Evaluating Sustainable Intermodal Transport Routes: A Hybrid Fuzzy Delphi-Factor Relationship (FARE)-Axial Distance Based Aggregated Measurement (ADAM) Model

Abstract

Intermodal transport (IT), which implies the combination of several different types of transport to achieve a more efficient and economical movement of goods, is of increasing importance in modern supply chains. In the conditions of globalization, growth of trade flows and increasingly pronounced requirements for sustainability, effective planning and management of intermodal routes have become crucial, which is why their evaluation and ranking are essential for making strategic and operational decisions. Accordingly, this paper aims to identify the most favorable alternative for developing intermodal transport. Deciding on the choice of the most important intermodal route requires consideration of a large number of criteria, often of a mutually conflicting nature, which places this problem in the domain of multi-criteria decision-making (MCDM). Accordingly, this paper develops a hybrid decision-making model in a fuzzy environment, which combines fuzzy DELPHI (FDELPHI), fuzzy factor relationship (FFARE), and fuzzy axial-distance-based aggregated measurement (FADAM) methods. The model enables the identification and evaluation of relevant criteria, as well as the ranking of defined variants under the requirements and attitudes of various stakeholders. The practical application and effectiveness of the developed model were demonstrated and confirmed by a case study for Bosnia and Herzegovina (B&H). The sensitivity analysis showed that even with changes in the weights of the criteria or the elimination of the most important criteria, the solution remains consistent and reliable. This indicates the robustness of the model and suggests that changes in the parameters do not lead to significant changes in the final results. This confirms the validity of the proposed model and increases confidence in its applicability in practice.

1. Introduction

The modern business environment, marked by intense globalization, increased movement of goods, and growing demands for sustainable development, is driving significant changes in the management of logistics and transport systems. The rising need for faster, more reliable, and environmentally friendly freight transport has underscored the growing importance of intermodal transport (IT) as an integral part of contemporary supply chains. IT is based on the use of at least two different modes of transport (e.g., road, rail, maritime) within a single transport chain, with minimal handling of goods during mode transfers. This approach enables cost optimization, improved time efficiency, and reduced negative external effects, particularly greenhouse gas emissions [1,2]. IT can be conceptualized as a network composed of nodes and links, where nodes represent various types of terminals (e.g., rail, road, maritime, and inland ports), and links denote the transport corridors connecting them. The efficiency of an intermodal system largely depends on the functionality and capacity of these nodes, as well as the quality, availability, and reliability of the connecting links. For the network to operate optimally, its infrastructural and operational components must be well-developed and harmonized, enabling reduced transport time and costs while enhancing sustainability and safety. The main objective of this paper is defined based on the network-based nature of IT systems and their importance in connecting to international logistics flows. In this regard, the research presented in this paper aims to address identified gaps by defining and evaluating major IT routes for Bosnia and Herzegovina (B&H). By applying a modern MCDM model that integrates the fuzzy DELPHI (FDELPHI), fuzzy factor relationship (FFARE), and fuzzy axial-distance-based aggregated measurement (FADAM) methods, the study enables the identification and ranking of IT routes based on a broad set of criteria and perspectives from various stakeholder groups. The methodological framework of this paper is based on the principles of uncertainty and expert subjective knowledge, enabling a more realistic analysis under conditions of limited precise data. Given the complexity and multidimensional nature of decision-making in IT, this approach supports more informed and sustainable decisions tailored to the specific context of B&H. This paper aims to evaluate the key IT routes for B&H, with a focus on analyzing seven route variants that connect B&H with major European trade and logistics hubs. Implementing IT chains in B&H presents a particular challenge due to the level of IT system development, network density, service availability, and other factors [3]. The contribution of this research lies in the analysis of seven key routes for IT development that link B&H to important European trade and logistics centers. The evaluation is carried out using economic, technical-legal, and environmental groups of criteria, from the standpoint of different stakeholder groups. A specific contribution of the study is the development and application of a hybrid MCDM model that integrates the DELPHI, FARE, and ADAM methods within a fuzzy environment. The paper is structured as follows: Section 2 provides a literature review on IT and MCDM methods used for evaluating IT routes. Section 3 presents the methodological framework, focusing on integrating the FDELPHI, FFARE, and FADAM methods. Section 4 outlines the most relevant transport routes for IT development in B&H and the evaluation criteria, while Section 5 discusses the model application results and sensitivity analysis. Finally, Section 6 and Section 7 summarize key findings and recommend improving intermodal policy and planning in B&H.

While each of the component methods (FDELPHI, FFARE, and FADAM) has been applied in earlier research, the novelty of this study lies in their structured integration. This hybridization enables the validation of expert-derived criteria, the modeling of interrelationships among criteria, and a robust, non-compensatory ranking process, addressing the limitations of more conventional hybrids such as FDELPHI–TOPSIS or fuzzy AHP–MARCOS.

2. Literature Review

The existing literature highlights numerous benefits of IT systems—from increased energy efficiency and reduced CO₂ emissions [4,5] to improved safety and cost rationalization [6,7]. Applying mathematical modeling, a hybrid optimization approach, and the MARCOS multi-criteria decision-making (MCDM) method, Kovač et al. [8] propose a new variant of the dry port (DP) concept and analyze its potential to enhance logistics networks in the Danube region. Authors such as Tsamboulas et al. [9] and Rizzoli et al. [10] emphasize the importance of understanding supply chain stakeholders’ needs, simulating flows, and testing various terminal management scenarios. Although there are studies of IT focused on specific regional contexts, they are mostly dealing with the IT development scenarios [11], modelling IT systems [8], and the establishment of terminals as the nodes in the IT networks, while the evaluation of IT routes in Western Balkan countries is underexplored.

