Fuzzy Techniques and Adjusted Mixture Design-Based Scenario Analysis in the CLMV (Cambodia, Lao PDR, Myanmar and Vietnam) Subregion for Multi-Criteria Decision Making in the Apparel Industry

Abstract

This research paper presents an all-encompassing methodology for multi-criteria decision-making in the apparel sector, with the particular objective of aiding in the determination of the most appropriate location within the CLMV subregion. The research is conducted in three crucial stages. The process began with the administration of a survey to proprietors of garment businesses in both Thailand and the CLMV countries. This survey resulted in the compilation of an exhaustive list of site-selection criteria and sub-criteria. Based on the findings of subject matter-expert interviews, Cambodia (C), Vietnam (V), and Myanmar (M) were identified as feasible alternatives. Subsequently, the questionnaire criteria and sub-criteria were evaluated utilizing the Fuzzy Analytic Network Process (Fuzzy ANP), which involved the utilization of meticulously designed pair-wise comparison matrices and local priorities. Five specialists from the Thai entrepreneurial community affirmed the effectiveness of Fuzzy ANP and expressed interest in expanding manufacturing operations in the CLMV subregion. The optimal location for Thai apparel manufacturers was subsequently determined using the Fuzzy Technique for Order of Preference by Similarity to Ideal Solution (Fuzzy TOPSIS). The results indicated that Vietnam is the most favorable option. In order to improve the dependability of results, an amended mixture-design scenario analysis was implemented. This analysis assessed the sensitivity and dependability of the proposed model in different scenarios, ensuring its applicability in real-world situations. In contrast to traditional models, this study integrates managerial judgments and preferences into the decision-making procedure, thereby accounting for the complex interdependencies among numerous criteria. The suggested methodology functions as a beneficial instrument for decision-makers, both domestic and international, as it integrates effortlessly into the organizational structure of the CLMV region. By harmonizing objectives pertaining to data acquisition, manipulation, retention, and dissemination, this framework not only enables enhanced decision-making processes, but also optimizes system efficiency.

1. Introduction

The apparel industry in Thailand is highly regarded on a global scale, due to its significant impact on the nation’s gross domestic product (GDP) and export figures. The flow of inputs and the apparel industry as a whole are entirely integrated into the supply chain in Thailand. In spite of this, this sector presents significant obstacles in comparison to other ASEAN nations. The challenges mentioned above arise from a multitude of sources, encompassing labor scarcity, escalating minimum wages leading to amplified manufacturing expenses, and a dearth of essential raw materials. Furthermore, a reduction in purchasing power has led to a decline in export values, which can be attributed to a variety of factors including the ongoing consequences of the pandemic, the conflict in Ukraine, and the global economic recession. To maintain competitiveness and achieve long-term growth in the Thai apparel industry, organizations must effectively address and overcome the numerous challenges they encounter. An extensive array of approaches and strategies are at one’s disposal in order to confront the myriad of challenges. One such approach involves the establishment of stronger brand relationships, the relocation of production facilities to foreign nations to increase manufacturing capabilities, and the implementation of industrial upgrading through innovation and automation. Apparel corporations have illustrated that effectively managing labor shortages and ensuring cost-effectiveness can be accomplished by strategically determining an optimal plant location [1].

In numerous real-world scenarios, decision-making processes are hampered by undefined objectives, constraints, and consequences. Strategic considerations regarding the location of a manufacturing facility are often crucial for the overall optimization of logistical systems. In order to attain a favorable result, an analysis of a multitude of qualitative and quantitative elements is necessary; therefore, it is imperative to investigate novel methodologies that can augment the likelihood of success. Given the substantial financial investments, formidable entry barriers, and enduring consequences associated with such decisions, organizations place the uttermost importance on them. Both operating expenses and profits are influenced by these variables. The procedure for choosing a site comprises several phases, which consist of recognizing, assessing, determining, and ultimately choosing among prospective alternatives [2]. This necessitates the evaluation of various competing standards and objectives. Multi-criteria Decision Making (MCDM) is an essential element in modern decision science [3] and offers an optimal framework for investigating this subject. The complexity of decision-making has led to uncertainty and difficulty in determining the qualities of possibilities [4]. Zadeh was the first to propose type-1 fuzzy sets (T1FS) and type-2 fuzzy sets (T2FS) [5,6]. Subsequently, this theory has been extensively implemented in MCDM scenarios, where it is employed to resolve ambiguous matters. Since then, this concept has seen extensive application in MCDM scenarios, where it is used to clarify otherwise murky circumstances. Combining fuzzy theory with MCDM yields a powerful framework for handling difficult decision issues with a significant level of uncertainty and fuzziness. It enables decision-makers to make decisions that are better informed and more resilient, particularly in circumstances where traditional methods may be insufficient. By incorporating both qualitative and quantitative factors, the integration of fuzzy theory and MCDM permits the implementation of a comprehensive strategy that improves the efficacy and dependability of decision-making processes.

In situations involving the evaluation of multiple objectives and values across a range of viable alternatives, MCDM can be an essential tool. Therefore, MCMD can be utilized to address challenges related to the selection of facility locations. In the literature, numerous MCDM techniques have been documented. The methods mentioned above include the Analytical Hierarchy Process (AHP), the Analytical Network Process (ANP), the Fuzzy Analytical Hierarchy Process (FAHP), Simple Additive Weighting (SAW), the Simple Multi-Attribute Rating Technique (SMART), the Maxi-Min and Maxi-Max techniques, and the Elimination and Choice Expressing Reality (ELECTRE) [7,8,9]. It is revealed which MCDM methodologies, namely AHP and TOPSIS, are the most commonly employed in academic research and practical applications. The implementation of the methodologies of AHP and ANP occurs within the framework of the pair-wise comparison approach. Both methodologies are executed to determine the relative importance of criteria through the utilization of pair-wise comparison matrices.

However, it could be argued that ANP, which builds upon AHP, offers greater flexibility in implementation due to its ability to handle intricate interdependencies among characteristics and decision levels. In contrast, the TOPSIS method is categorized as a distance-based strategy. In evaluating solutions, the degree of deviation from the ideal is considered. As a result of the intricacy of decision-making problems and the imprecise nature of human thought, decision-makers often encounter difficulties in precisely expressing their thoughts. Hence, through the integration of fuzzy set theory, which accommodates the inherent uncertainty in human judgment, MCDM techniques can produce more tangible, pertinent, and realistic results. The integration of Fuzzy ANP and Fuzzy TOPSIS with the current literature and novel developments pertaining to the location selection problem is an intriguing strategy.

The principal aims of this research are as follows: (1) ascertain the determinants that apparel manufacturers in the CLMV countries contemplate when making decisions regarding the location or expansion of their factories; (2) utilize Fuzzy ANP to rank the identified determinants and sub-determinants for international location determination; (3) employ Fuzzy TOPSIS to ascertain the most advantageous site for the apparel industry in the CLMV subregion. This study utilizes all available methodologies with the aim of improving the accuracy, comprehensiveness, and reliability of the decision-making process. Regarding this combination, there has been no investigation into the most advantageous apparel manufacturing option across nations. Fuzzy ANP establishes preference weights for criterion evaluation before continuing, enabling the description of critical requirements that directly affect corporate profit. Fuzzy TOPSIS then incorporates the weights to reduce the discrepancies between the actual performance values of the alternatives and their intended levels in each dimension/criterion, with the goal of identifying the optimal alternative site. Both objectives are advantageous to the economy and society as a whole. Moreover, it possesses the capacity to support governments in devising policies that incentivize investments from investors, and provides important agencies and practitioners with perceptive data that can improve the quality of decisions.

When selecting locations to operate in within the fiercely competitive and ever-evolving global market, garment manufacturers must exercise extreme caution. Achievement, competitiveness, and cost-efficiency are all influenced by the location of production. Sensitivity analysis is a crucial element of our comprehensive methodology, functioning as a beneficial tool for apparel manufacturers to enable informed decision-making. By integrating the Fuzzy ANP and Fuzzy TOPSIS methodologies, this exhaustive investigation ranks and evaluates industrial locations in the CLMV region. Real-world application of this approach is facilitated by its acceptance of the unpredictability inherent in site selection criteria. By employing both methodologies in this study, not only does the researcher showcase a dedication to methodological progress, but the process also underscores the significance of utilizing sophisticated tools to develop dependable and well-informed assessments within the dynamic apparel industry. Stability analysis is necessary to ascertain the manner in which input factors impact our model of site selection. The utilization of sensitivity analysis facilitates the evaluation of the consequences of ambiguity on the decision-making process and the determination of location prioritization.

The subsequent sections of this investigation will be organized as detailed below. In the following section, an evaluation of a locational decision utilizing MCDM techniques is presented. An explanation of the suggested methodologies is provided in Section 3. The findings of the investigation are subsequently elaborated upon in Section 4. The results of a sensitivity analysis performed utilizing a modified mixture design are subsequently detailed in Section 5. The concluding section of the manuscript comprises the discussion and conclusion.

2. Locational Decisions Based on MCDM Techniques

2.1. Introduction to Locational Decisions

The process of selecting a location is a strategic decision that holds significant importance in the development of profits and overall performance outcomes for a corporation. When a company starts up or develops its operations abroad, the difficulty is deciding in which one or many geographic places to locate [10]. Multiple models and algorithms have been used to address facility placement selection issues for decades. For instance, [11,12,13] provide summaries of the reviewed location studies. The Facility Location Problem (FLP) was designed to identify the ideal quantity and location of facilities while taking into account constraints including budget, schedule, sustainability, and efficiency [13,14]. According to Ogryczak [15], one of the most important factors to consider when determining where to locate a firm is the physical length of service facilities and the distance between demand points. Conventional research tended to optimize a single factor like cost, distance, or time using intricate mathematical formulations, but they ignored qualitative data like criterion values [16] and an unpredictable setting [17].

