Supply Chain Analytics 1 (2023) 100002

Contents lists available at ScienceDirect

Supply Chain Analytics

journal homepage: www.sciencedirect.com/journal/supply-chain-analytics

A spatial multi-criteria decision-making model for planning new logistic

centers in metropolitan areas ]]
]]]]]]
]]

İsmail Öndena, , Fahrettin Eldemirb, A. Zafer Acarc, Metin Çancıd

⁎

a
Smart System Research Group of TUBITAK TUSSIDE, Kocaeli, Turkey
b
Department of Industrial Engineering, University of Jeddah, Jeddah, Saudi Arabia
c
Faculty of Applied Sciences, Bilgi University, Istanbul, Turkey
d
Department of International Trade, Yalova University, Yalova, Turkey

A R T I C L E I N F O A B S T R A C T

Keywords: The logistics center concept has been discussed in the literature for over four decades. Logistics centers simplify
Spatial analysis the logistics network and have many advantages, such as lower transportation costs, an economy of scale, and
Multi-criteria decision-making integrated service capabilities. We propose a spatial multi-criteria decision-making model for new logistic
Urban logistics planning centers in metropolitan areas. The first focus of the study is identifying the logistic concerns, defining the factors
Logistics networks
affecting the replacement decisions and determining the weights of the factors in metropolitan areas with many
Geographic Information Systems (GIS)
expert opinions. The second focuses on spatial analysis to locate new logistics centers serving urban areas. We
Land suitability
present a case study in Istanbul, the most populous metropolis in Europe, to demonstrate the applicability and
exhibit efficacy of the method proposed in this study. Outputs of the study pointed out where the convenient
places are to locate new logistics centers.

1. Introduction literature [55] and the interest on the topic is growing day by day.
Primary advantages of having logistic centers can be counted as re-
The demand for logistic services is generally fulfilled by multiple duced costs, decreased travelled distance of the trucks, integrated and
logistics companies in the logistics centers with their facilities that better service capabilities [46]. Secondary advantages of the primary
provide integrated logistics services. In this manner the integration factors can be counted as less exhausted harmful gases, relieved traffic
means a holistic logistics service capability which creates harmony that congestion therefore decreased negative impacts on environment [43]
leads lower business costs for logistic service providers. Another aspect and lower variable costs. The relationship between logistic activities
of the logistics centers is the discordance of the transportation vehicles and transportation studies showed that important percentage of traffic
regarding size and the number. Inbound activities towards to the cen- are caused by transshipments [15], and the location of strategic logis-
ters generally done by larger trucks and the outbound activities such as tics facilities can have influences on traffic volume [64]. Therefore,
deliveries to final customers or distribution centers can be completed by right location decisions may lower traffic problems, logistics costs and
relatively smaller but more vehicles. Thus, the decision of establishing a carbon emissions. These studies also show that there is correlation
new logistics center will affect the whole logistics network strategies in between the logistics facilities’ locations and life quality of a city.
cities in the long term, and the consequences of this decision will be However, location selection is accepted as a sophisticated research
permanent or hard to change due to the decision’s initial investment area. There are different solution approaches and multi-criteria ap-
outlay. Hereby, it can be concluded that a full consideration should be proaches are successful to incorporate the various effective factors into
given to the decision of establishing a logistics center with different the problem. Moreover, spatial multi-criteria decision analysis ap-
aspects of the choice. proaches are powerful to add the spatial features to the analyses with
In addition to the complexity of the location decision, there are geographical suitability consideration [42,32,39]. Location analysis is
various advantages of these facilities discussed by researchers. The related with spatial information and affecting criteria’s effects. Multi-
concept has been debated for more than forty years in academic criteria spatial analysis is capable of integrating both of these inputs.

Corresponding author.
⁎

E-mail addresses: [email protected] (İ. Önden), [email protected] (F. Eldemir), [email protected], [email protected] (A.Z. Acar),
[email protected] (M. Çancı).

https://doi.org/10.1016/j.sca.2023.100002
Received 26 January 2023; Received in revised form 14 February 2023; Accepted 14 February 2023
2949-8635/© 2023 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/
by-nc-nd/4.0/).
İ. Önden, F. Eldemir, A.Z. Acar et al. Supply Chain Analytics 1 (2023) 100002