Selecting an efficient IT route can lead to cost and time optimization, reduced delays, lower risk of damage, and diminished environmental impact, while simultaneously increasing overall reliability. The evaluation of IT routes is considered crucial for IT network planning [12,13]. The goal of IT route evaluation is to enhance the performance of the entire transport system in terms of cost, time, environmental protection, reliability, and other factors. Numerous authors have explored the evaluation of IT routes. Chang [14] emphasizes that this evaluation is highly complex due to the multiplicity of goals, the mode of transport, delivery time, and economies of scale. Duan and Heragu [15] analyzed an intermodal network in the United States that includes highways, railways, and inland waterways across 15 states, aiming to identify IT routes that reduce cost, time, and CO₂ emissions. Among the factors considered in IT route evaluation, cost and time are consistently highlighted as the most critical [16]. McGinnis [17] defines key decision-making criteria when selecting IT routes, including transportation costs, reliability, transit time, handling of overage, shortage, and damaged freight, shipper market, carrier characteristics, and product features. His analysis revealed that transit time was deemed more important than cost. Barnhart and Ratliff [18] study various aspects of IT, including transport mode combinations (truck, rail, ship), and emphasize the importance of time factors, cost, infrastructure availability, and other relevant parameters. Using mathematical modeling, their work focuses on optimizing route selection to achieve efficient and sustainable logistics operations. Moon et al. [19] analyzed six IT routes between Korea and Europe, applying both quantitative (distance, transit time, cost) and qualitative (service, safety, flexibility) criteria, highlighting the complexity and need for a multi-criteria decision-making approach.

Several other authors further deepen the understanding of the criteria relevant to IT route evaluation. Bookbinder and Fox [20] investigated the optimal intermodal freight routes between Canada and Mexico within the context of the North American Free Trade Agreement (NAFTA). The authors examined various factors influencing route selection, including costs, delivery time, availability of transport networks, and political issues related to NAFTA. Pham and Yeo [21] conducted a study aimed at assessing competing door-to-door transport routes from Shenzhen (China) to Hai Phong (Vietnam) from the perspective of logistics service providers and freight forwarders. The results revealed that the most significant criteria included reliability, cost, capacity, and transport time. When evaluating the IT routes, Kim et al. [22] defined the reduction of transport distance, reduction of transport time, reliability, transport costs, infrastructure availability, customs procedures, political stability, and security as relevant criteria. Similarly, Wang and Yeo [23] list time, cost, route reliability, capacity of different transport modes, safety, and environmental impact as key criteria for evaluating IT routes. On the other hand, Collison [24] categorized IT route evaluation criteria into service timeliness, infrastructure and equipment, cost, and marketing services. Boardman et al. [25] aimed to identify the most favorable IT routes that minimize transport costs and time, taking into account factors such as distance, transport mode capacity, and delivery time. Methodological approaches to IT route evaluation also vary depending on problem complexity and data availability. Idri et al. [26] proposed a universal approach to studying IT routing problems using the shortest path method. Another group of authors applies mathematical algorithms, including mixed-integer programming and dynamic programming [27,28,29]. Tian and Cao [30], as well as Sun et al. [31], introduced a generalized interval fuzzy mixed-integer programming model to address decision-making under uncertainty.

Despite the substantial body of literature dedicated to IT, no existing study has focused on a systematic evaluation of IT routes in the context of B&H. This study aims to fill that research gap by identifying and assessing the main intermodal corridors relevant for the development of B&H’s transport system.

A significant portion of the literature also focuses on MCDM methods, which enable the incorporation of numerous quantitative and qualitative factors. Bostel and Dejax [32] and Macharis et al. [33] used operations research methods to evaluate IT routes. Kengpol et al. [34] developed an MCDM model that combines data envelopment analysis (DEA), analytic hierarchy process (AHP), and zero-one programming for route selection between Thailand and Vietnam, while Koohathongsumrit and Luangpaiboon [35] selected the optimal route from Thailand to Cambodia using zero-one programming and risk analysis.

Several authors have employed MCDM methods and combinations of various MCDM techniques in the evaluation of IT routes. Moon et al. [19] assessed six IT routes between Korea and Europe using the technique for order of preference by similarity to ideal solution (TOPSIS) method. Liang and Meng [36] developed enhanced tools for highway route selection, where TOPSIS was improved with fuzzy logic. Wang and Yeo [23] explored the selection process for intermodal freight routes from Korea to Central Asia. By applying FDELPHI and fuzzy Elimination Et Choix Traduisant la Réalité I (ELECTRE I) methods, they developed a methodology for evaluating and ranking various routes. The FDELPHI method was used to gather expert opinions and assess relevant criteria, while the fuzzy ELECTRE I method enabled the integration of these evaluations to reach a final decision on the most favorable route. Pham and Yeo [21] used a combination of the DELPHI method and consistent fuzzy preference relations (CFPR) to analyze route selection factors for the transport of electronic components from Shenzhen to Vietnam. Tadić et al. [3] applied a combination of FDELPHI, FFARE, and the fuzzy measurement of alternatives and ranking according to compromise solution (MARCOS) method to select the most favorable scenario for implementing IT systems in the Southeast European region. Tadić et al. [11] also employed a combination of the fuzzy step-wise weight assessment ratio analysis (SWARA) and fuzzy MARCOS methods to evaluate IT in Danube region countries. Despite the numerous papers that apply MCDM in transport planning, the contemporary scientific literature lacks documented examples that integrate FDELPHI-FFARE and FADAM methods. This paper aims to address the identified research gaps through the development of a hybrid fuzzy MCDM model, which enables the identification and validation of key criteria and the ranking of potential IT routes.