Those factories have had to supply new customers around the world efficiently and cheaply for a number of years, due to the expansion of global trade and rising consumer demand. As a result, the location of appropriate location facilities assumes great importance for foreign enterprises seeking entry into global markets. Lack of access to clients, raw materials, transportation, etc. [18] are just some of the problems that would arise from not holding events in appropriate locations. Considering several viable options is crucial to solving the location problem, which is complicated by a wide range of qualitative (such as accessibility to clients and service quality) and quantitative (such as cost and distance) considerations. Some aspects are more crucial than others, and these can often dictate the outcomes [19]. A business must consider important parameters affecting location selections in a certain industry if it is to gain a competitive edge in the global marketplace.

2.2. Role of Multi-Criteria Decision Making (MCDM) Techniques

The MCDM approach can help decision-makers consider a wide variety of factors [20,21]. It focuses on decision-making in situations where multiple criteria are against one another. The process of using multiple criteria to analyze a problem assists in identifying its core elements. Suitable for economic choice issues with certain, uncertain, or hazardous circumstances, the decision considers both quantitative and qualitative criteria [22]. MCDM problems usually fall into two different groups. In one group, there are problems with more than one possible answer. Another has to do with problems that can be solved in an infinite number of approaches. If an attribute’s value must be in a certain range, the possible results could be infinite. This is called a “multiple objective optimization problem” [23].

Even though MCDM problems can be very different depending on the case, there are some things they all have in common. For example, multiple criteria are often set up in a hierarchy, and criteria usually conflict with each other. After the problem is set up, a scale must be used to rank the performance of each option, considering all factors, so that the options can be ranked in the final stages. Several MCDM techniques have been adopted to solve location problems. The literature contains several studies examining placement decisions over the past decade, as evidenced by the following References: [8,10,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54]. The application details regarding location determinations are provided in

Table 1. Locational decisions based on MCDM techniques.

MCDM Approaches	Author(s)	Paper Contributions
Fuzzy ANP	[24]	To select the best location for greenhouses in Iran via Fuzzy ANP
Fuzzy AHP, TOPSIS and Fuzzy TOPSIS	[25]	To select the site of the warehouse in Bangladesh
AHP	[26]	To select a manufacturing plant from five locations in India
AHP	[27]	To select a particular manufacturing industry from five locations in India
AHP and ANP	[28]	To compare AHP and ANP approaches for approaching the best location for supplying mangoes to the different areas in Bangladesh
TOPSIS, AHP, and GIS	[29]	To determine locations for dam construction in Iran
AHP, TOPSIS, and Fuzzy TOPSIS	[30]	To assess and determine the optimal site for the Urban Distribution Center (UDC) in the province of Yogyakarta Special Region, Indonesia
TOPSIS, Fuzzy AHP, and DEA	[31]	To determine the optimal site for solar power facilities in Vietnam
ANP	[32]	To identify the optimal site for the construction of a dry port in Togo
AHP, and PROMETHEE	[10]	To rank 66 cities for business location investment in Europe
Fuzzy TODIM and Entropy	[33]	To pick an offshore wind-photovoltaic-seawater pumped storage power facility in China
TOPSIS	[34]	To select location of freight consolidation facility in Melbourne, Australia
AHP, Fuzzy AHP, TOPSIS, and PROMETHEE	[35]	To discover the precise location of the expressway segment within Poland
ArcGIS, AHP, PROMETHEE, and VIKOR	[36]	To select electric vehicle charging station (EVCS) sites in Istanbul
AHP, ArcGIS	[37]	To select wind farm in Alborz, Iran
TOPSIS and SAW	[38]	To compare and select textile manufacturing plants in ASEAN countries
AHP and TOPSIS	[39]	To evaluate the best countries to develop new manufacturing settlements in African countries
AHP, fuzzy AHP, and GIS	[40]	To compare fuzzy AHP and GIS-based AHP methods for the selection of hospital sites in Prayagraj City, India
AHP	[41]	To select optimal bus-stop locations in Bandar Seri Iskandar, Malaysia
Best Worst Method (BWM)	[42]	To select a wind farm in Izmir, Turkey
AHP and FAHP	[43]	To determine the optimal site for the furniture manufacturing sector in Bangladesh
AHP and ArcGIS	[44]	To select a wind farm in Iran
Fuzzy AHP and GIS	[45]	To select a wind farm in Sudan
SF-AHP and WASPAS	[46]	To determine the optimal site for an offshore wind power station (OWPS) in Vietnam
Fuzzy ANP and Fuzzy TOPSIS	[47]	To determine optimal sites for the establishment of wind-energy infrastructure
TOPSIS	[8]	To locate suitable locations for textile manufacturing plants in the ASEAN region
AHP, TOPSIS and PROMETHEE	[48]	To find a place for a medical waste disposal center in the province of Semnan in Iran
AHP and GIS	[49]	To select locations for wind farms in Wolaita, Ethiopia

.

Upon examining multiple MCDM-based location selection studies, it was evident that the framework could be applied to enhance decision-making in a range of situations. However, there was little research utilizing MCDM to promote economic integration in the region. Considering the aforementioned issue, it is essential to make use of hybrid MCDM techniques in order to identify the best possible location for the CLMV subregion within the context of the AEC Blueprint 2025.

2.3. Comparative Analysis of MCDM Techniques

According to Saaty [50], real-world problems exhibit interdependence among their components, specifically between the qualities of the criteria and/or alternatives. Similarly, a multitude of decision-making situations exhibit a lack of hierarchical structure because of their complex interrelationships and interdependencies among criteria and/or indices. The adoption of ANP presents a more advantageous method for addressing these scenarios as it facilitates a reciprocal interaction among each measurement criterion [50,51]. Pair-wise comparisons in the ANP-based method include managerial decisions using the discrete scale of 1–9. As the complexity of the choice circumstance increases, so does the likelihood of imprecision and uncertainty. Location selection and assessment under fuzzy settings is more accurate, since it considers more criteria and options. Fuzzy TOPSIS additionally prioritizes country options by minimizing uncontrollable elements and optimizing controllable factors [52,53]. Its straightforward approach, easily programmable computation mechanism, and improved outcomes make it popular. Table 1 demonstrates that only one study [47] looked at the combination of Fuzzy ANP and Fuzzy TOPSIS to determine the best location for a wind farm in Iran. Nevertheless, no study has examined real-world criteria for choosing an apparel manufacturing plant, particularly in the CLMV subregion.

To facilitate economic integration between Thailand and the neighboring CLMV and to establish a regional center, it is imperative to identify an appropriate site for the competitive apparel industry. To achieve this goal, the implementation of MCDM techniques must be integrated. An enhancement to the decision-making process of apparel manufacturers in Thailand can be achieved by incorporating the Fuzzy ANP and Fuzzy TOPSIS methodologies. These methodologies provide a robust framework for effectively managing complexity, uncertainty, and subjectivity during the site selection procedure. The apparel industry is capable of making well-informed and resilient decisions by taking into account a vast array of interrelated factors.

2.4. Challenges and Limitations

The application of a combination of MCDM methods is critical for the CLMV subregion for a number of reasons. The process of determining the most advantageous locations entails the consideration of various quantitative and qualitative factors. Methods for Order Preference by Similarity to the Ideal Solution (TOPSIS), the Analytical Hierarchy Process (AHP), and the Technique for Order Preference by Similarity to the Ideal Solution (AHP) are MCDM techniques that enable the consideration of a wide range of criteria, including socio-cultural and economic aspects. In a region as disparate as CLMV, the complexity of location decisions necessitates a comprehensive strategy.

Furthermore, regional economic integration necessitates complex interconnections among a multitude of decision-making levels and attributes. As an expansion of the AHP methodology, the ANP is highly suitable for the capture and analysis of these interdependencies. Through the examination of interconnected facets of decision-making, the MCDM framework enhances comprehension regarding the determinants that impact the selection of locations within the CLMV subregion.

Furthermore, in light of the ever-changing and unpredictable characteristics of economic environments, a decision-making approach that can effectively navigate ambiguity and vagueness is crucial. By incorporating Fuzzy ANP and Fuzzy TOPSIS into our research, we are able to account for the inherent uncertainty in preferences and generate more accurate results. This holds particular significance in a region that is experiencing economic transformations and progress.

In light of this, the MCDM framework offers a methodical and well-informed approach to making decisions, taking into account numerous dimensions and criteria. Well-informed decisions regarding optimal locations play a significant role in enhancing the overall efficiency and success of integrated economic activities within the CLMV subregion, with respect to economic integration. Through the application of a variety of MCDM techniques, this research endeavors to provide insights that may promote cooperation among CLMV nations. The process of determining the most advantageous sites takes into account a multitude of elements that can foster synergies between nations, thereby fostering a more cooperative and integrated economic climate.

3. Proposed Methodologies

This study proposes the importance of adopting advanced analytical techniques by utilizing two innovative MCDM methodologies, namely Fuzzy ANP and Fuzzy TOPSIS, to determine foreign placement decisions in the CLMV subregion for the apparel sector. By accepting a wide range of criteria and intricately capturing the linkages between them, these novel approaches not only speed up the evaluation process, but also improve the overall decision-making. These two techniques are able to encompass a large variety of aspects and the interconnection of those factors, which could make it easier to make judgments that are both well informed and resilient. In addition, the apparel company’s criteria, sub-criteria, and alternative options will be carefully gathered from industry professionals in the relevant sector through surveys and interviews. This approach not only underscores the methodology’s tailored relevance but also illustrates how creatively sector-specific insights may be utilized to inform strategic decision-making.

By employing advanced decision-making techniques like Fuzzy ANP and Fuzzy TOPSIS, as well as identifying key criteria influencing international location selection within the garment industry, countries in the CLMV subregion can strategically position themselves to attract foreign investment and businesses. This study also benefits decision-makers and investors who must choose a foreign location. As a result, they are more competitive as a group in the global market, which aids in economic integration worldwide.

This research project has been carried out in four stages. Figure 1 outlines the steps needed to complete this research. The figure shows the four main stages of the study: (1) establishing factors influencing international location selection, (2) investigating potential locations, (3) applying Fuzzy ANP to weight variables, and (4) choosing an alternative nation using Fuzzy TOPSIS.

