The concept of logistics villages has been discussed in various forms in the literature, including logistics centres, logistics villages, freight villages, and logistics centres [2]. Logistics villages encompass a wide range of national and international logistics activities, including transportation, warehouses, distribution centres, offices, banking, postal services, insurance, and customs services [6]. Many studies have primarily focused on the challenge of selecting suitable locations for logistics villages.

While many studies have primarily focused on the challenge of selecting suitable locations for logistics villages, several notable transitions and connections between these studies provide valuable insights into this complex issue.

For instance, Ballis and Mavrotas [7] addressed the logistics village selection problem in the Thriasio region near Athens, treating it as a multicriteria decision-making problem. They employed the Preference Ranking Organization Method for Enrichment of Evaluations PROMETHEE method to determine the most appropriate location among three alternatives, using criteria such as maximum space utilization, warehouse sizes, cross-docking facilities, railway access, distance to external roads, traffic density, and the number of road crossings.

Building on this foundation, Turskis and Zavadskas [8] addressed the logistics centre location selection problem in a fuzzy environment. They proposed the ARAS-F methodology, utilizing fuzzy sets for optimal location selection. Their criteria included investment cost, operation time, expansion potential, and proximity to demand markets, which were evaluated using Analytical Hierarchy Process (AHP), and the alternative ranking was carried out using ARAS-F.

Erkayman et al. [9] further expand on the location selection problem, incorporating the Fuzzy Technique for Order of Preference by Similarity to Ideal Solution (TOPSIS), method and considering geographical, physical, socioeconomic, and cost criteria. Their focus on the eastern provinces of Turkey-Erzurum, Diyarbakır, and Malatya_establishes a connection between the theoretical framework and practical application.

Li et al. [10] introduce the axiomatic fuzzy set-TOPSIS methodology, which incorporates a broader set of logistics centre location selection criteria, fostering a more comprehensive decision-making process. Their approach is particularly relevant when evaluating numerous logistics centre alternatives based on 13 criteria. Chen et al. [11] seamlessly integrate fuzzy TOPSIS and multichoice goal programming for logistics centre location selection. Their method, which uses five criteria to evaluate five different alternatives, highlights the importance of determining criteria weights with Fuzzy TOPSIS and achieving optimal location selection through multichoice goal programming.

As the studies progress, Żak and Węgliński [12] propose a two-stage multicriteria decision-making methodology for logistics centre location selection, emphasizing the significance of criteria weights with Multiple Criteria Decision Making (MCDM)/A in the first stage and alternative ranking with Elimination and Choice Expressing Reality (ELECTRE) III/IV in the second stage. This two-stage approach enhances the decision-making process's robustness.

Elevli [13] addresses the freight centre location selection problem using the fuzzy PROMETHEE method, thus continuing the exploration of fuzzy logic in logistics village selection, albeit in the context of evaluating alternative locations for Samsun City in Turkey.

Tomić et al. [14] introduce a combination of a greedy heuristic algorithm and AHP methods for logistics centre location selection in the Balkan Peninsula region. Their emphasis on selecting the most suitable alternative based on the environment-strategy-performance paradigm opens up new avenues for decision-making in complex logistics networks.

Yildirim and Önder [15] propose a two-stage AHP-PROMETHEE methodology for evaluating potential freight villages in Istanbul. This approach, which includes criteria related to transportation infrastructure, connections, proximity to the city centre, and total surface area, underscores the importance of a systematic decision-making process.

Özceylan et al. [16] introduce a GIS-MCDM methodology for solving the freight village location problem. Their approach, which combines GIS-based criteria determination, equal weighting of alternative locations, and criteria weighting with ANP, exemplifies the integration of Geographic Information Systems (GIS) into logistics village selection.

Pham et al. [17] bring a new perspective to logistics centre location selection with their fuzzy Delphi-TOPSIS methodology. This method, which incorporates criteria obtained from literature and expert opinions, bridges the gap between theoretical research and practical decision-making. Özmen and Aydoğan [4] developed a two-stage multicriteria decision-making methodology for selecting the location of a logistics centre in Kayseri, Turkey, emphasizing the use of Best-Worst Method (BWM) for criteria weighting and EDAS methods for alternative ranking.