The traditional DELPHI method, developed by Dalkey and Helmer [37], is used to obtain a consistent flow of responses through an iterative process. The DELPHI method is essentially an interactive approach based on surveys and feedback. Experts are surveyed on a particular issue or problem, and their responses are used to create a new survey. These iterative rounds continue until consensus is achieved or a saturation point in the responses is reached. The FDELPHI method extends the traditional DELPHI approach by incorporating fuzzy logic, which allows for better handling of uncertainties and linguistic variables in expert assessments. This method further enhances the capacity to process subjective and imprecise information by enabling experts to express their views in the form of linguistic variables, which are then translated into fuzzy numbers. The FDELPHI method has been applied in numerous fields. Govindan et al. [38] used the FDELPHI method to identify and analyze key barriers in reverse logistics. By combining fuzzy sets and the DELPHI approach, it became possible to collect and aggregate expert opinions and reach a consensus on the most significant obstacles in reverse logistics. Kuo et al. [39] applied an integrated FDELPHI and TOPSIS method for selecting the locations of logistics centers. FDELPHI was used to filter and prioritize criteria based on expert opinions, enabling a more objective decision-making process for identifying optimal locations. Amri et al. [40] presented a study focused on identifying relevant performance indicators in the edible oil logistics sector in Indonesia. The FDELPHI method was used to select 20 key indicators from an initial set of 34, based on expert consensus from the logistics sector. Ha and Tran [41] used the FDELPHI method to identify key attributes for implementing the physical internet in sustainable supply chains. Experts used this method to define and reduce the set of criteria, after which causal relationships among them were analyzed to better understand digital interoperability and sustainability. Awasthi and Mukhtar [42] presented a three-step approach for evaluating the quality of logistics services, where the FDELPHI method is used to identify criteria from the perspective of multiple stakeholders. The criteria are then weighted using the fuzzy AHP method, and the variants are ranked using the fuzzy TOPSIS method.

The FFARE method applies an extension of the FARE method in a fuzzy environment to reduce the need for a large number of evaluations, eliminate contradictions in comparison matrices, and improve the stability of results. The FFARE method is applied to identify and assess relationships among criteria in the decision-making process. Unlike other pairwise comparison-based methods such as AHP or analytic network process (ANP), FFARE reduces the number of necessary evaluations, thus decreasing complexity and increasing the reliability of the results. The FFARE method is simple to use, provides faster results, and is highly reliable. In addition, it is useful for solving large-scale problems because it does not require extensive comparisons and evaluations of criteria and variants. The FARE approach is based on defining the relationships between decision-making elements, most commonly criteria [43]. In its initial phase, the method requires a data set from decision-makers to identify the influences among individual decision-making elements, including their type and intensity. The influence between the remaining elements is analytically determined in the next phase, significantly reducing the number of assessments required from decision-makers. Krylovas et al. [44], Chatterjee et al. [45], and Kazan et al. [46] state that the main advantages of this method, compared to other methods based on pairwise comparison of decision-making elements (such as AHP, ANP, and decision-making trial and evaluation laboratory (DEMATEL)), include the small number of required evaluations, the elimination of contradictions that arise in comparison matrices, and the high reliability, consistency, and stability of the obtained results. The combination of FDELPHI and FFARE methods provides a synergistic effect in the decision-making process, as FDELPHI enables the consolidation of expert opinions under uncertain and subjective conditions, while FFARE uses those assessments to evaluate and determine the weights of the criteria. This combination enables more accurate and efficient evaluation of criteria, making the model more robust and reliable in complex decision-making situations. By using this combination, it is possible to identify the key factors influencing the selection of IT routes and the optimization of transport systems, while reducing uncertainty in evaluations. FDELPHI and FFARE have wide applications across various disciplines, including route selection, service evaluation, and maintenance strategy assessment, confirming their universality and effectiveness in decision-making.

The ADAM method represents one of the newer MCDM methods—a geometric method. This method was developed to provide a simple yet robust tool for decision-making in complex decision systems. In this approach, the ranking of variants relies on the evaluation of aggregated measurements of complex polyhedra defined by vertices established based on the values of the criteria weights and variants. The results obtained using this method indicate a high level of consistency with the results of other MCDM methods. Krstić et al. [47] applied the existing ADAM method in a fuzzy environment and created a new FADAM method.

Andrejić et al. [48] developed a decision-support system combining the full consistency method (FUCOM) and ADAM methods for selecting optimal distribution channels in logistics. The study evaluated six distribution channels based on nine criteria, revealing that third-party logistics (3PL) services are the most efficient option. This combination of methods offers practical tools for decision-making in real-world logistics scenarios. Popović et al. [49] evaluated ecosystems in selected European countries using the ADAM method with weights determined by the preference selection index (PSI) method. Zhang et al. [50] used the combination of the DEA-MEREC-ADAM methodology to measure the efficiency of logistics processes in the presence of a mismatch between sales and logistics.

Kovač et al. [8] applied mathematical programming and the FADAM method and introduced a new concept for urban logistics based on the development of various categories of logistics centers. Kalem et al. [51] addressed the performance evaluation of railway infrastructure managers by developing and applying a novel hybrid fuzzy MCDM model that integrates the FDELPHI method, an extended fuzzy AHP, and the ADAM method, aiming to identify, assess, and rank key performance indicators (KPIs) to support decision-making, improve efficiency, and enhance strategic planning in the railway sector.

Recent contributions have introduced advanced fuzzy MCDM techniques tailored to transportation and logistics decision-making. For instance, Nayeb-Pashaei et al. [52] utilized a fuzzy-integrated framework to identify critical sustainability criteria in urban transport systems. While their approach excels in holistic sustainability assessment, it is focused primarily on criteria identification, without incorporating a full decision model for ranking transport alternatives.

Kannan et al. [53] proposed the linear diophantine fuzzy CODAS (combinative distance-based assessment) method for logistic specialist selection. Although their method provides high discriminative power among alternatives, it is mathematically complex and requires solving Diophantine equations, which may reduce practical applicability in scenarios requiring broader expert participation or transparency.

Li et al. [54] developed an enhanced spherical cubic fuzzy WASPAS method for evaluating service quality in crowdsourced logistics. This method effectively handles high levels of uncertainty, but it relies heavily on ideal solution concepts and additive weighting, which can distort results in the presence of strongly conflicting or non-compensatory criteria.