3.1. Identification of Factors Influencing International Location Decisions

Different literature studies classify factors differently, since their definitions and importance vary among industries and businesses [19]. Entrepreneurial perceptions and attitudes play a significant role in location decision-making, due to their interdependence and varying degrees of achievement. The significance of location considerations varies by industry and operating environment [54,55]. Few studies have examined international placement decisions for modern manufacturing processes, especially in the textile and garment industry [8,38]. Therefore, the initial stage of the study aimed to ascertain the essential characteristics that contribute to the success of placement selections made in foreign markets, utilizing survey research methodology. Initially, a comprehensive interview was held with the management group possessing expertise in the relocation of manufacturing facilities to foreign countries. This interview was supplemented by consultations with the Thai Apparel Manufacturers Association (TGMA) and pertinent government officials. Subsequently, a meticulously designed questionnaire was produced, encompassing eight primary criteria and a total of thirty-two sub-criteria. These items were derived from extensive research of the relevant literature and insightful interviews conducted with prominent professionals in the apparel industry. This questionnaire tries to discover factors influencing location choices on a global scale. A preliminary investigation was undertaken involving a cohort of 30 apparel enterprises to evaluate the efficacy of the survey instrument in relation to its user-friendliness and suitability. According to [56], the Cronbach alpha analysis yielded a coefficient of 0.976, suggesting that the questionnaire had a good level of reliability for its intended application in the study.

This study’s target population consisted of 791 Thai apparel manufacturers. The sample size of this study, 266 individuals, was determined by applying the Taro Yamane formula [57] at a 95% confidence level. A total of 211 were returned for analysis, at a response rate of 79.3%. Likert scales with five points, from (1) “Not at all important” to (5) “Extremely important”, were used to rate each question. In addition, each respondent was asked to rate the significance of each of the eight major criteria on a scale from 1 (the most significant) to 8 (the least significant). The Rank-order centroid weight method (ROC) [58] is utilized to define weights for ranking and selecting the relevant criteria that will be incorporated into the proposed location selection model. The expression for the jth rank of ROC ($W_j^{ROC}$) is given by Equation (1).

$$W_j^{ROC}=\frac{1}{n}(\sum _{k=j}^n\frac{1}{r_k}),j=1,2,\dots ,n$$

(1)

where $r_k$ is the rank position of the kth criterion.

3.2. Identification of Alternative Location in CLMV Subregion

The establishment of ASEAN, or the Association of Southeast Asian Nations, occurred in 1967, with the purpose of fostering prosperous and peaceful communities as well as stimulating cultural development, social improvement, and economic growth throughout the region. To enhance the region’s development, ten member countries founded the ASEAN Economic Community (AEC) in 2015 [59], with the goal of achieving “a prosperous, secure, and exceptionally competitive ASEAN economy” [60]. Furthermore, the process of integrating into a unified market would provide investors with the opportunity to broaden their market reach [61], encompassing a population of over 684 million individuals as of 2023 [62]. The process of economic integration has undergone a transition from AEC 2015, which had the objective of establishing an economic community by the conclusion of 2015, to AEC 2025, which seeks further integration through regional economic integration, as outlined in the AEC Blueprint 2025 [63,64]. The AEC Blueprint 2025 aims to develop a highly connected, competitive, dynamic, innovative, inclusive, people-centered economy that reflects ASEAN’s role in the region and world. Given the region’s growing economies and social and economic prosperity, there is plenty of potential for an intra-ASEAN strategy. Furthermore, it is indisputable that ASEAN serves as a favorable place for enterprises to construct manufacturing and production facilities, facilitating seamless global trade. The geographic location of ASEAN and its comparatively low manufacturing costs contribute to this phenomenon [65].

CLMV is an ASEAN subregion representing Cambodia (C), Lao PDR (L), Myanmar (M), and Vietnam (V). The fast-growing CLMV neighbor may readily access Thailand by numerous means of transportation. These adjacent countries have abundant resources and low production costs. These countries also export a lot of labor to Thailand [1]. Shifting production to such nations will also help upgrade the industry by diminishing Thai apparel businesses’ domestic production base and addressing labor shortages. This would allow those enterprises to focus on functional upgrading and added value in the country.

Thailand launched a master plan for trade and investment relations with CLMV nations in 2017 to boost the regional economy and provide investors with an edge [65]. This approach also linked logistics and supply chain systems and built strong small and medium-sized enterprises. The AEC was established to liberalize and facilitate investment, harmonize customs policies, and promote greater economic integration among ASEAN member states. As a result, members might eliminate cross-border barriers and facilitate the free flow of goods, services, workers, and capital.

The literature on production migration to labor-sending countries, especially CLMV adjacent countries, for Thai apparel industry upgrading is scarce [1]. As a result, an MCDM framework model for the selection of apparel sector locations is proposed in this study. Although facility placement studies are comprehensive, the process of choosing a location is unpredictable. Numerous companies waste valuable resources on inefficient factories. Therefore, the concept of fuzzy-MCDM was developed through the combination of fuzzy set theory with MCDM. This approach enabled the development of decision-maker models capable of processing information and knowledge that is subjective in nature and subject to uncertainty. Since fuzzy set theory permits the incorporation of qualitative criteria in situations where exact measurement is challenging or where criteria precision is compromised by a lack of data, the fuzzy system has been increasingly utilized in conjunction with other MCDM methodologies to address challenges in a variety of fields [66]. To demonstrate the applicability of the research to a real-world apparel industrial context, the methodologies are applied by surveying critical success factors for international location decisions and proposing a framework model for choosing the best alternative country for a real-world industry.

After figuring out the most important factors that affect the apparel manufacturing sector, the second step began, which was to determine the potential countries to move to or to establish that sector. According to [67], the preliminary selection was a significant procedure for selecting candidate models. It is essential to compare and evaluate the characteristics of each country, such as population size, investment climate, exchange rates, etc. In addition, interviews were conducted with senior executives from Thai and CLMV apparel companies, and site visits were undertaken in these four nations. It was discovered that investors from numerous nations, such as the United States, Canada, and Korea, expanded their garment production facilities to Myanmar, Vietnam, and Cambodia, suggesting that the labor force in those nations is qualified, skilled, and experienced in producing garments. There is a tendency to increase the minimum wage in each of the countries. However, the minimum wage in these nations was roughly two to three times less than in Thailand. In terms of market size and labor force, which is proportional to the size of each country’s population, Vietnam is the largest country in CLMV, while Lao PDR is the smallest. In addition, Myanmar is eligible to receive benefits under Japan’s Generalized System of Preference (GSP), which allow qualified products to enter preference markets, whereas Cambodia can benefit from the GSP of the United States, the European Union, and Japan. Despite the fact that Lao PDR can also benefit from the GSP, experts believe that the country’s population is too small to support a new apparel manufacturing base in the future. Consequently, our proposed model designates Myanmar, Vietnam, and Cambodia (Figure 2). The industry can leverage these advantages, including a large population, cheaper wage costs, a favorable tariff structure, and other factors, to facilitate the development of its business in the CLMV subregion.

3.3. Fuzzy ANP

The third phase was the implementation of the ANP as a MADM technique, along with fuzzy logic, with the purpose of calculating the weights of crucial aspects that affect the apparel manufacturing sector. Before moving on to the development of Fuzzy ANP, an overview of the motivation behind fuzzy theory is provided. Zadeh [5] originally introduced the concept of fuzzy set theory as a means to address the inherent ambiguity associated with human decision-making. The utilization of informal language is more advantageous in elucidating concepts related to human interaction, since it better accommodates the rationality of doubt arising from imprecision or ambiguity. The degree of membership of an element in each set, denoted as µM(x), is represented by a fuzzy set. The membership function assigns a value between 0 and 1 to each possible value. For an element x in M, µM(x) = 1, not 0.

Verma et al. [68] claimed that triangular fuzzy numbers (TFNs) are often used when values or factors of a problem are not well known. The fuzzy problem is easier to explain with TFNs than with the interval value [69]. The representation of TFNs consists of three parameters, namely l, m, and u. The parameters l, m, and u indicate the lowest practical value, the most promising value, and the highest value that could account for a fuzzy event, respectively. The equation for the TFNs is denoted as Equation (2).

$$\mu M\left(x\right)=\left\{\begin{array}{cc}(x-l)/(m-l)& l\le x\le m\\ (u-x)/(u-m)& m\le x\le u\\ 0& otherwise\end{array}\right.$$

(2)

Saaty’s [50] Analytic Hierarchy Process (AHP) has been expanded upon in the form of the Analytic Network Process (ANP). The AHP presupposes that all elements within a given cluster exhibit preferential independence and that no correlation exists between clusters operating at distinct levels [70,71]. Nevertheless, numerous choice problems are not amenable to hierarchical approaches, due to the presence of interactions and dependencies between higher- and lower-level components [72]. In response to this concern, Saaty developed the ANP as a means of addressing complex and unstructured situations involving alternatives or criteria. The ANP does not have a strict hierarchical structure, so that relationships within and between cluster parts can interact and feedback [71,72]. Under the ANP, the relative value of each criterion must first be determined by carrying out a pair-wise comparison. To rank these factors, the comparisons must also be compared and added together.

The ANP uses Saaty’s nine-point system (1–9), which is similar to the AHP. ANP is a more effective method for accommodating these circumstances, since it permits a feedback relationship between each measurement criterion [72]. The ANP models contain two components. The first component is a control dimension, which is comprised of factors and sub-factors that determine the interactions in the system under consideration. Another is a network of impact or influence among concentrations of such problems. Since precise numerical judgment is difficult and decision-maker preferences are usually imprecise, integrating fuzzy set theory into the ANP model is the most practical way to handle complex decision-making. Consequently, it appears that the Fuzzy ANP approach is a more suitable method for attaining realistic outcomes.

With a pair-wise comparison matrix, Fuzzy ANP utilizes both interdependencies and intrinsic dependencies between criteria. A triangular fuzzy number (TFN) is employed to produce pair-wise comparison matrices for evaluating the strategy determiner’s decisions. In comparison to mathematical optimization models, the Fuzzy ANP is simpler to comprehend and requires less time to discover the optimal solution. The application of Fuzzy ANP has been documented in numerous complicated decision-making scenarios, such as supplier selection [73,74], performance measurement [75,76,77], portfolio selection [78], project selection [51,79,80], and technology selection [81], among other instances.