Uyanik et al. [18] address the logistics centre location selection problem in Istanbul, proposing a DEMATEL-IF-TOPSIS hybrid methodology that combines criteria weighting with DEMATEL and alternative evaluation with intuitionistic fuzzy TOPSIS.

Finally, Komchornrit [19] presents an AHP-TOPSIS hybrid methodology for logistics centre location selection in The Greater Mekong Subregion, further contributing to the evolving landscape of decision-making approaches in this field.

In summary, these studies build upon one another, showcasing the evolution of methodologies and approaches in selecting logistics village locations, ultimately contributing to developing a comprehensive decision-making framework for this complex issue. The literature summary is presented in Table 1.

Bayesian Best-Worst Method (B-BWM) is a distinct variation of the BWM methodology, initially introduced by Mohammadi and Rezaei [20], wherein it considers multiple decision-makers instead of just one. Yanilmaz et al. [21] extended the Federal Emergency Management Agency (FEMA) and Seriousness Manageability Urgency Growth (SMUG) models with B-BWM to enhance disaster risk reduction. This study employed B-BWM to prioritize nine different disaster hazards in a specified region.

Ak et al. [22] proposed a hybrid methodology, B-BWM-VlseKriterijumska Optimizacija I Kompromisno Resenje (VIKOR), for assessing occupational risks in the textile industry. They employed B-BWM to assign weights to six different criteria for occupational risk assessment and used the VIKOR method to rank potential hazards arising from these risks.

Gul and Yucesan [23] introduced a B-BWM-TOPSIS methodology to evaluate the performance of Turkish universities, considering five Main Criteria (MC) and 34 sub-criteria. B-BWM was applied to determine the weightings for performance criteria, while the TOPSIS method was used to rank the universities.

Gul et al. [24] investigated the prioritization of control measures based on the Fine-Kinney risk assessment method for quantitative risk assessment. They presented a hybrid approach that combines B-BWM and Fuzzy VIKOR methods to prioritize these measures based on relevant criteria.

Munim et al. [25] explored the adoption of Blockchain technologies in the oil and gas industry. They ranked the criteria with the most impact on adoption by employing B-BWM for weighting.

B-BWM has also been found to apply to logistics studies. Tsang et al. [26] emphasized using the B-BWM methodology to prioritize areas for improvement in the Environmental, Social, and Governance (ESG) performance measurements of small and medium-sized logistics companies. Gupta et al. [27] investigated the challenges associated with smart and sustainable logistics, identifying and prioritizing strategies to address these challenges. They identified 19 barriers to smart, sustainable logistics through literature review and practitioner discussions and used B-BWM to prioritize these barriers against the proposed strategies.

from actual operational functions.

Region/Area Key Findings/Contributions Studied Thriasio Developed a MDM framework for region near logistics village location selection.

Athens Not specified Introduced fuzzy logic into logistics centre location selection.

Eastern Identified the most suitable location for a provinces of logistics centre in Eastern Turkey.

Turkey

logistics centre alternatives with multiple criteria.

Not specified Integrated fuzzy logic and multichoice goal programming into location selection.

Not specified Developed a two-stage decision-making process for the logistics centre location.

Samsun city, Applied fuzzy logic to the selection of Turkey freight centre locations.

Balkan Introduced a novel approach based on Peninsula the environment-strategy-performance region paradigm.

Istanbul Proposed a systematic two-stage methodology for evaluating freight villages.

Not specified Incorporated GIS into location selection.

Not specified Integrated expert opinions and fuzzy logic into location selection.

location.

Istanbul Proposed a hybrid methodology combining DEMATEL and fuzzy logic for location selection.

Greater Presented an AHP-TOPSIS hybrid Mekong methodology for logistics centre location Subregion selection.

[10] Axiomatic fuzzy Multiple criteria based on 13 Not specified Provided a methodology for evaluating

[4] BWM and EDAS Criteria weighting with BWM Kayseri, Developed a two-stage decision-making methods EDAS for alternative ranking Turkey methodology for the logistics centre

The concept of logistics centres has been used for approximately three decades and has seen significant transformations in recent years. The ever-changing nature of production, storage, and distribution processes requires a continuous evolution and adjustment of the functional concept of logistics centres. Despite the presence of several definitions for logistics centres, a universally accepted definition has not been established yet [28]. The terminology related to logistics centres, including distribution centres, warehouse distribution centres, terminals, central warehouses, or logistics platforms, may sometimes differ in conceptual meanings This section will delve into logistics villages, recognized as regions where comprehensive logistics activities occur. We will begin by providing an overview of selected logistics villages in Europe, followed by those in Asia and America. Although the history of establishing logistics villages is relatively recent, these practices have become indispensable in European countries, with more than 60 logistics villages already in operation. These European logistics villages are home to approximately 2,400 transport operators who benefit from the services offered within these villages.