In contrast to existing approaches, the proposed FDELPHI–FFARE–FADAM hybrid model provides several practical and methodological advantages. By combining expert consensus through FDELPHI, relational weighting via FFARE, and a non-compensatory ranking process using FADAM, the model follows a clear, sequential structure that enhances interpretability. Unlike methods that rely on ideal or anti-ideal alternatives, this framework reduces the likelihood of biased rankings caused by the influence of extreme values. Additionally, the overall simplicity, transparency, and flexibility of the model make it particularly well-suited for real-world applications in IT planning, where decisions often involve multiple stakeholders and uncertain data. Such a hybrid structure ensures both methodological robustness and practical usability, setting it apart from more algebraically complex or fully integrated MCDM models.

3. Methodology

In this paper, the evaluation of IT routes was conducted using the hybrid FDELPHI–FFARE-FADAM model that involves the sequential and modular use of three distinct fuzzy MCDM methods, each serving a specific role in the decision-making process. The methods are not structurally or mathematically integrated, but rather combined in a complementary fashion to support different phases of evaluation. The proposed hybrid model was structured to leverage the complementary strengths of each method while mitigating the limitations found in widely used MCDM approaches.

The FDELPHI–FFARE method was used to determine the weights of the criteria, while the FADAM method was applied to rank the variants. FFARE was selected for determining criteria weights due to its unique ability to model interrelationships among decision elements with minimal expert input. In contrast to AHP and ANP, which require large numbers of pairwise comparisons and often result in inconsistency within judgment matrices [45], FARE requires only a single row of evaluations to analytically derive all other relationships, ensuring full consistency without the need for iterative revisions [45,46]. Compared to DEMATEL, which uses complex influence matrices and may struggle with stability in large systems [55], FARE is simpler and less data-intensive. Moreover, unlike BWM and FUCOM, which respectively rely on subjective best–worst identification or strict criterion ranking assumptions [56,57], FARE enables more delicate evaluations without imposing rigid decision structures. Its fuzzy extension allows experts to express imprecise judgments through linguistic terms, effectively capturing uncertainty in multi-actor transport evaluations.

Given that FFARE relies on expert input, the fuzzy DELPHI method was introduced to achieve consensus and unify divergent opinions across stakeholder groups. Unlike the classical DELPHI method, which is time-consuming and prone to low response rates, the fuzzy variant accelerates convergence and better handles ambiguity in expert assessments. This makes it particularly suited for decision-making environments where precise data are lacking and group-based judgments dominate.

For the final ranking of alternatives, FADAM was chosen due to its simplicity, transparency, and stability. It offers an intuitive geometric interpretation based on polyhedral volumes and avoids the compensatory effects often observed in methods like TOPSIS, VIKOR (visekriterijumska optimizacija I kompromisno resenje), or SAW (simple additive weighting). Unlike ELECTRE and PROMETHEE (preference ranking organization method for enrichment evaluation), which involve complex outranking procedures and thresholds [58], ADAM is easier to implement and interpret. Its results are highly consistent with those of other MCDM methods and show a low risk of rank reversal, especially in problems involving many criteria. This makes FADAM particularly robust and suitable for strategic transport planning, where both clarity and methodological soundness are essential [47]. Figure 1 presents the outline of the proposed model, and the following steps define the procedure for developing the model using the FDELPHI–FFARE and FADAM methods:

Step 1: Define the problem structure, the criteria and variants to be evaluated, as well as the stakeholder groups.

Step 2: Define the fuzzy scale to be used for the evaluation of criteria and variants. The nine-point fuzzy scale was selected based on established literature to ensure sufficient expressive resolution in expert assessments. It is a fuzzy version of the well-established Saaty scale [47]. The linguistic assessments that can be transformed into triangular fuzzy numbers (TFNs) are shown in

Table 1. Fuzzy evaluation scale.

Descriptive Value	Abbreviation	Fuzzy Scale
“None”	“N”	(1,1,2)
“Very low”	“VL”	(1,2,3)
“Low”	“L”	(2,3,4)
“Fairly low”	“FL”	(3,4,5)
“Medium”	“M”	(4,5,6)
“Fairly high”	“FH”	(5,6,7)
“High”	“H”	(6,7,8)
“Very high”	“VH”	(7,8,9)
“Extremely high”	“EH”	(8,9,10)

.

Step 3: Calculate the criteria weights using the FDELPHI–FFARE method. The detailed steps for determining the criteria weights are as follows:

Step 3.1: Form the criteria evaluation matrix $\widetilde{A_h}$. This matrix is formed by transforming the linguistic assessments of decision-makers, representing different stakeholder groups, into triangular fuzzy numbers using the scale from Table 1:

$${\tilde{A}}_h={\left[{\tilde{a}}_{ijh}\right]}_{{nxn}^{\prime }}\hspace{0.33em}\hspace{0.33em}\hspace{0.33em}\hspace{0.33em}\hspace{0.33em}\hspace{0.33em}\hspace{0.33em}\hspace{0.33em}\forall h\hspace{0.33em}=\hspace{0.33em}1,\dots ,p,$$

(1)

where ${\tilde{a}}_{ijh}=(l_h,m_h,u_h)$ represents the importance of the criterion $i$ relative to the criterion $j$ by decision-maker $h$.

$l_h,m_h$ and $u_h$ denote the lower, middle, and upper values of the triangular fuzzy evaluation ${\tilde{a}}_{ijh}$, $n$ which denotes the number of defined criteria, while $p$ denotes the number of decision-makers conducting the assessment. The following conditions must be observed when forming the matrix $\widetilde{A_h}$:

$${\tilde{a}}_{jih}=-{\tilde{a}}_{ijh},$$

(2)

and the evaluation will be considered consistent if the following condition is met:

$${\displaystyle \sum _{j=1}^n{u}_h=-{\displaystyle \sum _{j=1}^n{l}_h,\hspace{0.33em}\hspace{0.33em}\hspace{0.33em}\hspace{0.33em}\hspace{0.33em}\hspace{0.33em}\hspace{0.33em}\forall h=1,\dots ,p}}.$$

(3)

Step 3.2: Use the FDELPHI method to form the consolidated evaluation matrix of the criteria $\tilde{\Delta }$:

$$\tilde{\Delta }={\left[{\tilde{\delta }}_{ij}\right]}_{{nxn}^{\prime }}$$

(4)

$${\tilde{\delta }}_{ij}=\left(\alpha ,\beta ,\gamma \right)=\left(min(l_h),{\left({\prod }_{h=1}^p{m}_h\right)}^{\raisebox{1ex}{$1$}\!\left/ \!\raisebox{-1ex}{$p$}\right.},max(u_h)\right), h=1,\dots ,p$$

(5)

$\alpha ,\beta$ and $\gamma$ represent the lowest, middle, and highest values of the consolidated fuzzy evaluation ${\tilde{\delta }}_{ij}$, respectively, with $\alpha \le \beta \le \gamma$.