When spatial information is considered, emerging Geographic 2. Background

Information Systems (GIS) in location analysis is the case. In the be-
ginning of the seventies, Tobler’s statement [61] “everything is related to In this section, most relevant literature in centralization of the in-
everything else, but near things are more related than distant things” had a ventories (2.1) and logistics centers and location selection (2.2) for
huge impact on geography researches due to spreading use of GIS. The those centers have been reviewed.
statement is accepted as the first law of the geography, and decades
after it is proven by using big data [27]. In addition to the Tobler’s 2.1. Centralization of the Inventories
statement, when the spatial convenience is the subject, determining the
relevant criteria is crucial for the location analysts. Multi-criteria tools Inventory centralization aims to decrease the number of the facil-
are suitable for measuring criteria effects. And spatial multi-criteria ities into one or fewer numbers of facilities to create cost benefits to
decision making is a specialized subset of the multi-criteria analysis companies. However, there is a trade-off between distances to custo-
with their spatial analysis capabilities. Spatial multi-criteria decision- mers and inventory costs. Decreasing the numbers of the facilities
making approaches are emerged in the literature due to their analyzing causes lower investment and warehousing costs; yet, also causes higher
capabilities based on integration of GIS and multi-criteria approaches. distances to the customers. This complexity attracts researchers and
GIS are used with different methodologies for location analysis there are various studies which deal with these problems.
[9,14,30], and with multi-criteria spatial approach different location Centralization concept firstly emerged in textbooks in 1960 s without
selection processes are completed [10,63]. The main idea of the spatial mathematical proofs [56,7]. After the first examples, with the proposal
multi-criteria analysis is the integration of multi-criteria decision of the square root law, the positive effects of the centralization on in-
making and GIS by using the data that is generated by GIS in multi- ventory levels are measured practically [28]. According to the square
criteria analysis before reaching the final decision. There are various root law, in the case of centralization of 20 facilities into 1 facility cause
spatial and non-spatial factors affecting location selection decision of a 78% of decrement of inventory cost. Then, [38] proved the square root
logistic center. Also, this decision-making problem considers both law mathematically. As a following study, [20] proposed newsboy
quantitative and qualitative criteria [35,51,53]. In the literature, problem framework and calculated the total cost of decentralized
proximities to the transportation network [23,41,44,65], demand and system is higher than the centralized system. In another study, the
supply nodes [44,48], costs [12,21,23,37,41], transportation avail- question of “does the effect of centralization show a variety according
ability, capacity and quality [12,41,44], traffic [22,36,53], social ben- to the applied sectors?” is answered by [3], and the results show that all
efits [13], environmental considerations [13,53], land specialties sectors are affected positively by centralization with different degrees
[22,23], labor characteristics and availability [23,37,41,62] are used as of cost profits. The mentioned centralization literature expanded with
decision parameters. Addition to these micro parameters, [58] used the researches of lateral shipments [33,59] and portfolio effects [60]. In
Logistics Performance Index (LPI) data developed by the World Bank a recent study Stevic and colleagues are applied an economic analysis to
for the time periods 2012, 2014, 2016 and 2018 while selecting the analyze centralized vs. decentralized warehouse system for a paper
logistics center. production company and they asserted and proved that “total cost of
All in all, spatial multi-criteria analysis provides certain advantages decentralized system is higher than the centralized system” [57].
for location analysis. This assumption needs to be tested, and also it
should be scientifically supported in large areas where complex logistics 2.2. Logistics centers and location selection
are needed such as huge metropoles.
Regarding these considerations, the aim of the study is the evalua- Logistics center concept appeared in the literature more than forty
tion of the spatially convenient locations for new logistic centers that years ago [55]. During this long research period different names used
will serve for city logistics needs. For the implementation, Istanbul for a logistics center such as logistics center, distribution center, central
selected as the study area for testing the methodology due to Istanbul’s warehouse, freight/transport terminal, transport node, logistics plat-
high demand. Istanbul is the most populous city in Europe with its over form, freight village, logistics depot, distripark etc. [40,57]. Although
15 million population divided between two continents and excluding there is no clear consensus on the name logistics centers have been
over 10 million tourists annually. Additionally, the city has important placed at the hearth of the modern freight transport system and act as
industrial facilities, this characteristic even makes city’s transportation the strategic nodal points. Thus, it can be accepted as the most common
structure more complicate. Thus, the city stands out as an important term. In addition to the variety in the names, the functions, sizes,
case, not only as a human settlement, but also as an important pro- proposed service varieties of the facilities showed significant changes in
duction and trade center. time due to the changes in transshipment flow volumes. The logistics
GIS/Spatial Analysis and survey studies are combined experiences’ centres are the areas including various sets of logistics facilities that are
knowledge with GIS’ quantitative capabilities. Over twelve thousand being able to present carry out different logistics services on a com-
stakeholders are participated the questionnaire study. Their expressions mercial base to their customers relating to transportation and dis-
and judgments over transportation and logistics activities showed the tribution of goods and supporting multi-modal shipments and geo-
professionals thoughts on location selection decision and spatial ana- graphic coverages [19,55].
lysis abilities let us to illustrate these findings on the maps. During the Location selection problems for logistics centers also has been a
analysis phase, different scenarios are created the convenient locations popular research topic for researchers, and different solution ap-
and the findings expressed the spatially attractive regions for logistics proaches are discussed for different cases. Some of these studies con-
centers. tributed to the selection of criteria to be used in location selection,
In this respect the framework of the current research and its results while others provided information about the most appropriate metho-
are reported with the following structure. Background of the study is dology.
described in detail in the Section 2 with that a brief literature is given Electre III/IV is used for logistics center research in Poland by Żak
about centralization of the inventories and logistic activities. After ne- and Weglinski (2014). Wang and He used a robust optimization for
cessary definitions are given for the logistic centers; then, the char- location and allocation of the logistics center under uncertain en-
acteristics of the study area are given. Section 3 presents the metho- vironment [4]. Awasthi and colleagues used fuzzy TOPSIS for the lo-
dology and experimentations; firstly, the method is explained and then cation selection process of an urban distribution under uncertainty [1].
the survey results and the mappings are given. Section 4 gives the [49] presented a smart and sustainable operations and supply chain
findings of the experimentations, then, the final section is summarizing management in Industry 4.0. [23] used Analytic Hierarchy Process
the study and including the final discussion. (AHP) -a structure technique for organizing and analyzing complex

2
İ. Önden, F. Eldemir, A.Z. Acar et al. Supply Chain Analytics 1 (2023) 100002

decisions- for their location study of a logistic center that will serve for
textile sector in Marmara Region of Turkey. [18] proposed an analytics
approach to the decision alternative prioritization for zero-emission
zone logistics. [6] studied logistic autonomous vehicles assessment
using decision support model.
Recently many researchers attempted to integrate existing techni-
ques to reach better solutions. A conceptual model based on integration
of the artificial neural network and AHP is proposed for intermodal
freight logistics centers by [34]. Li and colleagues used axiomatic fuzzy
set and TOPSIS methodology to determine the location of the logistics
centers [36]. Basciftci and colleagues derived an exact mixed-integer
linear programming reformulation in order to solve a facility location
problem in terms of robust distribution [5]. The results of this approach
provided evidence of significant improvements in profitability and
service quality. Çakmak and colleagues proposed integrated metho-
dology to solve location selection problem which used Binary Particle
Swarm Optimization (BPSO) and GIS which provides the optimum so-
lutions faster than 0–1 Integer Programming [8]. [41] integrated ANP
and TOPSIS as an effective model to solve the distribution center lo-
cation problem. In this approach ANP method used to define the
weights of the selection criteria, while the TOPSIS applied to rank al-
ternatives. Decision support systems have been successfully applied to
many problems such as: Pamucar et al. [50], [2,16,47,52].

3. Methodology

In this study a continuous plane is considered for a facility location

problem. For the solution approach a multi-criteria spatial decision
analysis is carried out. For this analysis two main analysis tools are
combined to reach the decision. The first analysis tool is a questionnaire
Fig. 1. The flowchart of the methodology.
to represent experts’ evaluations. Because to do a comprehensive spatial
analysis it is necessary to select a suitable logistics center, Rikalovic and
be tested as different scenarios to understand how decision is changing.
colleagues suggested that Geographical Information Systems are ideal
The results of these scenario analysis will represent the suitability levels
[54]. Thus the second tool is selected as GIS/Spatial analysis. A multi-
of different geographical part of the analysis region. The final step of
criteria analysis reaches the decision with contribution of a set of de-
the methodology is to pointing out the most convenient areas in the
cision criteria. Spatial multi-criteria analysis uses GIS’ analysis cap-
city.
abilities to understand the effects of geographic criteria. There are
different analysis approaches such as distance-based algorithms, clus-
tering algorithms and so on. 3.1. Questionnaire
The application is completed in two main analysis tools. These are
Field Study and GIS Analysis. In field analysis, decision criteria for lo- In multi-criteria analysis, researchers use different techniques such
cation analysis are determined and these decision criteria are evaluated as: weighted sum, AHP, TOPSIS, ELECTRE, PROMETHEE, and so on
with different questions. The use of these techniques in the metho- with their fuzzy derivatives. These techniques are often used to reflect a
dology followed an iterative structure and Fig. 1 shows the flow of the limited number of expert opinions. Survey study can be used to reflect
used methodology in the study. views of higher participation. In the scope of the study, it will be ap-
In Fig. 1, there are two main analysis steps. The left-hand side in propriate to use the survey method since an important decision on a
Fig. 1 express how a field study to collect and calculate decision criteria complex metropolitan area will be taken. The questionnaire can accu-
weights can be created. According to this part of the figure, firstly de- rately reflect the opinions of the experts in case of high participation.
cision criteria should be determined, then, a questionnaire should be Due to this reason fuzzification of the decision judgements is not seen as
designed. In the design step what questions are adequate, who can be a necessity. It is suitable to report the results and weights of decision
correct sample for the analysis should be determined. After this step, criteria with crisp values.
data should be collected. After data collection, decision weight will be There are two stages in the survey study. These are the selection of
found as the results of the experts’ answers in the survey study and survey criteria and the selection of questionnaire questions which will
these values will be the input of the spatial analysis. be an input to the analysis. The second step is to determine what al-
The right hand side in Fig. 1 expressed the GIS analysis. These steps ternative scenarios should be considered for site selection analysis and
start with questionnaire study. The questions also express the geo- to determine these scenario parameters.
graphic decision criteria of selection of the logistics centers. Due to that
in accordance with survey study data collection and digitization of data 3.2. Decision criteria determination and questions
should be started as the first step. Then a geographic database (geodb)
should be created. After data are ready to analyze, spatial analysis can In the scope of the study, a survey data was analyzed for the analysis
be performed. In this step distances towards to focal entities such as of the site selection. At this point, the first step is to determine the
demand nodes, supply nodes or transportation network can be per- criteria and questions about the location of the logistics facility in the
formed. In the next step combining different maps, which also named as metropolitan area that constitutes the scope of the study. These ques-
overlaying can be performed to map different scenarios. In this step tions should be appropriate for site selection analysis. The point being
constraints can be added to eliminate inconvenient locations from de- evaluated is that expert opinions should be converted into numerical
cision environment. And also different decision criteria weights should values. The linguistic values are converted to numerical values as low,