Prominent countries hosting active logistics villages include France, Germany, Spain, Italy, Greece, Denmark, Netherlands, Belgium, Luxembourg, Poland, Ukraine, Hungary, and Portugal. Noteworthy logistics centres within these countries include HTC Hoeje and NTC Nordic in Denmark, Padova, Parma, Rogivo, and Verona in Italy, Dresden, Bremen, and Zal in Germany, and Barcelona in Spain [29]. Logistic villages, strategically located within a maximum distance of 40-50 km from industrial and urban centres, hold significant appeal for countries with advanced industries worldwide. Below, you will find detailed information about these logistics villages [30]:

I. Rotterdam (Holland).

II. Hamburg (Germany).

III. Quadrante Europa (Interporto Verona) (Italy).

IV. Singapur.

V. Hong-Kong.

VI. Alliance Global Logistics Hub/Texas/USA.

VII. Atlantic Gateway-Halifax Logistics Park/ Canada.

The success of logistics villages relies on the presence of critical elements. These elements encompass extensive infrastructure, strategic positioning, efficient storage and inventory management, effective transportation and distribution systems, robust technological infrastructure, and the seamless integration of logistics services. By carefully considering these factors during the establishment of logistics villages, logistics activities can be significantly improved in terms of efficiency, speed, and cost-effectiveness.

The criteria for this study were derived from highly cited articles in the field of logistics site selection, providing a valuable source of inspiration and guidance. Notably, the study by Uyanık et al. [18] offered comprehensive insights into these criteria. The analysis revealed that the most frequently mentioned criterion was cost, mentioned 33 times [9], [11], [31]. Following closely, the criterion of environmental impact was mentioned 31 times [32], making it one of the most frequently cited criteria. Transportation accessibility also ranked high in terms of frequency, appearing 21 times [33–36]. Additionally, logistics infrastructure and the availability of a skilled labor force were mentioned several times.

These relevant criteria will serve as the evaluation metrics for the study, and their definitions are provided as follows.

Transportation accessibility: this criterion assesses the ease of access and connectivity of the logistics village to transportation networks, including roads, railways, seaports, and airports. It considers the quality of transportation infrastructure, the condition of roadways, and the availability of transportation links essential for the seamless flow of logistics activities.

Logistics infrastructure: this criterion involves assessing the infrastructure services and facilities available within the logistics village, including storage, distribution, customs clearance, and cargo handling. It is essential to evaluate whether the logistics village possesses the equipment, technology, and facilities to facilitate efficient and effective logistics operations.

Labour force: the potential labour force, along with the availability of skilled workers, their education levels, and employment opportunities in the vicinity of the logistics village, should be considered. Access to a pool of talented and experienced employees within the logistics sector is vital for successfully executing logistics activities.

Costs: cost considerations are integral when selecting a logistics village. Factors such as rental or purchase expenses, operating costs, labour expenditures, and transportation fees will directly impact the economic sustainability of your logistics operations.

Environmental impacts: sustainability and environmental considerations are of paramount importance. This criterion evaluates the logistics village's adherence to environmental regulations, energy efficiency measures, waste management practices, green spaces, and environmentally friendly initiatives. Ensuring that logistics activities are environmentally sustainable is a critical aspect of site selection.

The sub-criteria developed accordingly are illustrated in Fig. 1.