The fuzzy scale established in

Table 2. Linguistic importance scales.

Linguistic Importance Scales	Triangular Fuzzy Scale	Triangular Fuzzy Reciprocal Scale
Absolutely more important	(5/2, 3, 7/2)	(2/7, 1/3, 2/5)
Very strongly more important	(2, 5/2, 3)	(1/3, 2/5, 1/2)
Strongly more important	(3/2, 2, 5/2)	(2/5, 1/2, 2/3)
Weakly more important	(1, 3/2, 2)	(1/2, 2/3, 1)
Equally important	(1/2, 1, 3/2)	(2/3, 1,2)
Just equal	(1, 1, 1)	(1, 1, 1)

Source: [82].

by [82] serves as the foundation for the computational method used in this paper. In pair-wise comparisons of each criterion and sub-criterion in Fuzzy ANP, linguistic variables are primarily employed to assess decision-makers’ linguistic judgments.

The current investigation utilized the extent analysis approach, as described in Reference [83], to assess fuzzy pair-wise comparisons. This was achieved by converting fuzzy numbers (TFNs) into crisp numbers. Let X = {x₁, x₂, …, x_n} be an object and G = {g₁, g₂, g₃, …, g_m} be a set of objectives. Even though the extent method is used, each criterion and sub-criterion is used to increase the number for each goal gi. This is shown by the following equation.

M¹_gi, M²_gi, M³_gi, M⁴_gi, M⁵_gi, …, M^m_gi (i = 1, 2, 3,…, n)

(3)

where M^j_gi (j = 1, 2, 3, …, m) represent TFNs.

The extension analysis proposed by Chang [83] consists of the subsequent stages:

$$S_i={\sum }_{j=1}^m{M}_{gi}^j\otimes {\left[{\sum }_{i=1}^n{\sum }_{j=1}^m{M}_{g_i}^j\right]}^{-1}$$

(4)

To derive ${\sum }_{j=1}^m{M}_{gi}^j$, apply the fuzzy addition operation to m extent analysis values with a specific matrix in the following manner:

$$\sum _{j=1}^m\underset{gi}{\overset{j}{M}}=\left(\sum _{j=1}^m{l}_j,\sum _{j=1}^m{m}_j,\sum _{j=1}^m{u}_j\right)$$

(5)

To obtain ${\sum }_{j=1}^m{M}_{gi}^j$, apply the fuzzy addition operation of m extent analysis values for a specific matrix, such that

$${\sum }_{i=1}^n{\sum }_{j=1}^m{M}_{g_i}^j=\left({\sum }_{j=1}^m{l}_j,{\sum }_{j=1}^m{m}_j,{\sum }_{j=1}^m{u}_j\right)$$

(6)

After that, use Equation (7) to find the vector’s inverse:

$${\left[{\sum }_{i=1}^n{\sum }_{j=1}^m{M}_{gi}^j\right]}^{-1}=\left(\frac{1}{{\sum }_{i=1}^n{u}_i},\frac{1}{{\sum }_{i=1}^n{m}_i},\frac{1}{{\sum }_{i=1}^n{l}_i}\right)$$

(7)

The definition of the degree of probability of $M_2=\left(l_2,m_2,u_2\right)\ge M_1=\left(l_1,m_1,u_1\right)$ is as follows:

$$V\left(M_2\ge M_1\right)=\sup \left[\min \right({\mu }_{M_1}\left(x\right),{\mu }_{M_2}\left(y\right)\left)\right]$$

(8)

The expression is as follows:

$$V(M_2\ge M_1)=\left\{\begin{array}{c}1\\ 0\\ \frac{l_1-u_2}{\left(m_2-u_2\right)-(m_1-l_1)}\end{array}\hspace{1em}\begin{array}{c}\mathrm{if}{m}_2\ge m_1,\\ \mathrm{if}{l}_1\ge u_2,\\ \mathrm{otherwise}\end{array}\right.$$

(9)

The definition of the degree probability that a convex fuzzy number $M_i$ exceeds k convex fuzzy numbers is

$$V\left(M\ge M_1,M_2,\dots ,M_k\right)=\min V\left(M\ge M_i\right),i=1,2,\dots ,k$$

(10)

Suppose that

$$d\left(S_i\right)=\min V\left(S_i\ge S_k\right),k=1,2,\dots ,n;k\ne i$$

(11)

The mathematical description of the weights vector is

$$W^{\prime }={\left(d\left(S_1\right),d\left(S_2\right),\dots ,d\left(S_n\right)\right)}^T$$

(12)

When n elements are represented by $A_i=(i=1,2,\dots n)$ through normalization, the normalized weight vectors are

$$W={\left(d\left(A_1\right),d\left(A_2\right),\dots ,d\left(A_n\right)\right)}^T$$

(13)

in which $W$ is a non-fuzzy variable.

Cheng’s extent analysis yields local criteria or sub-criteria weights [84]. Local weights become precise quantities to simplify Fuzzy ANP computation. The final global weights are obtained using sequenced computation. In assessing alternatives in relation to predetermined criteria, the decision-makers employ linguistic factors to determine which criteria are preferred. Pair-wise comparisons of decision variables from “Very bad” to “Excellent” were represented by the membership functions of TFNs M₂, M₄, M₆, and M₈. The TFNs M₂, M₄, M₆, and M₈ represent the values of midrange preference among them. Fuzzy ANP compares pairs using linguistic factors, illustrated as triangular numbers in Table 2.

In accordance with the MCDM technique, the specialists were given a second questionnaire in the form of matrices to fill out, in addition to being questioned personally. The second questionnaire was constructed in accordance with the model that was provided for location selection. This model includes criteria, sub-criteria, and alternative countries, which were identified via the process of surveying and interviewing individuals, as mentioned in Section 3.1 and Section 3.2. The primary criteria for selection were reviewed against one another using the pair-wise comparison method, and then the sub-criteria were compared against one another. Based on these relationships, pair-wise comparison matrices were formed by the application of triangular fuzzy scales.

Five Thai senior directors of apparel manufacturers involved in the expansion or relocation of a manufacturing facility in CLMV countries provided information. These experts were chosen on the basis of their knowledge, background, and goals for the establishment of production plants in the CLMV subregion. In addition, participants were provided with instructions in advance regarding the completion of the matrices and a paired comparison exercise. The reliability of the evaluations is crucial when making use of pair-wise comparisons. Therefore, Equation (15) must be used to obtain the consistency index (CI) and consistency ratio (CR) of comparison matrices. The average random consistency index RI depends on matrix n size and is given by [85]. If CR is 0.1 or less, expert opinions within that matrix are consistent and logical solutions can be found; otherwise, the evaluation process must be examined and reassessed.

$$CR=\frac{CI}{RI}$$

(14)

when

$$CI=\frac{\left({\lambda }_{max}-n\right)}{(n-1)}\mathrm{and}{\lambda }_{max}={\sum }_{i=1}^n\left[{\sum }_{j=1}^n{a}_{ij}{W}_j\right]$$

(15)

where ${\lambda }_{max}$ is the maximum eigenvalue of the matrix and n is the size of the matrix.

3.4. Fuzzy TOPSIS

TOPSIS, or the Technique for Order Performance by Similarity to Ideal Solution, was originally suggested by [86] and further expanded upon by Yoon in 1987 and Tzeng and Huang in 2011 [87,88], in various deliberative scenarios. It determines the best choice based on multiple parameters. TOPSIS was well designed and was reasonable for ranking solutions by their similarity to the optimal one. A positive ideal solution (PIS) maximizes desired advantages while minimizing undesirable expenses, while an anti-ideal solution (NIS) accomplishes the opposite. An ideal solution combines the best feasible values for each criterion, while a negative ideal solution does the opposite. TOPSIS is the leading MCDM technique because of its benefits [89].

The applicability of the TOPSIS method to this inquiry was justified, due to its simplicity and consistent solution procedure, which remains constant even when dealing with various criteria and alternatives. However, in various practical contexts, particularly when considering the viewpoints of individuals involved in the decision-making process, it becomes evident that exact data alone are inadequate for effectively representing real-world situations [90]. Therefore, it is possible for decision-makers to assess difficulties by employing interval judgment and linguistic concepts as opposed to relying solely on a single value [91,92]. In today’s competitive business scene, choosing an appropriate location is a difficult undertaking that requires consideration of numerous factors, including raw materials, labor, and infrastructure. The process of decision-making becomes intricate, due to the need to minimize the influence of uncontrollable elements while enhancing the impact of controllable factors. According to [25,51], the fuzzy approach, especially Fuzzy TOPSIS, is suggested as a useful tool for evaluating location choice, due to its straightforward and easily programmable calculation process. The technique also excels at tackling group decision-making in fuzzy environments. Hence, it can be concluded that the utilization of Fuzzy TOPSIS presents a more pragmatic approach in the context of this study when it comes to the selection of the most suitable location for an apparel manufacturing facility.

The goal of this section is to map linguistic variables to numerical variables by extending TOPSIS to the ambiguous environment presented by [5]. The following is an outline of the procedures involved in the Fuzzy TOPSIS algorithm.

• Step 1: Create a decision matrix and assign weights to each criterion

Typically, decision factors are utilized to rate alternatives using linguistic variables that are then turned into Triangular Fuzzy Numbers (TFNs). The linguistic variables and their associated TFN are displayed in

Table 3. Linguistic values and mean of fuzzy numbers for ranking alternatives.

Importance	Abbreviation	Mean Fuzzy Number
Extremely low or inadequate	VL	0
Low or poor	L	0.25
Moderate to fair	M	0.50
Excellent or high	H	0.75
Extremely high or excellent	VH	1.00

Source: [93,94].

.