3
İ. Önden, F. Eldemir, A.Z. Acar et al. Supply Chain Analytics 1 (2023) 100002

42,46
Fig. 2. Spatial analysis tools: Euclidean distance, reclassification, overlay .

1; medium, 3; high; 5. The main motivation for using 3 basic evaluation All the counted analysis approaches have different analysis structures,
judgments in evaluation is to see the difference in the opinions of the and their capabilities show differences. In case of there are high number
experts in detail. of data for a map, creating a density map [24] will illustrate how the
considered data sprawled on the study area; on the other hand, in the
3.3. Scenario creation and weight calculation event of studying with low number of the data Euclidean distance
analysis [26] will illustrate the geographic sprawl of the data. After
The survey will provide input to GIS analyzes, the second step of the creating a density or distance map, these maps are needed to be re-
method. In this phase, the criteria to be assessed were determined in the classified to perform the overlay analysis [25]. Jenks natural breaking
previous phase of the survey study. At this stage, the assessment of the points [31] is suitable to measure the impact areas of these decision
effect of the relevant criteria will be carried out. Metropolitan areas criteria geographically.
have complex structures with crowded population. For this reason, However, the amount of data does not always the correct parameter
different expertise has different opinions to reflect this complexity. for the GIS analyses. Suitability analysis for the location decision is
Within the scope of the study, scenarios reflecting different expertise based on combining expert thoughts and spatial analysis results. That is
considerations should be discussed. why an overlay analysis is needed to combine weights and spatial in-
Two views should be reflected in the scenario analysis design. These formation.
are the evaluation of decision criteria by all experts for site selection. Overlaying is the integration tool of outputs of the applied spatial
This assessment will give a general idea. The second is to evaluate the analysis for the decision criteria. The idea of overlaying is to create a
criteria according to the expertise. With these two approaches, it will be final map that draws the borders of convenient levels of the regions.
possible to determine the difference in the points of view of different This approach needs input parameters which are mainly weights of
specializations. The values to be used in the scenario will be derived input maps in the study. Overlaying is often used in GIS and multi-
from the quantification of the opinions of the linguistic experts men- criteria integration in previous studies [1,54]. Fig. 2 illustrated the how
tioned in the questionnaire. Criteria weights are measured with the given GIS analyses work. In this study, it is suggested to create final
weighted average method. The formulation of the method is given in maps which include classes that represent suitability levels. The over-
Eq. (1). The weights there were evaluated as at least 1 and maximum 9 laying results will show how the experts’ thoughts can be illustrated on
according to the evaluation scale. In the Eq. (1), wi values represents the a plane, and these results will need a final evaluation.
weights of the spatial decision criteria and xi represent the each deci- The considered facilities should not be in some sensitive areas.
sion criteria from i to n. Normalization with total value of the weights These musts can be solved with constraint maps. While creating suit-
will give the x value. ability maps for convenient areas the sensitive areas will erase from the
decision environment.
n
wx
i=1 i i
x = n
w
i=1 i (1) 3.4.1. Point density analysis
Density analysis can determine the distribution of point data over
the considered plane. In the scope of this study, this analysis method
3.4. GIS/spatial analysis was used in the analysis of high number of points. Density analysis
output is a raster dataset. The inputs are vector and point data set. For
The next step in the article study is GIS / Spatial Analysis. This step the analysis, a neighborhood around each raster cell center defined.
will give the convenient areas for logistics centers under different sce- And the number of points that fall within the neighborhood is totaled
nario considerations. The input criteria and weights will be provided by and divided by the area of the neighborhood [24].
the previous step. As a novelty a vast number of decision makers’ ex-
pression is analyzed with GIS. 3.4.2. Euclidean distance analysis
During the analysis, there is a need of geographic data for both il- When a continuous plane is considered, Euclidean distance analysis
lustration and analysis. Thus, geographic data should be gathered or applied for calculating the distances between each grid to the con-
created to be able to perform the spatial analysis. The data to be col- sidered feature. The input features of the analysis are vector data such
lected in this step will be determined by the decision criteria de- as point, line or polygon.
termined in the survey study. Then, a geographic database should be
created for the analysis. After data preparation, spatial analysis can 3.4.3. Mapping operations
perform to analyze the decision criteria. In addition to the analysis tools described in this methodology, two
There are different ways to analyze the decision criteria under GIS more mapping operations were used. One of these is the erase tool. This
such as spatial analysis, spatial statistics analysis or network analysis. function removes the restriction maps from the current workspace. In

4
İ. Önden, F. Eldemir, A.Z. Acar et al. Supply Chain Analytics 1 (2023) 100002

Fig. 3. Transportation network of the study area.

this context, placement of the installation in undesired zones can be contribution is needed to be taken into account. Due to this problem,
avoided. investments continue to expand the transportation network. The cur-
The second used tool is the weighted overlay tool [26]. This tool rent transportation infrastructure is illustrated in Fig. 3. On the other
allows you to combine multiple maps according to their importance and hand, it is known the logistics activities aggravate traffic problem, and
calculate final suitability map for the location analysis. Input maps logistics centers’ positive effects over congestion is identified
must be in raster format. These maps may have been obtained from [15,34,64].
Euclidean distance or density analysis results. While raster maps con- The logistics network structure of Istanbul is an example of the
verge at different ratios, class ratios must carry the same values. For this decentralized systems. Each logistics company has to build its logistics
reason, it is necessary to reclassify the results of the distance or density network and companies have to operate their own fleet and complete
analysis. This will ensure that the values of the maps to be merged are the transshipments by their individual capability. Even the logistics
included in the same class. This could lead to a meaningful output of the structure is decentralized, the previous study [45] showed that the lo-
result map. gistics companies are clustered in some districts. That means even the
Another GIS tool used within the scope of the study is the mean city management does not regulate logistics companies geographically,
center algorithm. This algorithm considers all features and analyzes the nature of the logistics business environment forces them to be in
where the midpoint is. When it is thought that this assessment is made some regions. Furthermore, city management is expressed that the city
for a city, it can be said that all data are taken into account and the is in a need of the logistics zones [11]. This statement makes Istanbul an
mean center can be determined as weighted or weightless. interesting study area for logistics center studies.