Step 3.3: Calculate the potential impact of the criteria as follows:

$$\tilde{P}=\tilde{H}\left(n-1\right),$$

(6)

where $\tilde{P}$ implies the potential impact of all criteria defined in the MCDM model, while $\tilde{H}$ implies the highest value used in the evaluation scale.

Step 3.4: Calculate the total importance of the criterion ${\tilde{P}}_j$ using the following equations:

$${\tilde{P}}_j={\displaystyle \sum _{i=1}^n{\tilde{\delta }}_{ij},\hspace{0.33em}\hspace{0.33em}\hspace{0.33em}\hspace{0.33em}\hspace{0.33em}\forall j=1,\dots ,n,\hspace{0.33em}\hspace{0.33em}\hspace{0.33em}j\ne i}.$$

(7)

Step 3.5: Calculate the weight of the criterion ${\tilde{w}}_j$ using the following equations:

$${\tilde{w}}_j={\displaystyle \frac{{\tilde{P}}_j^r}{{\tilde{P}}_H}},\hspace{0.33em}\hspace{0.33em}\hspace{0.33em}\hspace{0.33em}\hspace{0.33em}\hspace{0.33em}\forall j=1,\dots ,n,$$

(8)

${\tilde{P}}_H$ represents the potential impact of the considered criteria, obtained using the following formula:

$${\tilde{P}}_H=n\times \tilde{P}$$

(9)

where ${\tilde{P}}_j^r$ represents the actual overall impact of the criterion $j$, obtained by applying the following equation:

$${\tilde{P}}_r^j={\tilde{P}}_j+\tilde{P},\hspace{0.33em}\hspace{0.33em}\hspace{0.33em}\hspace{0.33em}\hspace{0.33em}\hspace{0.33em}\forall j=1,\dots ,n.$$

(10)

Step 4: Rank the variants using the FADAM method [47].

Step 4.1: Define the fuzzy decision-making matrix:

$$\tilde{E}={\left[{\tilde{e}}_{vj}\right]}_{{mxn}^{\prime }}$$

(11)

where ${\tilde{e}}_{vj}=(l^e,m^e,u^e)$ represents the evaluation of variants $v$ regarding the criteria $j$, while $m$ is the number of variants.

Step 4.2: Define the matrix.

$$\tilde{F}={\left[{\tilde{f}}_{vj}\right]}_{{mxn}^{\prime }}$$

(12)

where ${\tilde{f}}_{vj}=(l^f,m^f,u^f)$ indicates the normalized ratings ${\tilde{e}}_{vj}$ calculated as follows:

$$\left(l^f,m^f,u^f\right)=\left(l^e/max{u}^e,m^e/max{u}^e,u^e/max{u}^e\right)$$

(13)

Step 4.3: Define the matrix.

$$\tilde{S}={\left[{\tilde{s}}_{vj}\right]}_{{mxn}^{\prime }}$$

(14)

where ${\tilde{s}}_{vj}=(l^s,m^s,u^s)$ indicates the sorted evaluation ratings (in descending order).

Step 4.4: Find the fuzzy coordinates $({\tilde{x}}_{vj},{\tilde{y}}_{vj},{\tilde{z}}_{vj})$ of the fuzzy reference point ${\tilde{R}}_{vj}$ and the fuzzy weighted reference point ${\tilde{P}}_{vj}$:

$${\tilde{x}}_{vj}=(l^{x_{vj}},m^{x_{vj}},u^{x_{vj}})=(l^{s_{vj}}\times \sin {\alpha }_j,m^{s_{vj}}\times \sin {\alpha }_j,u^{s_{vj}}\times \sin {\alpha }_j),\forall j=1,\dots ,n,\forall v=1,\dots ,m,$$

(15)

$${\tilde{y}}_{vj}=(l^{y_{vj}},m^{y_{vj}},u^{y_{vj}})=(l^{s_{vj}}\times \cos {\alpha }_j,m^{s_{vj}}\times \cos {\alpha }_j,u^{s_{vj}}\times \cos {\alpha }_j),\forall j=1,\dots ,n;\forall v=1,\dots ,m,$$

(16)

$${\tilde{z}}_{vj}=(l^{z_{vj}},m^{z_{vj}},u^{z_{vj}})=\left\{\begin{array}{c}(0,0,0), for {\tilde{R}}_{vj}\\ (l^{wj},m^{wj},u^{zj}), for {\tilde{P}}_{vj}\end{array},\forall j=1,\dots ,n;\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\forall v=1,\dots ,m\right.$$

(17)

where ${\alpha }_j$ is obtained using the following equation:

$${\alpha }_j=(j-1){\displaystyle \frac{{90}^{\circ }}{n-1}},\forall j=1,\dots ,n.$$

(18)

Step 4.5: Determine the fuzzy volume value of the complex polyhedra as follows:

$${\tilde{V}}_v^C={\oplus }_{k=1}^{n-1}{\tilde{V}}_k,\forall v=1,\dots ,m,$$

(19)

where ${\tilde{V}}_k$ are the fuzzy volumes of pyramids defined by each pair of successive criteria are calculated as follows:

$${\tilde{V}}_k={\displaystyle \frac{1}{3}}{\tilde{B}}_k\otimes {\tilde{h}}_k,\forall k=1,\dots ,n-1,$$