The decision matrix, designated as $X^k$, is generated by utilizing the ratings of m alternatives with respect to n criteria provided by K experts, as follows:

$$X^k=\left[\begin{array}{ccc}\begin{array}{c}\begin{array}{cc}{X}_{11}^k& X_{12}^k\end{array}\\ \begin{array}{cc}{X}_{21}^k& X_{22}^k\end{array}\end{array}& \begin{array}{c}\cdots \\ \cdots \end{array}& \begin{array}{c}{X}_{1n}^k\\ X_{2n}^k\end{array}\\ \begin{array}{cc}⋮& ⋮\end{array}& \ddots & ⋮\\ \begin{array}{cc}{X}_{m1}^k& X_{m2}^k\end{array}& \cdots & X_{mn}^k\end{array}\right]$$

(16)

$X_{ij}^k$ represents the kth expert’s evaluation of option i with respect to criterion j, and $X_{ij}^k$ is equal to $\left(a_{ij}^k,b_{ij}^k,c_{ij}^k\right)$.

• Step 2: Evaluate the entire decision matrix

By combining the ratings of K decision-makers with Equation (17), the resulting group decision matrix, $\widehat{X_{ij}}$ can be constructed as follows:

$$a_{ij}={\mathit{min}}_k\left(a_{ij}^k\right),b_{ij}=\frac{1}{k}\sum _{k=1}^K{b}_{ij}^k,c_{ij}={\mathit{max}}_k{c}_{ij}^k$$

(17)

The entire group decision matrix, $\widehat{X_{ij}}$ is shown below:

$$\widehat{X_{ij}}=\left[\begin{array}{ccc}\begin{array}{c}\begin{array}{cc}\widehat{X_{11}}& \widehat{X_{12}}\end{array}\\ \begin{array}{cc}\widehat{X_{21}}& \widehat{X_{22}}\end{array}\end{array}& \begin{array}{c}\cdots \\ \cdots \end{array}& \begin{array}{c}\widehat{X_{1n}}\\ \widehat{X_{2n}}\end{array}\\ \begin{array}{cc}⋮& ⋮\end{array}& \ddots & ⋮\\ \begin{array}{cc}\widehat{X_{m1}}& \widehat{X_{m2}}\end{array}& \cdots & \widehat{X_{nm}}\end{array}\right]$$

(18)

where $\widehat{X_{ij}}=(a_{ij},b_{ij},c_{ij})$.

• Step 3: Determine the normalized fuzzy decision matrix, $\widetilde{p_{ij}}$

In this step, both the positive and negative criteria are determined. Efficiency, for example, is seen as an advantage, whereas expense is viewed as a negative criterion whose value should be minimized.

The normalization criterion for the benefit is expressed as:

$$\widetilde{p_{ij}}=\left(\frac{a_{ij}}{c_j^{+}},\frac{b_{ij}}{c_j^{+}},\frac{c_{ij}}{c_j^{+}}\right),\mathrm{if}jϵG,c_j^{+}={\mathit{max}}_i\left(c_{ij}\right)$$

(19)

In contrast, the non-beneficial criteria are normalized as follows:

$$\widetilde{p_{ij}}=\left(\frac{a_j^{-}}{a_{ij}},\frac{a_j^{-}}{b_{ij}},\frac{a_j^{-}}{c_{ij}}\right),\mathrm{if}jϵH,a_j^{-}={\min }_i\left(a_{ij}\right)$$

(20)

in which G and H represent the advantageous and expense criteria, respectively.

• Step 4: Determine the weighted normalized fuzzy decision matrix, $\widetilde{v_{ij}}$

$\widetilde{v_{ij}}$ is determined by multiplying the normalized matrix by the weight of each criterion as follows:

$$\widetilde{v_{ij}}=\widetilde{p_{ij}}\times w_j$$

(21)

where w_j represents criterion j’s fuzzy importance weight. When categorizing the degree of importance of criteria with the participation of more than one decision-maker, the resulting weight is known as the “combined group criteria weight”. For instance, if K people are involved in the decision-making process, the group criterion weight could be written as follows:

$$W_j^k=\left[w_j^1,w_j^2,w_j^3,\dots ,w_j^k\right]$$

(22)

where $w_j^k$ represents the ambiguous weight that the kth decision-maker assigns to criterion j, and $W_j^k=\left(a_j^{/k},b_j^{/k},c_j^{/k}\right)$.

The weight of the combined group criteria is derived as follows [94,95]:

$$a_j^{/}={\min }_k\left(a_j^{/k}\right),b_j^{/}=\frac{1}{k}\sum _{k=1}^K{b}_j^{/k},c_j^{/}={\max }_k\left(c_j^{/k}\right)$$

(23)

Consequently, the combined fuzzy weight of criterion j is computed as $Z{w}_j=\left(a_j^{/},b_j^{/},c_j^{/}\right)$.

• Step 5: Assess the fuzzy positive ideal solution (FPIS), $A^{+}$, and the fuzzy negative ideal solution (FNIS), $A^{-}$

The procedure for selecting $A^{+}$ and $A^{-}$ is as follows:

$$A^{+}=\left({\tilde{v}}_1^{+},{\tilde{v}}_2^{+},\dots ,{\tilde{v}}_n^{+}\right),\mathrm{where}{\tilde{v}}_j^{+}={\max }_i{\tilde{v}}_{ij}$$

(24)

$$A^{-}=\left({\tilde{v}}_1^{-},{\tilde{v}}_2^{-},\dots ,{\tilde{v}}_n^{-}\right),\mathrm{where}{\tilde{v}}_j^{-}={\min }_i{\tilde{v}}_{ij}$$

(25)

• Step 6: Determine the separation of each possible option from $A^{+}$ and $A^{-}$

The vertex approach can be used to compute the distances of each alternative from $A^{+}$. Given ${\tilde{v}}_{ij}=\left(a_{ij},b_{ij}\mathrm{a}\mathrm{n}\mathrm{d}{c}_{ij}\right)$ and ${\tilde{v}}_j^{+}=\left(a_j^{+},b_j^{+}\mathrm{a}\mathrm{n}\mathrm{d}{c}_j^{+}\right)$, the distance between them or $d\left({\tilde{v}}_{ij},{\tilde{v}}_j^{+}\right)$ can be represented in the following manner [96]:

$$d\left({\tilde{v}}_{ij},{\tilde{v}}_{ij}^{+}\right)=\sqrt{\frac{1}{3}\left[{\left(a_{ij}-a_j^{+}\right)}^2+{\left(b_{ij}-b_j^{+}\right)}^2+{\left(c_{ij}-c_j^{+}\right)}^2\right]}$$

(26)

By employing a similar process, it is possible to determine the distance between each option and $A^{-}$ in the following manner:

$$d\left({\tilde{v}}_{ij},{\tilde{v}}_{ij}^{-}\right)=\sqrt{\frac{1}{3}\left[{\left(a_{ij}-a_j^{-}\right)}^2+{\left(b_{ij}-b_j^{-}\right)}^2+{\left(c_{ij}-c_j^{-}\right)}^2\right]}$$

(27)

With respect to the complete set of decision criteria, the calculation for the distance between each alternative and $A^{+}$ and $A^{-}$ is as follows [97]:

$$Y_i^{+}={\sum }_{j=1}^{n}d\left({\tilde{v}}_{ij},{\tilde{v}}_j^{+}\right),i=\mathrm{1,2},\dots m;j=\mathrm{1,2},\dots n$$

(28)

$$Y_i^{-}={\sum }_{j=1}^{n}d\left({\tilde{v}}_{ij},{\tilde{v}}_j^{-}\right),i=\mathrm{1,2},\dots m;j=\mathrm{1,2},\dots n$$

(29)

• Step 7: Calculate the closeness coefficient ${CC}_i$ for each choice

Each option’s closeness coefficient (${CC}_i$) is calculated by using the following criteria:

$${CC}_i=\frac{Y_i^{-}}{Y_i^{-}+Y_i^{+}}$$

(30)

The options are assessed and ordered according to their ${CC}_i$ numbers, and the alternative with the greatest value is identified as the optimal option.

3.5. Adjusted Mixture Design (AMD)

Organizations employ both fundamental and advanced statistical methodologies, which may include optimization techniques, statistical analysis, and experimental design. Experimental design is a widely used statistical approach that involves conducting experiments with extreme precision to collect data regarding the efficacy of various combinations of the variables under consideration. Experimental design enhances process parameters and input–output variables through the application of statistical methods and the development of models that capture the complex interrelationships between various components and their respective intended outcomes [98,99].

Experimental designs are frequently employed to identify the independent components that minimize quality variation [100]. Through the application of research findings, it is possible to determine the optimal combination that optimizes desired outcomes, frequently while adhering to predetermined constraints. Experimental design is an indispensable method for obtaining system knowledge through the fitting of data to empirical functions. Its practicality is particularly evident in situations involving limited proportions of constituents or when determining the precise blend of elements, substances, or operational variables [101].

Experiments may be structured by employing levels and parameters that are determined by observed products or processes. This enables practitioners and researchers to arrive at informed conclusions through comprehensive examination of different combinations and analysis of their effects on outcomes [102]. Experimental designs differ in the extent to which they require knowledge of the product or process, as well as scenario sensitivity. The systematic approach provided by this method streamlines the process of experimenting, analyzing, and refining combinations, thereby accelerating the identification of optimal solutions. As a result of the researcher’s responsibility to identify the independent and response components, this methodology facilitates the achievement of intended objectives.

Mixture design is a specific form of RSM, wherein the factors are represented by the components of a mixture and the response changes in accordance with the proportions; that is, the response is dependent on the proportional variation. One is the aggregate of the proportions added together. Standard design approaches are inappropriate and cannot be implemented in such circumstances. For a q-component mixture, see Equation (31).

$$0\le x_i\le 1;i=1,2,\dots ,q\mathrm{a}\mathrm{n}\mathrm{d}\sum _{\mathrm{i}=1}^q{x}_i=1$$

(31)

where $x_i$ is the proportion of the ith component in the mixture. The q-components constitute a simplex with a dimension of (q − 1). Several considerations must be taken into account when selecting the most suitable mixture design. These include the quantity of factors and interactions that need to be investigated, the intricacy of each design, the statistical validity and efficacy of each design, the feasibility of implementation, and the financial and temporal limitations associated with each design. The subsequent subsections provide descriptions of the mixture design categories that are most commonly employed.