4. Application 4.2. Dataset

4.1. Study area According to the methodology two types of data are necessary for
the analysis processes. The first data is the judgments of the experts on
Istanbul, a megacity which is located at the intersection point of the decision criteria. Logistics questionnaire of Metropolitan Planning
Europe and Asia serves as a logistical gateway between these con- Department of Istanbul Greater Municipality (IMP) is used for that data
tinents. In addition to being a logistical gateway, the city is the most requirement [29]. Survey data represents the opinions and experiences
populated and largest metropolitan area in Europe as well. On one of the stakeholders of Istanbul’s transportation and logistics sector. The
hand, the city has a deep historical background, on the other hand, second data type is geographic data which gives chance to illustrate and
there are very active industrial and trade activities within and the analyze geographic patterns of decision criteria. The details of the used
periphery of the city. Because of these attractions, city has over 15 data are explained in the following chapters.
million inhabitants [17] and make the city a huge demand and supply
center. Active industrial and touristic activities increase both service 4.2.1. Logistics questionnaire
needs of inhabitants and manufacturing centers’ demand. Conse- IMP, the planning agency of Istanbul conducted a large ques-
quently, total volumes of the local and international logistics activities tionnaire and a part of this study dedicated for logistics to understand
are very intense in the region. Additionally, city struggles one of the how logistics activities take place in Istanbul. For that aim different
worst traffic problems in the world, and to overcome this problem any aspects of the logistics business covered by the survey such as inventory

5
İ. Önden, F. Eldemir, A.Z. Acar et al. Supply Chain Analytics 1 (2023) 100002

Fig. 4. Demand and supply nodes of the study area.

statistics, expert opinions on several topics, logistics movements, and so • Organized Industry Zones (OZ) - 4297,
on. This survey firstly describes the statistics of logistics equipment • Small Industry Sites (IS) - 4135,
inventory in the city. Types and numbers of the equipment, equipment • Major gas stations (MS) - 20,
age, and transshipments between the city districts and amounts are • Ports (P) - 36,
analyzed and described to manage the logistics activities in the city. • Shipyards (S) - 124,
The questionnaire secondly determines the expert opinions in the • Distribution Centers (DC) - 592,
logistics field. The related part of this survey to our study is that the • Bonded Warehouses (BW) - 231,
survey asks numerous experts how a new logistics center should be • Fresh fruit and vegetables wholesale market hall (FWM) - 1746,
located in the city. Furthermore, the survey drew the framework of • Road transportation distribution warehouses (RDW) - 68,
location selection criteria and results expressed expert thought. • Container warehouses (CW) - 60,
Additionally, the expert judgements are described according to their • International transportation firms (IF) - 862,
professions in the logistics management environment. The number of • Retail Distribution Centers (RC) − 12,
the participants for the questionnaire is found as 12592. This value is • Manufacturing Distribution Centers (MC) - 76,
collected by the local authority for such analysis. These experts an- • Manufacturing Facilities (M) - 332.
swered the survey questions due to their experience in the business
field. They represented the general business environment with their
4.2.2. Geographic data
answers. The working fields of the participated logistics experts are
Spatial analysis is used in the study. To carry out this analysis, there is
given as follows:
a need for geographical data related to decision criteria of a new logistics
center selection. In this context, the collection or creation of relevant
geographic data is completed. The logistics environment and its all com-
Table 1 ponents such as demand nodes, supply nodes, transportation network and
What criteria did effect to your current location decision?.
constraint maps are created as geographic data. The transportation net-
Criteria Count Normalized Value work that contains railway, motorway, airway, and railway transportation
infrastructures are created, and illustrated in the Fig. 3. The second data
Insufficient land space 1815 0.169
type is the demand and supply nodes of the city. Point data type is used for
Located in logistics center? 1705 0.159
Proximity to demand and supply centers 1690 0.158 such features. M, OZ, CW and P locations are represented in Fig. 4.
Infrastructure availability 1562 0.146 GCS_WGS_1984 used as coordinate system and D_WGS_1984 used as
Convenience to goods flows 1387 0.129 datum for created data. The data sets used in the analysis are 163 points
Affordable facility rents 947 0.088
for the warehouse location, 2 for the airport location, 4 for the seaport,
No other location alternatives for the 570 0.053
logistics facility location
and 1.506.896 land use points, which represent the demand points.
(Regulation constraint of land use) Another created data is the constraints of the decision analysis. In
Affordable land cost 515 0.048 the study, forestry lands and fresh water supply areas are taken out of
Proximity to seaports 334 0.031 consideration. In addition to the environmental consideration, land
Others 189 0.018
uses such as residential area and industrial area are also taken out of
Total 10714 1
consideration. Fig. 4 illustrated the constraints of the study area.

6
İ. Önden, F. Eldemir, A.Z. Acar et al. Supply Chain Analytics 1 (2023) 100002

Table 2 The first evaluation question is what criteria was affected for the
What criteria would you consider if you were to move to a new location?. present logistics facility. Experts were asked whether the various
Criteria Quantity Normalized Value
parameters related to this question were effective. The survey partici-
pants also expressed their participation in whether the given para-
Facility expansion limitation due to 1036 0.1816 meters are effective in the decision of the current facility choice.
the insufficient land Table 1 gives what parameters are directed to the experts. The table
Insufficient area 964 0.169
Increment of the rent costs 678 0.1188
also gives the rates of influence of the parameters expressed by the
Land convenience 555 0.0973 experts on location selection. Experts were able to select more effective
Infrastructure availability 506 0.0887 parameters for evaluation. The number of evaluations completed by
Land cost 467 0.0819 this method is 10714. The results obtained were normalized to a total of
Inbound flow problems 439 0.077
1. Priority order and importance ratios of parameters are calculated.
Decrement of the demand orders due 430 0.0754
to the firms’ locations The second question to be answered is what parameters are influ-
Outbound flow problems 272 0.0477 ential in selecting a new plant site. Regarding this question, the method
Being located in the logistics zone 141 0.0247 followed in the first question was followed. The survey results are given
Being not located in the logistics zone 113 0.0198 in Table 2.
Others 104 0.0182
The third question is how the spatial criteria affect the selection of a
Total 5705 1
new logistics center. Experts were asked whether the spatial criteria
should be effective. This question has been reported according to dif-
4.3. Results of the logistics questionnaire ferent expertise. Responses from OZ, IS, BW, FWM and IF experts show
the variation of responses according to different specializations. 10671
The article summarizes the responses of 4 questions related to the experts answered to this question and their participation rates are given
study of the IMP logistic questionnaire. These questions are given in Table 3.
below: The last question reported in the survey study is which features
should be close to a new logistics facility. In this context, twelve fea-
1. According to what criteria did you choose the current plant loca- tures were determined by city logistics management. These are:
tion? Accessibility to supply and demand nodes (C1), Proximity to highway
2. What criteria would you consider if you were to move to a new network (C2), Proximity to ports (C3), Proximity to railway network
location? (C4), Proximity to airports (C5), Expansion capabilities of the facilities
3. What spatial criteria are important for a new logistics center? (C6), Geological suitability of the land (earthquake, landslide, etc.) (C7),
4. What are the criteria weights of a new logistics facility location infrastructure availability (C8), land cost (C9), proximity to the city
decision? center (C10), accessibility to the city center (C11), availability of the

Table 3
What spatial criteria are important for a new logistics center?.