(20)

where ${\tilde{B}}_k$ is the fuzzy value of the base area of each pyramid, obtained as follows:

$${\tilde{B}}_k={\tilde{c}}_k\otimes {\tilde{a}}_k\oplus {\displaystyle \frac{{\tilde{a}}_k\otimes ({\tilde{b}}_k\mathsf{\Theta }{\tilde{c}}_k)}{2}},$$

(21)

where ${\tilde{a}}_k$ represents the fuzzy values of the Euclidean distances:

$${\tilde{a}}_k=\sqrt{{\left({\tilde{x}}_{j+1}-{\tilde{x}}_j\right)}^2+{\left({\tilde{y}}_{j+1}-{\tilde{y}}_j\right)}^2}$$

(22)

with expressions ${\tilde{b}}_k=(l^{b_k},m^{b_k},u^{b_k})$ and ${\tilde{c}}_k=(l^{c_k},m^{c_k},u^{c_k})$ equal to

$${\tilde{b}}_k={\tilde{z}}_j,$$

(23)

$${\tilde{c}}_k={\tilde{z}}_{j+1}.$$

(24)

And the fuzzy value of the pyramid height is as follows:

$${\tilde{h}}_k={\displaystyle \frac{\sqrt[2]{{\tilde{s}}_k({\tilde{s}}_k-{\tilde{a}}_k)({\tilde{s}}_k-{\tilde{d}}_k)({\tilde{s}}_k-{\tilde{e}}_k)}}{{\tilde{a}}_k}}$$

(25)

where ${\tilde{s}}_k$ is the fuzzy value of the semiperimeter of the triangles defined by the two successive criteria reference points and the coordinate origin.

$${\tilde{s}}_k={\displaystyle \frac{{\tilde{a}}_k\oplus {\tilde{d}}_k\oplus {\tilde{e}}_k}{2}}$$

(26)

where ${\tilde{d}}_k$ is calculated as follows:

$${\tilde{d}}_k=\sqrt{{{\tilde{x}}_j}^2+{{\tilde{y}}_j}^2}$$

(27)

while ${\tilde{e}}_k$ is calculated as follows:

$${\tilde{e}}_k=\sqrt{{{\tilde{x}}_{j+1}}^2+{{\tilde{y}}_{j+1}}^2}$$

(28)

Step 5: Rank the variants (adopted from Rahmani et al. [59])

$$Crisp({\tilde{V}}_v^C)=(4\times m^{V_v^C}+u^{V_v^C}-2l^{V_v^C})/3(u^{V_v^C}-2l^{V_v^C})$$

(29)

4. Case Study

As part of the analysis of IT routes for B&H, seven variants (V) of IT routes were identified that connect B&H with key European ports and logistics hubs. The analyzed routes include ports on the North Sea, Baltic Sea, Black Sea, and Adriatic Sea, with various transit corridors passing through Central and Southeastern Europe. The variants differ in terms of infrastructure development level, IT terminal efficiency, quality of logistics services, and degree of regulatory harmonization. Some variants involve a combination of road and rail transport, while others include the possibility of utilizing inland waterways, which further influences the flexibility and sustainability of supply chains. The evaluation of each route encompasses trade volumes, infrastructure readiness, environmental aspects, and the potential for achieving economies of scale. The analysis aims to identify the most significant IT routes that can contribute to enhancing the competitiveness of B&H’s economy through modernization, diversification, and IT development. Below are detailed characteristics of each of the seven defined variants, focusing on their advantages, challenges, and specific features.

4.1. Port of Hamburg–Stuttgart–Munich–Salzburg–Ljubljana–Zagreb–Sarajevo–Variant 1 (V1)

Variant 1 (Figure 2) connects Hamburg, a key transport hub on the North Sea, with B&H through major industrial and logistics centers in Central Europe, thereby opening substantial opportunities for B&H to integrate into European and international transport flows. Given its strong connectivity to industrial hubs such as Stuttgart and Munich, this route demonstrates high potential for achieving economies of scale. The capacity and infrastructure along most of the route are at a solid level, although there are bottlenecks in Croatia and B&H. Infrastructure maintenance costs are relatively low, which contributes to the overall efficiency of the route. However, due to high energy consumption and significant environmental impact, this route faces challenges in terms of sustainability and compliance with modern environmental and energy standards.

4.2. Port of Gdansk–Lodz–Brno–Bratislava–Gyor–Pécs–Slavonski Brod–Sarajevo–Variant 2 (V2)

Variant 2 (Figure 3) connects the Baltic Sea with B&H through developed logistics and industrial centers in Central Europe, thereby opening significant opportunities for B&H’s inclusion in European and international transport flows. Thanks to the growing importance of the Port of Gdansk and the developed logistics hubs in Poland and Slovakia, this route demonstrates high potential for developing new markets and achieving economies of scale through cargo consolidation in hubs such as Lodz and Bratislava. Capacity and infrastructure are at a satisfactory level in most of the countries along the route, although the segments through Croatia and particularly B&H require modernization. High regulatory harmonization and efficient customs procedures along most of the route further enhance its competitiveness. However, the volume of goods currently transported via this route is limited, while low energy efficiency and environmental performance present challenges for its sustainability in the context of modern demands for green transport.

4.3. Port of Constanța–Bucharest–Craiova–Sremska Mitrovica–Sarajevo–Variant 3 (V3)

Variant 3 (Figure 4) connects the Black Sea with B&H, providing an important route for the integration of trade flows between Southeastern Europe and the broader Asian market. This route includes road, rail, and inland waterway transport, with significant potential for developing new markets, especially through connections with Black Sea and Asian ports. Environmental indicators and energy efficiency are very high, giving this variant an advantage in terms of sustainable transport. However, infrastructure quality is uneven, and significant investments are required, particularly for the modernization of railway lines and terminals in Serbia and B&H. Administrative barriers are prominent due to varying legal regulations within and outside the EU, coupled with low efficiency of customs procedures at border crossings. The volume of goods currently transported along this route is relatively small, further emphasizing the need for strategic development to capitalize on the geographical advantages of this corridor.

