Supply Chain Responsiveness