Criteria OZ IS BW FWM IF

Proximity to the city center 77% 89% 81% 90% 84%

Proximity to main transportation arterial 88% 95% 91% 95% 94%
Proximity to manufacturing sites 87% 91% 74% 59% 75%
Being in a logistics zone 96% 94% 81% 90% 68%

Table 4
What are the criteria weights of a new logistics facility location decision? (According to different professions).

Expertise Area

OZ IS P DC BW CW FWM IF RDW MC M

Participant 4297 4135 36 592 231 60 1746 862 68 76 20

Criteria c1 0098 0100 0096 0106 0087 0080 0099 0085 0073 0115 0090
c2 0105 0104 0113 0117 0108 0119 0110 0105 0127 0120 0095
c3 0047 0038 0102 0043 0081 0054 0044 0084 0131 0043 0085
c4 0035 0032 0077 0028 0050 0044 0036 0054 0061 0041 0048
c5 0048 0040 0055 0033 0064 0086 0045 0065 0032 0043 0072
c6 0096 0095 0096 0099 0083 0083 0097 0086 0098 0099 0089
c7 0113 0118 0102 0104 0113 0110 0106 0102 0085 0106 0101
c8 0106 0107 0090 0107 0107 0110 0107 0100 0094 0103 0109
c9 0095 0095 0083 0098 0081 0080 0086 0094 0106 0071 0085
c10 0063 0069 0048 0063 0057 0041 0070 0055 0036 0060 0057
c11 0081 0089 0055 0098 0067 0081 0098 0066 0065 0091 0070
c12 0114 0114 0084 0104 0104 0113 0102 0103 0093 0107 0100

Table 5
Criteria weights of a new logistics facility location decision.

Criterion C7 C12 C8 C2 C1 C6 C9 C11 C10 C3 C5 C4

Priority 0112 0111 0106 0106 0098 0095 0093 0085 0065 0047 0046 0036
Used in the GIS Analysis Not used Used Not used Used Used Not used Not used Used Used Used Used Used

7
İ. Önden, F. Eldemir, A.Z. Acar et al. Supply Chain Analytics 1 (2023) 100002

Fig. 5. Distance based transportation criteria maps.

Fig. 6. Demand and supply criteria maps and analysis results.

8
İ. Önden, F. Eldemir, A.Z. Acar et al. Supply Chain Analytics 1 (2023) 100002

Fig. 7. Proximity to city center criteria map.

personal accessibility (C12). Criterion evaluation was carried out by criterion was not taken into consideration for this reason. Infrastructure
12123 experts. availability is considered equal for every region with the reason that the
Experts assessed how close they should be to the new logistics fa- analyzed city is Istanbul. The land cost criterion has been excluded from
cility. However, as a compromise is expected a triple Likert scale was the scope of the need for the construction activity to be carried out on
used in the evaluation to force abstaining participants make a decision. state subsidies.
This scale offers two polar point along with a neutral option. In the
current research expert judgments are linguistically expressed as very 4.4. GIS/spatial analysis for determination of the convenient areas
important (VI), important (I) and not significant (NS). These linguistic
judgments are digitized with VI = 9, I = 5 and NS = 1 coefficients. The methodology section explains how to perform GIS/Spatial
Linguistic scale values’ numeric equivalent and the number of experts analysis in the study. The information obtained as a result of the
expressing that linguistic value in the questionnaire are multiplied. The questionnaire survey will be combined with the geographical data and
values obtained are normalized to a total of 1 according to the expertise the suitability analysis will be performed. According to expert opinions,
groups. The obtained values are given in Table 4. The table also reflects various scenarios were determined. And analyzes were repeated ac-
how many expert opinions are collected from the different expertise. cording to alternative scenarios. The calculated suitable regions for a
All the evaluations were pooled and the criterion weights were new logistics center are expressed in this section.
evaluated with normalization process. The calculated criteria weights
are provided in Table 5. The table also explains if the criteria are used 4.4.1. Criteria maps
in spatial analysis. In the analyses, some criteria that are discussed in In order for spatial analyzes to be carried out, it is first necessary to
the field study have not taken into consideration. There are different collect or create the required data. In the data creation stages, the maps
reasons for not being considered. The expansion capabilities of the fa- related to the criteria determined in the survey study must be created in
cilities are excluded. Because a new logistics facility will provide suf- order to complete the analyses. Since the assessed facility is related to
ficient space for all logistics companies. Geographical suitability is an logistics, data items related to supply, demand and transportation
output determined by the result of this research. The regions to be network have been prepared for analysis. ESRI’s data source is used for
found at the end of the study will be geographically convenient. The railway network; highway network and city borders are provided by

9
İ. Önden, F. Eldemir, A.Z. Acar et al. Supply Chain Analytics 1 (2023) 100002

Fig. 8. Personnel accessibility criteria map.

local authority. Other considered geographic data is created by research obtained. As a result of the analysis, the distance map is obtained and
group. These data are illustrated in Figs. 5 and 6. illustrated the results in Fig. 7.
The geographic data sets prepared for analysis were evaluated ac- In the survey study, it was found that personnel access to the company
cording to the analysis steps expressed in the methodology section. was an important criterion. The distance to the bus stops was taken as a
Considering the data, the railway and highway line are in data format. criterion for access. Istanbul has a complex network of buses. Therefore,
Ports and airports are in point datatype format. Euclidean distance 12136 stops have been taken into consideration. Point density analysis is
analysis was applied to both features. preferred due to there are a vast number of point data in the study area.
Another criterion that needs to be evaluated for logistic center lo- The input data and the results are illustrated in Fig. 8.
cation analysis is supply and demand points. Within this scope, building
doors were taken as demand nodes. There are 1506,897 features in this 4.4.2. Spatial suitability analysis
data set. Density analysis was performed on the basis of the high Criterion maps were created to complete the geographic suitability
number of features in this data set. Production centers and logistics analysis. In order to obtain the results of the geographical suitability
facilities are taken into account as supply nodes. In this context, 163 analysis, the overlay analysis should be run as stated in the metho-
features have been analyzed. Euclidean distance analysis was per- dology. The overlay analysis is expressed as a weighted overlay analysis
formed with the reason that the number of features is low. Base maps as it creates the suitability map with the criterial weights. The weights
and analysis results are shown in Fig. 6. of the decision criteria come from the results of the logistics survey. In
Another criterion affecting the logistics center decision is the dis- the survey study, experts were asked about different logistics questions
tance to the city centers. The city land use data has been taken into and more than one result was obtained. For this reason, multiple sce-
account for this analysis. Center points according to each district are narios and suitability values were analyzed to see different results
calculated using the mean center algorithm with ArcGIS. Then, spatially. In this respect it was possible to observe changes in suitability
Euclidean distance analysis performed according to the center point values according to different expertise judgements.