Transportation Accessibility (MC1) • Road Accessibility (SC11) • Rail Accessibility (SC12) • Seaway Accessibility (SC13) • Airline Accessibility (SC14) • Combined Transportation (SC15) Logistics Infrastructure (MC2) • Storage and Stocking Services (SC21) • Distribution Network and Logistics Services (SC22) • Customs Clearance and Export/Import Services (SC23) • Cargo Operations and Logistics Operations (SC24) • Supporting and Affiliated Industries (SC25) Labor (MC3) • Skilled Labor Potential (SC31) • Education Level and Competencies (SC32) • Employment Opportunities and Labor Market (SC33) Costs (MC4) • Rental and Purchase Costs (SC41) • Operating and Labor Costs (SC42) • Transport and Logistics Costs (SC43) Environmental Impacts (MC5) • Sustainability and Green Energy Use (SC51) • Waste Management and Recycling (SC52) • Environmentally Friendly Design and Applications (SC53)

In Turkey's strategically important Marmara Region, Sakarya stands out with significant logistics potential. According to data from 2021, Sakarya ranks among the top 10 provinces in exports and holds the 11th position for imports within the country. To assess Sakarya's logistics capacity, one must consider its geographical location, well-developed transportation infrastructure, organized industrial zones, and logistics infrastructure elements.

From a geographical perspective, Sakarya benefits from its proximity to major cities such as Istanbul and Ankara, facilitating efficient logistics operations in national and international trade. Additionally, its adjacency to the Black Sea Region contributes to regional trade growth.

Sakarya takes pride in its advanced transportation infrastructure, with the TEM Highway and E-5 Highway running through its provincial borders, facilitating efficient road transportation. Additionally, there are established rail links, providing opportunities for rail transport. Consequently, Sakarya offers a range of transportation options for logistics operations, ensuring versatility and accessibility. The presence of organized industrial zones significantly amplifies Sakarya's logistics potential. The province hosts numerous organized industrial zones spread across its territory, accommodating a multitude of factories and production facilities. This concentration of logistics activities in regions with high goods flow underscores the pivotal role played by organized industrial zones. These zones promote the seamless integration of supply chain processes and the effective management of logistics activities. Sakarya province is home to seven organized industrial zones, including the 1st, 2nd, and 3rd organized industrial zones, Ferizli organized industrial zone, Eastern Marmara machinery manufacturers specialized organized industrial zone, Kaynarca furniture specialized organized industrial zone, and Karasu organized industrial zone. These industrial zones, strategically located close to various transportation systems, bolster Sakarya's industrial and production capacity, create employment opportunities, and serve as focal points for concentrated logistics operations. Ongoing efforts are underway to expand the existing organized industrial zones in Sakarya and establish new ones. Particularly noteworthy is the work to create a logistics-oriented organized industrial zone in Sakarya1.

Sakarya has undergone significant developments in its logistics infrastructure, encompassing storage facilities, distribution centres, and logistics service providers. This robust infrastructure is pivotal in supporting logistics businesses, enabling the smooth flow of goods and services, enhancing operational efficiency, and offering cost advantages. Furthermore, Sakarya has emerged as a logistics education and research hub, with its universities offering logistics programs and conducting research in this field. This contribution aids in the training of skilled logistics professionals and the overall advancement of the logistics sector.

Considering all these factors, it becomes evident that Sakarya possesses substantial potential in logistics. Its advantageous geographical location, well-established transportation infrastructure, presence of organized industrial zones, comprehensive logistics infrastructure and services, and commitment to logistics education and research position the province as a key player in the logistics sector. Sakarya's strategic location not only facilitates efficient logistics operations but also fosters trade facilitation and the growth of the logistics industry.

A recent study by Ateş and Esen [37] evaluated Sakarya's potential as a logistics base. They concluded that Sakarya's location, advantages related to efficient transportation, suitability for combined transportation, proximity to markets, well-established collection and distribution networks, and eligibility for investment incentives, among other criteria, make it a promising candidate to become a regional logistics hub. Additionally, Sakarya's abundance of potential, human resources, and financial infrastructure further reinforce its position.

MCDM methods encompass a variety of approaches and computational techniques, each with its unique methodology and calculation method. Notable examples of these methods include the AHP, TOPSIS, ELECTRE, Preference Ranking Organization Method for Enrichment Evaluation (PROMETHEE), VIKOR, and BWM. These methods together form a comprehensive toolkit that assists decision-makers in navigating complex decision-making processes and analyzing preferences.

The choice of a specific method, characterized by its distinct properties and computational methodologies, depends on the preferences of the decision-maker, the structure of the problem at hand, and the availability of relevant data. These methods find effective applications across various domains, including business

https://www.satso.org.tr/operations, industrial projects, market research, and related areas, enhancing and streamlining the decision-making processes.