4.4. Vienna–Bratislava–Budapest–Novi Sad–Belgrade–Šabac–Sremska Mitrovica–Sarajevo–Variant 4 (V4)

Variant 4 (Figure 5) represents a strong transport link between Central Europe and B&H, relying on well-developed infrastructure that includes road, rail, and inland waterway transport. Thanks to its connection with major IT terminals in Vienna, Budapest, and Belgrade, this route has a high capacity for achieving economies of scale and significantly contributes to reducing energy consumption, making it one of the more energy-efficient routes. Its environmental impact is very favorable, though further investment is needed in modernizing certain infrastructure segments. The volume of goods exchanged along this route is substantial. However, double customs procedures (EU–Serbia and Serbia–B&H) pose an administrative challenge, and the legal framework requires further harmonization. The availability of all three modes of transport and access to major logistics hubs make this variant the most technically advanced and sustainable in the long term.

4.5. Port of Koper–Ljubljana–Zagreb–Sarajevo–Variant 5 (V5)

Variant 5 (Figure 6) represents one of the shorter and more efficient links between the Adriatic Sea and B&H, with the Port of Koper playing a strategic role as the main maritime gateway for Central and Southeastern Europe. The infrastructure and capacity along the route are at a medium level. This variant achieves a solid volume of goods exchange thanks to well-developed business relations between B&H, Slovenia, and Croatia, while the relative proximity contributes to lower energy consumption and a more favorable environmental impact. A high rating of regulatory compliance further facilitates cross-border flows. Although opportunities for expansion into new markets and the realization of economies of scale are not particularly pronounced, the stable institutional framework and geographic convenience make this route a suitable option for the development of an IT route.

4.6. Port of Rijeka–Zagreb–Sarajevo–Variant 6 (V6)

The Port of Rijeka is an important maritime port for B&H, offering direct access to the Adriatic Sea and relatively short distances to final destinations within B&H (Figure 7). The route is characterized by a high volume of imports and exports, which highlights its strong role in trade. The level of infrastructure is satisfactory, although there are capacity limitations that may affect long-term efficiency. Owing to the presence of only one border crossing between Croatia and B&H, this route has an advantage in terms of regulatory frameworks and the efficiency of customs procedures. Required investments in modernization are relatively low, and the shorter distance contributes to reduced energy consumption and a favorable environmental impact. The potential for developing new markets and achieving economies of scale remains moderate, indicating that the route is efficient but more focused on existing flows rather than expansion.

4.7. Port of Trieste–Ljubljana–Zagreb–Sarajevo–Variant 7 (V7)

The route connecting the Port of Trieste with B&H via Slovenia and Croatia holds strategic importance due to its linkage with three developed markets, Italy, Slovenia, and Croatia (Figure 8). This variant enables stable goods exchange with a moderate traffic volume. Its connection with developed markets contributes to greater potential for trade expansion and access to new markets. The route requires somewhat higher infrastructure investments, while maintenance costs are moderate. Despite the presence of multiple border crossings, the regulatory framework and customs procedures are assessed as relatively efficient. Energy consumption and environmental performance are solid, though not optimal, highlighting the need to increase the share of rail transport to improve the sustainability of this route.

4.8. Defined Criteria for Evaluating IT Routes

This section describes the criteria used to evaluate IT routes. A total of 10 criteria have been defined, grouped into three categories: economic (C1, C3, C4, C5, and C6); technical-legal (C2, C7, and C8); and environmental (C9 and C10). These criteria play a key role in assessing the performance and potential development of IT routes that connect B&H with its main trade partners.

Volume of imported and exported goods (C1) refers to the scale of trade exchange between B&H and the countries along the route. A higher volume of trade indicates greater importance and, thus, higher development priority for the route.

Capacity and infrastructure readiness (C2) represents the measure of the infrastructure’s ability and readiness to support efficient goods transport and trade. This criterion assesses the physical and technical capacities of transport infrastructure (roads, railways, ports) and their suitability for receiving and distributing goods in accordance with market needs.

Potential for new market development (C3) represents the extent to which trade expansion can occur through the opening of new markets for goods and services. This criterion evaluates how the development of a route can help B&H companies find and grow new markets.

Required investments in infrastructure development and modernization (C4) measure the financial investments needed for infrastructure development and expansion.

Infrastructure maintenance costs (C5) refer to the costs necessary to maintain the functionality and competitiveness of the existing network infrastructure.

The possibility of achieving economies of scale through cargo consolidation and increased transport volume (C6) measures efficiency and cost-effectiveness. This criterion evaluates how increasing transport volumes and consolidating shipments can reduce the average cost per unit of cargo.

Legal regulation (C7) measures the alignment of legal frameworks and prescribed laws governing trade, customs policy, product safety, intellectual property protection, taxation, and other economic aspects related to international goods exchange.

Efficiency and transparency of customs procedures at border crossings (C8) measures the degree of effectiveness and speed in customs control and the cross-border movement of goods.

Energy consumption (C9) measures the total energy consumption along the defined route, depending on the modes of transport used and their combinations.

Environmental impact (C10) refers to the emission of harmful substances, noise, and vibrations, as well as safety improvements.

Using the proposed MCDM methodology, the criteria were evaluated using linguistic evaluations by three stakeholder groups: operators, service users, and authorities (

Table 2. Linguistic evaluation of criteria by interest groups.