10
İ. Önden, F. Eldemir, A.Z. Acar et al. Supply Chain Analytics 1 (2023) 100002

Fig. 9. Illustration of the land suitability values of the Scenario I.

• Scenario 1: Spatial suitability of a new logistics center based on analysis geographic analysis model and GIS toolbox is created and the
solely spatial criteria result of the analysis is illustrated in the Fig. 9. Obtained suitability
values are from 1 to 8 where 1 expressed the most suitable regions for
In the survey, there is a question about the spatial suitability of a the new logistics center. Level 1 suitable region is also shown in the
new logistics center. Table 3 gives the evaluated criteria list by the Fig. 9.
survey question and summarizes the related results. This question gives
the weight of the spatial criteria evaluated for a new logistic center in • Scenario 2: What is the final suitability for a new logistics center
the study area. The scenario was primarily designed to calculate the
spatial suitability of these results. The Table 5 expresses the weights of location selection criteria with
As can be seen in Tables 3 and 4 spatial features are taken into all profession’s judgements. The second scenario is devoted to under-
account. Being in a logistics center is not considered in this analysis. standing how the suitability values can be obtained if the average va-
Remaining criteria which are distance to city centers, distance to lues is considered. For that reason, all considered criteria maps used as
transportation network and distance to production centers are con- input maps and calculated the suitability values expressed in the
sidered with the weights of 0.31, 0.33, 0.34. These coefficients used Fig. 10. As a result, 5 suitability levels are found and the first and
with together criteria maps in the weighted overlay analysis. For the second levels are also given in the same figure.

Fig. 10. Illustration of the land suitability values of the Scenario 2.

11
İ. Önden, F. Eldemir, A.Z. Acar et al. Supply Chain Analytics 1 (2023) 100002

Table 6 • Final Suitability Calculation

Scenario 2–5 criteria weights based on stakeholders’ groups.

Decision Scenario Scenario Scenario Scenario 5

According to explained in methodology section, the final suitability
Makers 2 - OZ 3 - BW 4 - IF - IS regions have calculated. Intersection tool used to determine final suit-
ability map with the results of Scenarios 2–5. The most suitable areas of
Participant / 4297 231 862 4135 these scenarios taken into consideration and intersection gave the final
Criteria
c1 0098 0087 0085 0100
suitability. The calculated suitability classes illustrated in Fig. 11. After
c2 0105 0108 0105 0104 finding the suitable areas, constraint maps have erased from those re-
c3 0047 0081 0084 0038 gions and Fig. 12 is found suitable for the facility.
c4 0035 0050 0054 0032
c5 0048 0064 0065 0040
5. Results and discussion
c10 0063 0057 0055 0069
c11 0081 0067 0066 0089
c12 0114 0104 0103 0114 The findings of the study can be grouped in two categories. The first
is the findings gathered from the logistics questionnaire survey. In this
field survey 12592 expert opinions were collected and a database was
generated. Thanks to wide participation, the views of different spe-
cialties can be measured. This has shown the mindset of experts in a
• Scenarios 3–6: Spatial convenience for a new logistics center based complex metropolis about the location of logistics facilities.
on different professions According to revealed results experts expressed that they want their
facilities to be close to city centers, demand nodes and supply nodes. It
Table 6 is a secondary table created based on expert judgements is also found that experts consider space constraints and land suitability
expressed in Table 4. It summarizes the priorities of location selection crucially for a new facility. Logistics centers are able to provide solution
criteria for a new logistics facility. These scenarios extended the ana- for such problems.
lysis carried out in the Scenario 2 and expressed the different profes- In the current study, judgements of the experts have been calculated
sion’s point of view. Four professions are selected due to their dense use to represent a consensus on the mean of the numeric values. According
of logistics network. The results expressed in Table 6 presents there is a to results, suitability, personnel accessibility, infrastructure availability,
significant difference in the answers, which is why considering different accessibility to demand and supply nodes, expansion capabilities are
expertise evaluations is important. found as the five important criteria in a sequence.
In Table 6, the results were grouped by the stakeholders’ clusters. Experts also have expressed their sentiments regarding some pro-
Scenarios 2, 3, 4 and 5 are focused to create suitability maps based on blems. Some of the found problems which are bureaucracy, time con-
each expertise. The result maps are given in the Fig. 11. Eight pre- straints for transportation vehicles, regulation, technical infrastructure
ference levels for calculated suitability map are found in the study area. are caused due to lack of organization and insufficient communication
between logistics sector and city management,

Fig. 11. Suitability results of the Scenario 3–6.

12
İ. Önden, F. Eldemir, A.Z. Acar et al. Supply Chain Analytics 1 (2023) 100002

Fig. 12. Calculated suitable areas for the logistics center.