An extremely prevalent variety of mixture design is the simplex-lattice design, which can be described as follows (Equation (32)). The “{q, m}” nodes comprising the simplex-lattice design for q components are delineated by the subsequent coordinate configurations:

$$x_i=0,\frac{1}{m},\frac{2}{m},\dots ,\frac{m-1}{m},1;i=1,2,\dots ,q$$

(32)

The proportions assumed by each component accept values from 0 to 1 that are equally spaced apart by m − 1, and every conceivable mixture of the ratios specified in Equation (33) is implemented. The standard formula for determining the number of points in a “q, m” simplex-lattice design is given below. Cornell provides exhaustive explanations of the simplex-lattice design [98].

$$\frac{\left(q+m-1\right)!}{m!\left(q-1\right)!}$$

(33)

The simplex centroid design is an option that can be used instead of the simplex-lattice design. It is possible to use a simplex-centroid mixture design when all of the components have the same range (between 0 and 1) and there are no limits on the design space. In any recipe, there is always a “center-point run” that contains equal parts of all the components. The whole and comprehensive strategy of the AMD is as follows:

• Step 1: Define the Major (Ma) and Minor (Mi) of the weight of criteria.
  • Step 1a: Collect the entire Ma weight of criteria ($W_i^{Ma}$) for all possible simplex-lattice or simplex-centroid mixture designs to form the mixture design.
  • Step 1b: Assign the remaining weight of the Minor criteria ($W_i^{Mi}$) in order to modify their newly selected levels in accordance with the Ma criteria.
• Step 2: Ascertain the modified Ma weight of the criteria to satisfy the constraint of $\sum _i^{Ma}{W}_i^{Ma}$.
• Step 3: Determine the updated level of the Mi weight of criteria by utilizing ${\theta W}_i^{Ma}$, where 0 ≤ $\theta$ ≤ 1, and reassess the requirement that the minimum weights of the Ma criteria are greater in magnitude than the maximum weights of the Mi criteria.
• Step 4: Converge and iterate while implementing the Step 4 procedure. Upon accomplishment (${\theta }^{*}$), proceed to Step 5.
• Step 5: To conduct a sensitivity analysis, compute the adjusted Ma (${AdjW}_i^{Ma}$) and Mi (${AdjW}_i^{Mi}$) values of the weights of criteria, as follows:

  $${AdjW}_i^{Ma}=W_i^{Ma}+{\theta }^{*}\sum W_i^{Mi}; \mathrm{for} \mathrm{all} Ma \mathrm{criteria} \mathrm{weights}$$

  (34)

  $${AdjW}_i^{Mi}=({1-{\theta }^{*})W}_i^{Mi}; \mathrm{for} \mathrm{all} Mi \mathrm{criteria} \mathrm{weights}$$

  (35)

4. Illustrative Results

This section examines empirical data and analyses that support the proposed model’s ability to integrate management preferences, judgment, and complex multiple-choice criteria. It shows how this method helps foreign and Thai decision-makers integrate into the CLMV region’s organizational framework. This study’s findings are given below.

4.1. Identification of Factors Influencing International Location Decisions

Following the survey, the key criteria weights were determined using the ROC weighting formulas shown in Equation (1).

Table 4. Weight of primary criteria using ROC and their respective rankings.

Criteria	${\mathit{W}}_{\mathit{j}}^{\mathit{R}\mathit{O}\mathit{C}}$	Rank
Logistics System	0.151	3
Risk	0.016	8
Labor	0.340	1
Raw Materials	0.215	2
Economic condition	0.054	6
Infrastructure	0.111	4
Government Policies	0.079	5
Location Facilities	0.033	7

shows the weights of those eight location selection criteria and their rankings.

In this study, the top six ranking criteria from Table 4 were selected for inclusion in the proposed model. The description of these six criteria and their sub-criteria are provided in

Table 5. Importance criteria together with the details of each sub-criteria.

Criteria	Sub-Criteria	Description
Labor (R1)	Low labor cost (R11)	Labor cost includes the sum of all wages, benefits, and payroll taxes paid to employees.
	Labor availability (R12)	The primary location selection driver is the availability of a qualified workforce after the enterprise has considered its logistics network.
	Flexibility of wage determination (R13)	Wage flexibility is a significant aspect of labor market flexibility, which is a key requirement for stable macroeconomic development. It is very critical in managing hiring and firing practices.
	High labor skill (R14)	Skilled labor is capable of assuming sophisticated duties and have the aptitude to adapt quickly to changes and routine job functions, resulting in productivity for any business.
Raw materials (R2)	Material quality (R21)	Material quality is essential for ensuring high-quality products to meet all the requirements and expectations of the consumers, increasing customer demand.
	Supplier quality (R22)	The quality of suppliers plays a significant part in taking responsibility for delivering an extraordinary quality of raw materials to commit to the end customer’s anticipations.
	Material availability (R23)	Material availability and input reliability shape productivity and have the capability of execution of an allocated assignment at a given time.
	Supplier reliability (R24)	The reliability of suppliers could provide the most suitable materials at the right periods when selecting a location site.
Logistics system (R3)	Low-cost transport (R31)	Transport costs include the total costs of the transport system and its use. Low transportation costs will help companies in reducing the costs of logistics, and thus the total costs of goods and services can be decreased.
	Transport lead time (R32)	Transportation lead time is the time required for preparing products from locations where they are sourced to locations where they are demanded. Reducing lead time in manufacturing can increase output and revenue.
	Ease of transport document (R33)	Transport documents give the kind of information about cargo that is being transported, such as air waybill, bill of lading, customs documents, etc. Reducing transportation time for requirement documents would help products to transport to the end customers more quickly and effectively.
	Low investment in import/export (R34)	Low investment in import/export could help businesses in reducing costs, and grow and expand national economies.
Infrastructure (R4)	Available of road transport (R41)	Available road transport provides the movement of goods and materials to reach customers within a designated time.
	Reliability of road transport (R42)	If the required transportation is not reliable, it can have an adverse consequence on the lead time and responsiveness.
	Cost of vehicle transport (R43)	The cost of vehicle transport is an important factor in determining how much it costs to ship products to end customers.
	Connectivity (R44)	Transport infrastructure is vital for connectivity between countries and aids in supporting international trade, growth, and regional integration.
Government policies (R5)	Government security (R51)	Government security indicates the situation of the host country. It is preferable to locate facilities in transparent, safe, and secure countries.
	High government potential (R52)	One important factor for seeking location decisions is the high potential of local government, which will help to support investors in doing business in such a country.
	Law enforcement (R53)	The local government can encourage business location by using law enforcement to support investors.
	Funding for infrastructures project (R54)	Funding for infrastructure project investment by the host country would enhance the success of the business of the investors.
Economy condition (R6)	Exchange rate (R61)	The exchange rate between two currencies is the amount for which one currency will be exchenged for another.
	High-rate inflation (R62)	Inflation is the percentage increase or decrease in prices during a specified time. A country with a high inflation rate will not be suitable for locating a facility.
	Interest rate spread (R63)	Interest rate is a significant factor if the company would like to use loans for constructing the facility or for its operations.
	Domestic market size (R64)	Domestic market size is an important factor that companies must be considered when locating their facilities abroad.

. The results were then used to develop a theoretical research framework to locate promising locations, as shown in Figure 3.

4.2. Using Fuzzy ANP to Weight Criteria and Sub-Criteria

After five experts evaluated the pair comparison matrices using verbal expression, they were transformed into TFNs. This section utilizes the Fuzzy ANP to determine the factor and sub-factor weights. Here are some illustrations of the proposed model’s computation procedure and its outcomes:

Step 1: All experts were asked to rank criteria and sub-criteria by comparing them in pairs, i.e., the relevance of the i criterion over the j criterion, where i, j = 1, 2, …, n. By doing this, a fuzzy pair-wise comparison matrix A is made.

Step 2: Due to more than one expert, the geometric average approach was used to calculate the aggregate fuzzy judgment matrix A˜^∗ [72]. The average pair-wise comparison of all experts is computed using Equation (36) as follows:

$${\tilde{r}}_{ij}={({\tilde{a}}_{ij1}\otimes {\tilde{a}}_{ij2}\otimes \dots .\otimes {\tilde{a}}_{ijk})}^{1/k}$$

(36)

when ${\tilde{a}}_{ijk}$ is the importance weight of comparing row $i$ and column $j$ of all experts.

Step 3: Calculate the consistency ratio (CR) from the average pair-wise comparison; CR is less than 0.1, indicating the consistency of expert opinion.

Step 4: Compute the fuzzy evaluation matrix and the local weights as follows:

$$S_{Ri}=e\otimes f=\sum _{j=1}^m{M}_{gi}^j\otimes {\left[\sum _{i=1}^n\sum _{j=1}^m{M}_{gi}^j\right]}^{-1}$$

(37)

Then, calculate the value degree of possibility of a convex fuzzy number, using Equation (9).

Step 5: Calculate inner dependence weight by considering criteria of weight by cutting the row and column of that criteria. For example, the weight of inner dependence on labor criterion (R₁) made by cutting the row and column of raw material criterion (R₂). Then, calculate the weight in the same manner as the weight of the local weights, as in Step 4.

Step 6: Multiply the local criteria weights by the inner dependence matrix to obtain the interdependent criteria weights.

Step 7: Then, by multiplying the sub-criteria’s local weight by the weight of the criterion ($W_i$), the global weights of the sub-criteria could be analyzed. Computed values based on the experts’ opinions are shown in

Table 6. Global weight of sub-criteria.

Criteria (i)	$\mathrm{Weight}({\mathit{W}}_{\mathit{i}}$)	Sub-Criteria	Local Weight	Global Weight
R1	0.31	R11	0.23	0.07
		R12	0.23	0.07
		R13	0.29	0.09
		R14	0.26	0.08
R2	0.28	R21	0.74	0.21
		R22	0.00	0.00
		R23	0.15	0.04
		R24	0.11	0.03
R3	0.12	R31	0.49	0.06
		R32	0.22	0.03
		R33	0.00	0.00
		R34	0.29	0.03
R4	0.14	R41	0.28	0.04
		R42	0.20	0.03
		R43	0.27	0.04
		R44	0.25	0.04
R5	0.03	R51	0.22	0.01
		R52	0.07	0.00
		R53	0.07	0.00
		R54	0.64	0.03
R6	0.12	R61	0.25	0.03
		R62	0.31	0.04
		R63	0.08	0.01
		R64	0.35	0.04

. These weights would subsequently be employed in the subsequent phase’s Fuzzy TOPSIS technique evaluation of the candidate country.