BWM is an effective approach in multicriteria decision-making. This method involves collecting input from decision-makers to identify the most favourable and unfavourable alternatives within a set of criteria. Decision-makers establish the criteria' relative importance or priority order while selecting the best and worst alternatives for each criterion. BWM provides a convenient and efficient way to uncover decision-makers' preferences. Through this method, decision-makers evaluate how criteria impact their differentiation between the best and worst alternatives. The best alternative demonstrates the highest values across specific criteria, while the worst alternative exhibits the lowest values. Consequently, decision-makers can rank the alternatives and determine their preferred choice.

BWM has proven its usefulness and effectiveness in decision-making processes across various fields. For instance, in healthcare, BWM has been applied to prioritize patient safety measures in hospitals [38], [39] and assess healthcare service quality [40], [41]. In environmental management, BWM has been used to evaluate the sustainability of renewable energy sources [42], [43] and rank environmental risks in urban planning [44]. Additionally, BWM is applicable in marketing research to assess consumer preferences for product attributes [45]. In supply chain management, BWM has been employed to optimize supplier selection criteria [45] and assess logistics performance indicators [46], [47]. These examples demonstrate the versatility of BWM and its effectiveness in aiding decision-making processes by providing valuable insights and prioritizing key factors. To implement BWM, the weights of the criterion W=(w1, w2, … , wn) must first be determined, where n is the number of criteria and m is the number of alternatives. After that, BWM involves several key steps to facilitate decision-making processes. Firstly, a set of criteria or attributes relevant to the decision problem is identified (C1, C2, … , Cm). These criteria should capture the essential aspects to consider when evaluating alternatives. Secondly, decision-makers are asked to compare and rank the criteria regarding their importance or weight. This step involves a pairwise comparison process, where decision-makers assess the relative significance of each criterion concerning others. Thirdly, decision-makers use a comparative rating scale to evaluate the alternatives against each criterion. The scale typically ranges from the "best" (Cbest) to the "worst" (Cworst) performance. It allows decision-makers to assess the performance of each alternative to identified criteria. Fourthly, each criterion's best and worst performances are determined based on the decision-makers' ratings. Similarly, all criteria are compared with the worst criterion and the vector AW = (aW1, aW2, … , aWn) is generated. The scores obtained from the comparative ratings are used to calculate the relative weights of the

∗ ∗ ∗ ∗

criteria W = (w1, w2, … wn). Finally, the overall score for each alternative is computed by aggregating the weighted scores of the criteria. W is the best weight vector, wj specifies the optimum weight taken by the j. criterion. The calculation of optimum weights using mathematical modelling is done using the BWM method.

wB wj

The aim is to ensure that the absolute differences are maximum. Thus, each = aBj and = ajw the j

wj ww

value is calculated to minimize the absolute differences for (|wB − aBjwj|, |wj − ajWww|). It means that the weights found only have one best weight. In this respect, this method has an important advantage in terms of the validity of the weights in the BWM method. In this study, the linear mathematical model in the BWM method is solved using an Excel solver. The mathematical model created in the BWM method is as follows:

minξ,

wB wj (1)

| − aBj| ≤ ξ, for all j → | − ajW| ≤ ξ, for all j.

wj wW

∑wj = 1 and wj ≥ 0, for all j. (2)

The Consistency Ratio (CR) measures the consistency of expert pairwise comparisons. It is calculated by dividing the objective function value (ξ ) obtained from the optimization process by the Consistency Index (CI). The CI is calculated by subtracting the number of criteria from the sum of the criteria weights and dividing the result by the number of criteria minus one. If the CR value is less than or equal to 0.1, the consistency of the comparisons is considered acceptable. If greater than 0.1, comparisons need to be reevaluated.

CR = . (3)

CI

The CR takes values between 0 and 1, a lower CR ratio indicates better consistency. Threshold values for CR are from the Rezaei [48] study given in Table 2.

Table 3 shows the CI for various dimensions of the decision matrix.

Separate weights are calculated for the criteria and sub-criteria considered in this study. Eq. 4 is used in the calculation of weight for all sub-criteria. Here, uij shows the normalized values of the alternative. Normalized values are obtained by dividing by the total value. This process is calculated as calculated in Eq. 5.

Vi = ∑j=1 wj × uij for all i. (4)

xij

uij = . (5)

∑xij