These findings are critical to discuss the location of a new logistics location selection, and then those locations were weighted based on
center. In the existing studies represented the limited number of ex- experts’ opinions regarding the logistics center location. So, by per-
perts’ opinions. This situation can be seen as a problem due to differ- forming this process, the survey data has been converted input into
ences between thoughts. For overcoming that subjectivity, including as numerical values, and the formation of criterion maps is discussed by
much as experts who really have experience in the field is a necessity. the method of digitization of expert opinions. It has been determined
The second category is gathered from the suitability analysis based that this method can also provide a solution to the problem.
on the created scenarios. The GIS/spatial analysis is capable of de-
termining the suitability levels for an urban area for a logistics center 6.2. Limitations and future directions
location selection problem, and scenario results successfully are
mapped based on experts’ judgments over decision criteria. Different This study has examined the location analysis of the logistic center
regions can be clustered in the same preference levels and the most decision in Istanbul according to a proposed methodology. Because the
convenient areas can be found spatially distributed. This finding shows study has been focused on a particular city, similar results may not be
a difference from multi-criteria decision-making approaches which obtained even in metropolitan cities of the same size. However, in fu-
mostly reaches a best alternative in an alternative set. With another ture studies, the authors recommend that the methods of the current
words, a limited alternative set might not be able to represent the whole study can be performed by using the specific parameters of each city.
decision alternative set. Thus, considering the continuous plane set will Additionally, this study can be considered a s a judgmental study based
give better results compared to a limited set of discrete decision alter- on gathered data from logistics experts. Therefore, although their opinions
natives. were requested in the context of the whole city, their assumptions about
their field of duty may have created bias in their answers to the ques-
6. Conclusion tionnaire. Thus, in future studies, the authors recommend that the results
can be provided or integrated with secondary data.
6.1. Implications and contributions
Declaration of Competing Interest
In this study, a case study application is performed for a location
analysis of logistic center decision in Istanbul. The findings from spatial The authors declare that they have no known competing financial
evaluation model showed that different professions’ expectations de- interests or personal relationships that could have appeared to influ-
viate from each other, and the spatial convenient areas might be found ence the work reported in this paper.
in different areas based on the stakeholders’ judgments. Thus, it can be
said that it is required to work with professionals in different positions References
to represent different aspects of the decision environment. The high
[1] A. Awasthi, S.S. Chauhan, S.K. Goyal, A multi-criteria decision making approach for
number of experts participating in the survey study provides an im- location planning for urban distribution centers under uncertainty, Math. Comput.
portant contribution in terms of literature. In other studies, studies with Model. 53 (1–2) (2011) 98–109.
a limited number of expert opinions pointed to important findings. The [2] N. Aydin, O. Yilmaz, M. Deveci, Z. Lv, Heuristics based optimization for multidepot
drone location and routing problem to detect post-earthquake damages, IEEE Trans.
number of participants in this study was able to reflect the opinion of Intell. Transp. Syst. (2022), https://doi.org/10.1109/TITS.2022.3190698
almost all experts in the region for a strategic decision such as a lo- [3] R.H. Ballou, Estimating and auditing aggregate inventory levels at multiple stocking
gistics center. Another contribution of the study is that these expert points, J. Oper. Manag. 1 (3) (1981) 143–153.
[4] W. Baohua, H.E. Shiwei, Robust optimization model and algorithm for logistics
opinions are combined with spatial criteria. The survey results are center location and allocation under uncertain environment, J. Transp. Syst. Eng.
important within this context and the findings led us to reach spatial Inf. Technol. 9 (2) (2009) 69–74.
convenient areas. In the future studies, other metropolitan areas’ survey [5] B. Basciftci, S. Ahmed, S. Shen, Distributionally robust facility location problem under
decision-dependent stochastic demand, Eur. J. Oper. Res. 292 (2) (2021) 548–561.
results, the differences of experts’ thoughts, spatial regions and map- [6] S.R. Bonab, S.J. Ghoushchi, M. Deveci, G. Haseli, Logistic autonomous vehicles
ping result can be compared with this study. The comparisons might assessment using decision support model under spherical fuzzy set integrated
provide better evaluation of different logistics systems. Choquet Integral approach, Expert Syst. Appl. 214 (2023) 119205.
[7] R.G. Brown, Decision Rules for Inventory Management, Dryden Press,, Hinsdale, IL, 1967.
Another contribution of the work is that it shows how spatial ap- [8] E. Çakmak, İ. Önden, A.Z. Acar, F. Eldemir, Analyzing the location of city logistics
propriateness can be calculated on a plane. At this point, possible lo- centers in Istanbul by integrating Geographic Information Systems with Binary Particle
cations were determined by digitizing the parameters affecting the Swarm Optimization algorithm, Case Stud. Transp. Policy 9 (1) (2021) 59–67.