The five experts were asked to compare these sub-factors between each alternative nation using linguistic variables. The alternatives’ overall priorities were then determined using the linguistic variables in Table 3. Since all factors and sub-factors in this investigation were cost criteria, Equation (20) was utilized to normalize the decision matrix. The normalized decision matrix was multiplied by the global weight of each Fuzzy ANP sub-factor in Equation (21) to obtain the weighted normalized decision matrix. After performing the calculations outlined in Equations (22)–(30), the relative ranking of the three alternatives was determined by sorting the proximity coefficient values in descending order. According to the latest findings, Vietnam was the most suitable area, followed by Myanmar and Cambodia, as displayed in

Table 7. The final weighting and ranking of choices for location.

Country	Score	Rank
Cambodia	0.226	3
Myanmar	0.399	2
Vietnam	0.401	1

.

5. Sensitivity Analysis via Adjusted Mixture Design

A critical component that must be assessed when evaluating the suitability and reliability of the presented MCDM approach for determining the optimal CLMV subregion site for the garment industry is sensitivity analysis. By employing an adjusted mixture design and conducting scenario assessment, we shall examine the model’s responsiveness to variations in the relative importance of the criteria outlined in this section. By conducting a sensitivity analysis, one can determine whether the results remain consistent in different contexts and in relation to the inclinations of management decision-makers. An essential component of the proposed model is its incorporation of management personnel’s preferences and judgments at every stage of the decision-making procedure. The objective of the conducted sensitivity analysis via the adjusted mixture design (AMD) with the optimal adjusted parameter (${\theta }^{*}=1/3$) was to enhance comprehension regarding the potential impact of diverse managerial inputs or preferences on the ultimate decision. The task in hand entailed assessing the model’s adaptability to changing preferences while ensuring its ongoing utility for decision-makers.

5.1. Importance Criteria and Their Impact

Among the advantages of the proposed model is its capacity to effectively handle intricate interdependencies among the criteria that are depicted by multiple-choice options. The objective of the sensitivity analysis was to ascertain the impact of variations in the strength of these interrelationships on the site selection process. This ensures that the model’s capability to accurately represent the dynamics of the CLMV subregion alternatives will persist. Furthermore, a scenario analysis was conducted to examine the collective impact of multiple elements, in addition to determining the degree to which individual changes affect the model. To accomplish this, a multitude of fictitious scenarios were constructed to mirror actual conditions, and the model’s performance was evaluated in accordance with these diverse scenarios.

The results obtained from the sensitivity analysis provide evidence that the proposed MCDM approach is dependable and robust across a range of circumstances. It permits adjustments to be implemented in the weights of criteria and sub-criteria, input data, management preferences, and intricate interrelationships. The model’s robustness and resilience in the context of evolving circumstances render it a valuable instrument for decision-makers within the CLMV organizational framework who are tasked with identifying suitable sites in foreign nations. In summary, the sensitivity analysis proves that the model is practical and effective, providing decision-makers with a robust tool for making informed and reliable site selections in the dynamic garment industry subregions of the CLMV (

Table 8. A total of 16 scenarios generated via the adjusted mixture design (AMD).

Scenario	R1	R2	R3	R4	R5	R6
1	0.790	0.060	0.040	0.060	0.010	0.040
2	0.547	0.060	0.040	0.303	0.010	0.040
3	0.303	0.060	0.040	0.547	0.010	0.040
4	0.060	0.303	0.040	0.547	0.010	0.040
5	0.303	0.547	0.040	0.060	0.010	0.040
6	0.547	0.303	0.040	0.060	0.010	0.040
7	0.547	0.182	0.040	0.182	0.010	0.040
8	0.182	0.182	0.040	0.547	0.010	0.040
9	0.182	0.547	0.040	0.182	0.010	0.040
10	0.303	0.303	0.040	0.303	0.010	0.040
11	0.060	0.790	0.040	0.060	0.010	0.040
12	0.060	0.060	0.040	0.790	0.010	0.040
13	0.425	0.425	0.040	0.060	0.010	0.040
14	0.425	0.060	0.040	0.425	0.010	0.040
15	0.060	0.425	0.040	0.425	0.010	0.040
16	0.060	0.547	0.040	0.303	0.010	0.040

).

5.2. Comparative Analysis with Alternative Scenarios

The weight designations in Scenario 1 demonstrate the subsequent preferences regarding the CLMV countries: Vietnam (0.375) is identified as the preferred location, indicating that, among the three nations considered, it is the most appealing option for apparel manufacturers in this particular circumstance. The Myanmar rating of 0.235 is ranked second. Cambodia (0.120) is the least-favored option, due to its lightest weight. With a significance score of 0.79, labor (R₁) is deemed to have the most substantial impact on the decision-making process. An increased value indicates that the quality and accessibility of the labor force are pivotal elements. In comparison to labor, the importance score of 0.06 for raw materials (R₂) is comparatively low. This implies that the availability and cost of basic materials are not as crucial in this situation. Similar to raw material, logistics system (R₃) and infrastructure (R₄) each have an importance score of 0.06. The weightage assigned to these factors is comparatively lighter in the given scenario.

The economic climate (R₆) and the government policies (R₅) each have an importance score of 0.04, signifying that they exert a moderate level of influence on the decision. It is critical to take into account the state of the economy and the stability of the government. Vietnam is the most preferable option in Scenario 1, notwithstanding the fact that labor (R₁) has the maximum importance score, on account of its weight assignment. This implies that, within the confines of this situation, the remaining criteria substantially augment Vietnam’s appeal as a prospective site for apparel production. Myanmar, which was assigned the second-highest weight, is regarded as the secondary option, whereas Cambodia, owing to its lower weight, is regarded as the least favored (Figure 4).

In summary, for other scenarios (Figure 4), Scenario 2: Vietnam is proposed as the location, and R₂ and R₄ are the most crucial aspects to consider when selecting the ultimate result of this option. Scenario 3: Vietnam is selected as the best possible location, and R₄ is an important factor in the decision-making process. Scenario 4: Vietnam is brought up as a potential location, and both the R₂ and R₄ scores are taken into consideration before making a final decision. Scenario 5: Myanmar is chosen as the location of preference, and R₂ is the key criterion that plays a role in selecting the final decision. Scenario 6: it has been suggested that Vietnam be selected as the location for the facility, and R₂ will have some input on the decision-making process. Scenario 7: Vietnam is the preferred choice, and R₁ and R₃ are taken into consideration during the decision-making process. Scenario 8: it has been suggested that Vietnam be selected as the location for the site, and R₄ will be the key consideration in making the ultimate decision. Scenario 9: Myanmar is the leading contender for the location, and R₂ is an important factor in the decision-making process.

Scenario 10: Vietnam is the best available choice, and the decision was influenced by a variety of factors, the most important of which were R₁, R₂, and R₆. Scenario 11: Vietnam is the best choice, due to the high weight that was awarded to it, which was mostly decided by R₂’s performance. Scenario 12: R₄ played a crucial part in the decision-making process, which ultimately led to Vietnam being chosen as the choice that provides the greatest benefit. Scenario 13: the location of Myanmar is proposed as the best option, owing to several factors, the most important of which are R₄ and R₂. Scenario 14: Vietnam is the best choice, with the decision being influenced by several different factors, including both R₁ and R₄. Scenario 15: it is determined that Myanmar is the best location for the facility, and factors such as R₂ and R₄ are taken into account in making this decision. Scenario 16: Myanmar is the location that is recommended, due to the comparatively large weight assignment that it was given, which was influenced by R₂ and R₄ (Figure 5).

Regarding scenario variations, an individual set of criteria and their influence on the decision are underscored in each scenario. For instance, Scenario 1, Scenario 2 and Scenario 3 emphasize the criticality of R₄, whereas Scenario 9 prioritizes R₂. These discrepancies indicate that the process of decision-making is malleable and susceptible to variation. Vietnam emerges as a formidable contender in the majority of situations. The substantial weight ascribed to Vietnam is predominantly attributable to factors such as R₁ and R₂. To comprehend why Vietnam is favored, it is critical to comprehend the significance of these particular factors. Myanmar is regarded as a potential substitute site in certain circumstances, specifically when considerations such as R₂ and R₄ are given priority. The justification for Myanmar’s applicability should be investigated, given that it could provide distinct benefits under particular conditions.

Regarding the function of particular criteria, the scenarios illustrate the function of a variety of criteria. For example, R₄ is commonly referenced as a pivotal factor in numerous situations. This implies that various elements such as market accessibility, infrastructure, and economic stability (depending on the value of R₄) significantly influence the ultimate determination of the location. Certain situations (e.g., Scenario 10) illustrate that the determination is not solely determined by one element, but rather is impacted by an amalgamation of components. It is crucial to comprehend the manner in which these elements interact and impact the decision. It is emphasized in the conclusion that the process of decision-making is not fixed. It is capable of adjusting to shifting priorities or conditions. Establishing a mechanism that allows for the re-evaluation of factor weighting in response to changing circumstances is of utmost importance. The ultimate objective in numerous situations is to determine the location that offers the most advantageous outcomes. It is essential to the decision-making process that these benefits, whether they pertain to cost, market access, or other factors, be precisely defined and quantified.