13
İ. Önden, F. Eldemir, A.Z. Acar et al. Supply Chain Analytics 1 (2023) 100002

[9] S. Chakhar, V. Mousseau, Spatial multicriteria decision making, in: S. Shehkar, [37] B.L. MacCarthy, W. Atthirawong, Factors affecting location decisions in interna-
H. Xiong (Eds.), Encyclopedia of GIS, Springer- Verlag, New York, 2008, pp. 747–753. tional operations–a Delphi study, Int. J. Oper. Prod. Manag. 23 (7) (2003) 794–818.
[10] Y. Charabi, A. Gastli, PV site suitability analysis using GIS-based spatial fuzzy multi- [38] D.H. Maister, Centralisation of inventories and the “square root law”, Int. J. Phys.
criteria evaluation, Renew. Energy 36 (9) (2011) 2554–2561. Distrib. 6 (3) (1976) 124–134.
[11] İMP, Şehiriçi Lojistik Planlama Sentez Raporu (The Report of Urban Logistics [39] J. Malczewski, GIS‐based multicriteria decision analysis: a survey of the literature,
Planning and Synthesis), İstanbul Metropolitan Planning Center, 2005. Int. J. Geogr. Inf. Sci. 20 (7) (2006) 703–726.
[12] C.L. Chen, T.W. Yuan, W.C. Lee, Multi-criteria fuzzy optimization for locating [40] I. Meidute, Comparative analysis of the definitions of logistics centres, Transport 20
warehouses and distribution centers in a supply chain network, J. Chin. Inst. Chem. (3) (2005) 106–110.
Eng. 38 (5–6) (2007) 393–407. [41] T.N.M. Nong, A hybrid model for distribution center location selection, Asian J.
[13] Y. Chen, L. Qu, Evaluating the selection of logistics centre location using fuzzy Shipp. Logist. 38 (1) (2022) 40–49.
MCDM model based on entropy weight, 6th World Congr. Intell. Control Autom. [42] I. Onden, F. Eldemir, M. Çancı, Metropol bölgelerde lojistik tesislerin
(2006) 7128–7132 (June). merkezileştirilmesi kararının çevresel etkileri, 11. Ulaştırma Kongr. (2015)
[14] R.L. Church, Geographical information systems and location science, Comput. Oper. 181–190.
Res. 29 (6) (2002) 541–562. [43] I. Önden, F. Eldemir, GIS and f-AHP integration for locating a new textile manu-
[15] L. Dablanc, Goods transport in large European cities: Difficult to organize, difficult facturing facility, Fibres Text. East. Eur. 5 (113) (2015) 18–22.
to modernize, Transp. Res. Part A: Policy Pract. 41 (3) (2007) 280–285. [44] I. Önden, A.Z. Acar, F. Eldemir, Evaluation of the logistics center locations using a
[16] M. Deveci, I. Gokasar, A.R. Mishra, P. Rani, Z. Ye, Evaluation of climate change- multi-criteria spatial approach, Transport 33 (2) (2018) 322–334.
resilient transportation alternatives using fuzzy Hamacher aggregation operators [45] İ. Önden, F. Eldemir, M. Çancı, Clustering logistics facilities in a metropolitan area
based group decision-making model, Eng. Appl. Artif. Intell. 119 (2023) 105824. via a hot-spot analysis, İşletme Araşt. Derg. 6 (4) (2014) 6–15.
[17] TUIK, Population size of the districts at the study area, The Results of Address Based [46] İ. Önden, F. Eldemir, M. Çancı, Logistics center concept and location decision cri-
Population Registration System, Turkish Institute of Statistics,, 2020. teria, Sigma J. Eng. Nat. Sci. 33 (3) (2015) 325–340.
[18] M. Deveci, D. Pamucar, I. Gokasar, D. Delen, Q. Wu, V. Simic, An analytics ap- [47] İ. Önden, M. Deveci, M. Çancı, M. Çal, A. Önden, A spatial analytics decision
proach to decision alternative prioritization for zero-emission zone logistics, J. Bus. support system for analyzing the role of sea transport in public transportation,
Res. 146 (2022) 554–570. Decis. Anal. J. 6 (2023) 100149.
[19] EEIG 2004. Logistics Centres, Directions for Use. The European Logistic Platforms [48] D. Özdemir, Strategic choice for Istanbul: A domestic or international orientation
Association (EUROPLATFORMS) EEIG, accessed April 9, 2022. 〈https://unece.org/ for logistics?", Cities 27 (3) (2010) 154–163.
DAM/trans/main/eatl/docs/EN-REV-What_is_a_Freight_VillageFinalcorretto.pdf〉. [49] T. Paksoy, M. Deveci (Eds.), Smart and Sustainable Operations and Supply Chain
[20] G.D. Eppen, Effects of centralization on expected costs in a multi-location newsboy Management in Industry 4.0, CRC Press, 2023.
problem, Manag. Sci. 25 (5) (1979) 498–501. [50] D. Pamucar, I. Gokasar, A.E. Torkayesh, M. Deveci, L. Martínez, Q. Wu,
[21] M. Erdal, M. Çancı, Lojistik Yönetimi, UTİKAD Yayınları,, İstanbul, 2003. Prioritization of unmanned aerial vehicles in transportation systems using the in-
[22] B. Erkayman, E. Gundogar, G. Akkaya, M. Ipek, A fuzzy TOPSIS approach for logistics tegrated stratified fuzzy rough decision-making approach with the hamacher op-
center location selection. Journal of Business Case, Stud. (JBCS) 7 (3) (2011) 49–54. erator, Inf. Sci. 622 (2023) 374–404.
[23] S.H. Eryürük, F. Kalaoğlu, M. Baskak, A site selection model for establishing a [51] S. Pramanik, S. Dalapati, T. Kumar Roy, Logistics center location selection approach
clothing logistics center, Text. Appar. 22 (1) (2012) 40–47. based on neutrosophic multi-criteria decision making, New Trends in Neutrosophic
[24] ESRI 2014. How point density works. ESRI GIS Dictionary, accessed March 20, Theory and Applications, Pons-Editions,, Brussels, 2016, pp. 161–174.
2022. 〈https://pro.arcgis.com/en/pro-app/2.8/tool-reference/spatial-analyst/ [52] S. Qahtan, H.A. Alsattar, A.A. Zaidan, M. Deveci, D. Pamucar, W. Ding, A novel fuel
how-point-density-works.htm〉. supply system modelling approach for electric vehicles under Pythagorean prob-
[25] ESRI 2018. How weighted overlay works. ESRI GIS Dictionary, accessed March 20, abilistic hesitant fuzzy sets, Inf. Sci. 622 (2023) 1014–1032.
2022. 〈http://desktop.arcgis.com/en/arcmap/10.3/tools/spatial-analyst-toolbox/ [53] C. Rao, M. Goh, Y. Zhao, J. Zheng, Location selection of city logistics centers under
how-weighted-overlay-works.htm〉. sustainability, Transp. Res. Part D: Transp. Environ. 36 (2015) 29–44.
[26] ESRI 2018. Understanding Euclidean distance analysis. ESRI GIS Dictionary, ac- [54] A. Rikalović, G.A. Soares, J. Ignjatić, Spatial analysis of logistics center location: a
cessed March 20, 2022. 〈http://desktop.arcgis.com/en/arcmap/10.3/tools/spatial- comprehensive approach, Decis. Mak.: Appl. Manag. Eng. 1 (1) (2018) 38–50.
analyst-toolbox/euclidean-distance.htm〉. [55] K. Rimienė, D. Grundey, Logistics centre concept through evolution and definition,
[27] B. Hecht, E. Moxley, Terabytes of tobler: evaluating the first law in a massive, Eng. Econ. 54 (4) (2007) 87–95.
domain-neutral representation of world knowledge, Int. Conf. Spat. Inf. Theory [56] K. Star Martin, W.M. David, Inventory Control: Theory and Practice, Prantice Hall,,
(2009) 88–105. Englewood, 1962.
[28] J.L. Heskett, N.A. Glaskowsky, R.M. Ivie, Business Logistics: Physical Distribution [57] Ž. Stević, E. Mulalić, Z. Božičković, S. Vesković, I. Đalić, Economic analysis of the
and Materials Management, The Ronald Press,, New York, 1973. project of warehouse centralization in the paper production company, Serb. J.
[29] IMP 2007. Greater Istanbul Municipality 1/25000 Master Plan Report. Manag. 13 (1) (2018) 47–62.
[30] P. Jankowski, L. Richard, Integration of GIS-based suitability analysis and multi- [58] I. Stojanović, A. Puška, Logistics performances of Gulf Cooperation Council’s
criteria evaluation in a spatial decision support system for route selection, Environ. Countries in global supply chains, Decis. Mak.: Appl. Manag. Eng. 4 (1) (2021)
Plan. B: Plan. Des. 21 (3) (1994) 323–340. 174–193, https://doi.org/10.31181/dmame2104174s
[31] G.F. Jenks, The data model concept in statistical mapping, Int. Yearb. Cartogr. 7 [59] G. Tagaras, Effects of pooling on the optimization and service levels of two-location
(1967) 186–190. inventory systems, IIE Trans. 21 (3) (1989) 250–257.
[32] F. Joerin, M. Thériault, A. Musy, Using GIS and outranking multicriteria analysis for [60] W.J. Tallon, The impact of inventory centralization on aggregate safety stock: The
land-use suitability assessment, Int. J. Geogr. Inf. Sci. 15 (2) (2001) 153–174. variable supply lead time case, J. Bus. Logist. 14 (1) (1993) 185–203.
[33] H. Jönsson, E.A. Silver, Analysis of a two-echelon inventory control system with [61] W.R. Tobler, A computer movie simulating urban growth in the Detroit region,
complete redistribution, Manag. Sci. 33 (2) (1987) 215–227. Econ. Geogr. 46 (1) (1970) 234–240.
[34] Y. Kayikci, A conceptual model for intermodal freight logistics centre location de- [62] D.A. Tsamboulas, S. Kapros, Freight village evaluation under uncertainty with
cisions, Procedia-Soc. Behav. Sci. 2 (3) (2010) 6297–6311. public and private financing, Transp. Policy 10 (2) (2003) 141–156.
[35] Y. Kazançoğlu, M. Özbiltekin, Y.D. Özkan-Özen, Sustainability benchmarking for [63] M.H. Vahidnia, A.A. Alesheikh, A. Alimohammadi, Hospital site selection using
logistics center location decision: An example from an emerging country, Manag. fuzzy AHP and its derivatives, J. Environ. Manag. 90 (10) (2009) 3048–3056.
Environ. Qual. 31 (5) (2020) 1239–1260. [64] T. Wagner, Regional traffic impacts of logistics-related land use, Transp. Policy 17
[36] Y. Li, X. Liu, Y. Chen, Selection of logistics center location using axiomatic fuzzy set (4) (2010) 224–229.
and TOPSIS methodology in logistics management, Expert Syst. Appl. 38 (6) (2011) [65] J. Żak, S. Węgliński, The selection of the logistics center location based on MCDM/A
7901–7908. methodology, Transp. Res. Procedia 3 (2014) 555–564.

14