5.3. Implications for Decision-Makers

Sensitivity analysis indicates that Vietnam is frequently an outstanding rival when decision-makers take labor (R₁) and raw materials (R₂) into account. In light of the labor-intensive nature of the industry, it is crucial for the sector to maintain a strategic emphasis on these criteria. For businesses looking for affordable manufacturing, Vietnam is a desirable alternative, due to its competitive labor prices. This particular component has a substantial role in generating cost reductions for enterprises in Thailand. Furthermore, Vietnam’s skilled labor pool is rising, giving enterprises access to talented workers. Skilled personnel and lower prices boost production efficiency and productivity. At the moment, Vietnam can have access to a wide variety of raw materials, which are required for the production processes of various goods. This assures a consistent and cost-effective supply chain, which is essential for sectors that rely on particular commodities. The resilience of the industry is contingent upon its ability to effectively procure a dependable workforce and obtain raw materials of superior quality, hence mitigating disruptions and maintaining a consistent level of production. These essential factors are readily accessible within the context of Vietnam.

Myanmar possesses a distinct advantage in terms of raw materials in the majority of instances (Scenarios 5, 9, 13 and 15). One possible explanation for this phenomenon is the consistent growth of Myanmar’s textile and garment sectors, which has subsequently garnered attention and investment from both domestic and international enterprises. Several international apparel companies and manufacturers have explored or established operations in Myanmar. The utilization of raw resources (R₂) has the potential to reduce lead times and enhance operational efficiency across the whole supply chain. As a result, the reduction in dependence on imports and the associated costs of transportation can be achieved. The utilization of local supply chains presents a compelling alternative for businesses looking to maximize their financial benefits, as it can enhance production efficiency and reduce costs. Furthermore, in Scenarios 13 and 15, Myanmar exhibits a discernible competitive advantage under specific situations in both R₂ and R₄. The success of manufacturing operations is contingent upon not only the procurement of raw materials, but also the quality and efficiency of a nation’s infrastructure, including its transportation networks. Factors such as connectivity, affordability, availability, and reliability play a crucial role in this regard (R₄). Due to its economic benefits in terms of connection to both the EU market and Thailand for the purpose of importing goods, Myanmar may prove to be a desirable location for operations by Thai garment manufacturers.

6. Discussion and Conclusions

The integration of Fuzzy ANP, Fuzzy TOPSIS, and Adjusted Mixture Design-based Scenario Analysis in our study presents a robust methodology for discerning optimal locations within the CLMV subregion for the apparel industry. As we assessed Cambodia, Myanmar, and Vietnam as potential sites, our model underwent four distinctive phases. Initially, a survey pinpointed factors influencing international apparel location decisions, succeeded by interviews with key executives from Thai and CLMV apparel firms to scrutinize potential site alternatives. The Fuzzy ANP model played a pivotal role in constructing hierarchies and linkages, thereby establishing criteria and sub-criteria weights for global location selection. Additionally, our methodology incorporated Adjusted Mixture Design-based Scenario Analysis to enhance the comprehensiveness of our decision-making framework. This dual approach allowed us to not only account for the inherent uncertainties in decision-making but also to evaluate the model’s performance across various scenarios, ensuring its adaptability and reliability.

Our findings suggest that by combining Fuzzy ANP and Fuzzy TOPSIS-based selection of location modeling, the optimal site for the CLMV apparel industry might have been determined. In the CLMV subregion, Cambodia, Myanmar, and Vietnam were all considered as potential replacements. The proposed model was created over a sequence of four stages. First, a survey was utilized to determine factors influencing apparel international location considerations. Second, key executives from apparel firms in Thailand and the CLMV were interviewed and extensive investigations were conducted to analyze possible CLMV subregion site alternatives. Additionally, the Fuzzy ANP model was used to build hierarchies and linkages within and across levels and calculate the weights of criteria and sub-criteria for the apparel industry’s worldwide location selection. Finally, Fuzzy TOPSIS calculated the three candidates’ ranking, based on their preferences.

According to this study, Vietnam is an appealing CLMV subregion choice for Thai firms looking to relocate or expand their garment manufacturing operations. These findings agree with [8,38], which found Vietnam was suitable for ASEAN textile manufacturing facility expansion. For years, Vietnam has had one of Asia’s fastest and most stable economies. Vietnam’s FDI was USD 19.74 billion in 2021 [103]. Textiles and garments contribute significantly to Vietnam’s GDP and exports. Geography makes transporting products overseas easier in Vietnam. Exporters like the apparel industry benefit from its proximity to major Asian markets and transportation routes. Transportation and infrastructure are expanding and modernizing nationwide [104]. According to [105], the textile and garment industry employed 2.5 million people, 12% of the industrial sector’s labor force. This implies that Vietnamese citizens are skilled in this field. Vietnam’s apparel industry also outperforms Myanmar and Cambodia in worker productivity. Vietnam’s competitive garment manufacturing industry attracts investment due to its trained workers, huge young population, and strong supplier base. Vietnam also signed the EU–Vietnam Free Trade Agreement and CPTPP. These agreements facilitate Vietnam’s entry into key markets. These facts and the evidence were also confirmed by research by [106], which discussed the elements that make Vietnam an appealing destination for investments seeking to establish a new global manufacturing base. As a result of these facts, Vietnam was identified as the leading location for CLMV garment manufacturers.

Selecting the best CLMV location for garment manufacturing is a complex process influenced by several factors. To maximize the benefits of the chosen location and adapt to changing circumstances, sensitivity analysis with adjusted mixture design was used to evaluate these criteria. Despite Vietnam’s potentially, Myanmar was considered under certain conditions. Thai investors may favor Myanmar if raw materials are most important. Even though Myanmar might not have enough raw materials on its own to support the global apparel sector, its proximity to Thailand and other nations like China gives it a competitive edge. It can give Myanmar-based manufacturers of apparel access to a wide variety of suppliers and sources of raw materials. This can facilitate the establishment of a reliable and consistent supply of the necessary supplies. In accordance with [107], aligning industrial, regional, and migration policies may help to gradually lower the proportion of foreign laborers in Thailand’s garment and textile sector and promote the transfer of labor-intensive production operations to nations that export labor. Both labor-sending and labor-receiving nations can advance industrialization by pooling their production networks. Additionally, bilateral and regional trade agreements may make importing raw resources from neighboring nations easier and cheaper [108].

Cambodia remained Thai investors’ least-preferred option, according to sensitivity analysis. There may be various factors, such as garment industry workers’ rights and circumstances in that country. Cambodia’s infrastructure and talent base growth creates substantial challenges for businesses. These challenges are exacerbated by a lack of product and market diversification, an overreliance on imported raw materials, and the need to transition from a cut-make-trim (CMT) to a free-on-board (FOB) business model [109]. Cambodia’s export and import procedures are disrupted by the country’s weak logistics systems and infrastructure, which in turn increases the overall supply chain’s lead time [110]. These results are consistent with those of [8], which ranked Cambodia as the least attractive ASEAN destination for relocating textile manufacturing plants.

The primary contributions of the current work are as follows. This study contributes to the growing body of knowledge in the field by proposing a novel methodology for location selection through the implementation of an integrated model incorporating Fuzzy ANP and Fuzzy TOPSIS. The methodology has the capacity to consider any ambiguity present in the decision-making process, while also incorporating interactions that occur between each level. Furthermore, the proposed model was structured to encompass all relevant criteria and sub-factors in the selection of overseas locations, derived from the practical knowledge and expertise of the apparel sector. An essential addition of the present study is the simultaneous consideration of six criteria and twenty-four sub-criteria in the suggested model. The utilization of the Fuzzy ANP technique allows for the comprehensive analysis of the interrelationships among criteria and elements. Additionally, this technique effectively captures the inherent ambiguities found in the preferences of management judgments, resulting in a more accurate reflection of reality. Survey research and data collecting from key specialists in the pair-wise comparison process reflecting their preferences and screening alternative sites are also significant. Our methodology can be applied across several industries attempting to enhance their international manufacturing capabilities. When making investment decisions, Thai apparel investors might take these findings into account, along with their own company’s needs and long-term objectives. However, it is imperative to consistently observe and investigate the ever-changing political, economic, and social circumstances in these countries before making any determinations or engaging in any activities.

Nonetheless, this investigation is not without its constraints. For example, in a pair-wise comparison, the participation of only five experts is required. In addition, it is imperative to conduct a comparative analysis of the study’s findings alongside alternative methodologies, such as PROMETHEE II, to assess the procedure’s practicality and fairness. Furthermore, considering the dynamic nature of political, economic, and social factors in the CLMV subregion, it would be beneficial for future studies to include an analysis of the dynamic environmental conditions. This would necessitate ongoing surveillance of regional developments, enabling adjustments to the decision-making framework in real time. An analysis of the ways in which these dynamic factors impact the adaptability and resilience of the proposed model would yield significant knowledge.

By extending the participation of stakeholders beyond executive-level individuals and incorporating a wider range of viewpoints, including those of government officials, local communities, and environmental experts, a more comprehensive comprehension of the decision-making environment can be achieved. An examination of the ways in which various stakeholder concerns and interests impact location determinations would enhance the decision-making framework. Conducting subsequent research that specifically examines the enduring viability of investments made in the apparel industry within the chosen CLMV locations would enhance the practical efficacy evaluation of the suggested methodology. Monitoring critical performance indicators (KPIs) for a prolonged duration would provide significant insights regarding the viability and achievement of the selected sites.

In light of the escalating concern for sustainability in international business operations, forthcoming investigations may consider incorporating sustainability criteria into the decision-making framework. Evaluating potential locations in terms of their environmental impact, social responsibility, and ethical implications would be in accordance with current business trends and facilitate the making of more conscientious decisions. By broadening the purview to encompass a comparative analysis of various industries within the CLMV subregion, a more comprehensive outlook on the challenges associated with location selection could be achieved. Gaining an understanding of the ways in which the distinctive attributes of the apparel industry coincide or deviate from those of other sectors can yield significant insights across industries.

With the ongoing progression of technology, an examination of how emergent technologies, including machine learning and artificial intelligence, can be incorporated into the decision-making process may bolster the predictive capabilities of the model. This has the potential to facilitate more advanced examinations of the interconnections among criteria and possible future situations. Given the interdependence of global economies, it may be fruitful to investigate how changes in the broader economic environment may affect the viability of particular locations. Conducting a comprehensive examination of worldwide economic patterns and their ramifications for the clothing sector in the CLMV subregion would be required.