Cj	C1	C2	C3	C4	C5	C6	C7	C8	C9	C10
C1	/, /, /	/, /, H	/, /, FH	M, /, FH	H, /, FL	/, /, VL	/, /, L	/, /, L	/, /, M	M, /, FH
C2	M, VL, /	/, /, /	FH, H, H	VL, L, FH	L, L, M	/, FH, VL	H, H, M	M, VH, L	FH, FH, M	VL, M, FH
C3	FH, L, /	/, /, /	/, /, /	VL, M, H	L, M, FH	/, H, VL	/, /, M	FH, /, FL	H, M, FH	VL, FH, H
C4	/, H, /	/, /, /	/, /, /	/, /, /	/, VH, FH	/, /, L	/, /, M	/, /, FL	/, /, H	VH, /, /
C5	/, H, /	/, /, /	/, /, /	H, /, /	/, /, /	/, /, FL	/, /, H	/, /, M	/, /, /	H, /, /
C6	FL, FL, /	H, /, /	FH, /, /	VL, M, /	L, M, /	/, /, /	H, /, /	M, /, /	FH, FH, /	VL, H, /
C7	M, L, /	/, /, /	H, H, /	VL, FL, /	L, FL, /	/, FH, M	/, /, /	M, /, H	FH, M, /	VL, FH, /
C8	FH, VL, /	/, /, /	/, H, /	FL, FL, /	M, FL, /	/, FH, H	/, H, /	/, /, /	/, M, /	FL, M, /
C9	FH, M /	/, /, /	/, /, /	L, H, /	FL, H, H	/, /, L	/, /, FL	H, /, M	/, /, /	L, /, /
C10	/, M, /	/, /, /	/, /, /	/, FH, VH	/, FH, H	/, /, L	/, /, M	/, /, FL	/, H, H	/, /, /

). Stakeholder groups were selected using purposeful sampling, focusing on those directly engaged in intermodal transport. To ensure representativeness and avoid bias, we ensured diversity in roles and regions, conducted anonymous and independent evaluations, and used fuzzy aggregation to absorb variability.

The evaluation was conducted with 12 domain experts belonging to one of the stakeholders’ groups. Although modest in size, this expert panel aligns with established practices in MCDM research. Similar sample sizes are commonly reported in recent expert-based decision-making studies in transport and energy sectors [60,61].

The FDELPHI method was implemented in three iterative rounds. Consensus was defined as at least 75% agreement among experts, measured by overlapping fuzzy evaluations. Criteria not meeting this threshold were re-evaluated in subsequent rounds. By the final round, consensus was achieved for all criteria.

The obtained evaluations were then transformed into fuzzy values using the relations presented in Table 1, forming evaluation matrices (1), in compliance with conditions (2) and (3). For these matrices, a consolidated evaluation matrix was formed using Equations (4) and (5). Using Equation (6), the potential impact of each criterion was calculated, while the overall importance of each criterion was determined using Equation (7). Final criterion weights were derived using Equations (8)–(11). The resulting weights were C1 (0.082), C2 (0.200), C3 (0.144), C4 (0.071), C5 (0.057), C6 (0.099), C7 (0.096), C8 (0.088), C9 (0.087), C10 (0.076).

The variants were also evaluated by the 12 experts. The fuzzy approach effectively manages uncertainty and imprecision by enabling experts to express judgments using linguistic terms converted into fuzzy numbers. In this study, experts supplemented their experience and domain knowledge with available quantitative data from official statistics. However, as no complete data sets were adequate for direct application to the defined problem, these sources were used to inform and support expert-driven, data-guided evaluations. This integration allowed the fuzzy model to reflect both empirical insights and expert interpretation, ensuring realistic and robust decision-making under limited precise data conditions. The experts’ evaluations were transformed into triangular fuzzy numbers using Table 1, forming the fuzzy decision matrix (14), (

Table 3. Evaluation of variants according to criteria.

	V1	V2	V3	V4	V5	V6	V7
C1	VH	L	FL	H	M	EH	FH
C2	VH	H	FL	EH	M	FL	M
C3	FH	EH	FH	L	M	FL	M
C4	H	FL	VL	VH	FH	L	FL
C5	VH	FL	L	H	FH	FH	M
C6	EH	H	M	EH	M	FL	FL
C7	VH	VH	FL	M	VH	VH	H
C8	VH	VH	VL	M	M	M	FL
C9	FL	L	VH	EH	FH	H	FH
C10	M	FL	VH	VH	FH	H	FH

).

By applying Equations (11)–(13), a normalized fuzzy decision matrix was formed. Then, using Equation (14), a sorted fuzzy decision matrix was created, from which the coordinates of fuzzy reference and fuzzy weighted reference points were calculated using Equations (15)–(18). Afterward, fuzzy values representing the fuzzy volumes of complex polyhedra were calculated using Equations (19)–(28). In the next step, variant ranking was performed using Equation (29). Variant V1, that is, the IT route from the Port of Hamburg, was identified as the best solution (Sc.0) (

Table 4. Variant ranking.

	V1	V2	V3	V4	V5	V6	V7
$\mathrm{C}\mathrm{r}\mathrm{i}\mathrm{s}\mathrm{p} ({\tilde{\mathit{V}}}_{\mathit{v}}^{\mathit{C}})$	0.0379041	0.03029671	0.02033713	0.03406055	0.02350591	0.02543129	0.02068653
Rang	1	3	7	2	5	4	6

).

5. Sensitivity Analysis and Validation

A sensitivity analysis was conducted using twelve additional scenarios, alongside the baseline scenario (Sc.0), to verify the stability of the obtained solution. In Sc.1, all criteria were treated as equally important, thereby testing the impact of a uniform weight distribution. In Scenarios 2 through 6 (Sc.2 to Sc.6), one of the five highest-weighted criteria was excluded, specifically C2, C3, C6, C7, and C8, respectively. In Scenarios 7, 8, and 9, two of the most important criteria were removed in different combinations (Sc.7—C2 and C3; Sc.8—C2 and C7; Sc.9—C3 and C7). The final three scenarios explored more extreme changes in the criteria structure to test the robustness of the developed model. In Sc.10, three of the most influential criteria (C2, C3, and C6) were simultaneously excluded, simulating a scenario with a significant loss of key factors. In Sc.11, an additional highly important criterion (C7) was added to the previously removed three (C2, C3, and C6), making it a total of four excluded high-impact criteria. In the last, most demanding scenario, Sc.12, as many as five top-weighted criteria were excluded (C2, C3, C6, C7, and C8), testing the model’s resilience to major changes in input parameters and radical shifts in preferences. In most scenarios, the criteria excluded were those with the highest weights based on the weighting results. This stepwise exclusion tested the model’s stability against the removal of critical factors, demonstrating robust ranking outcomes even under these extreme assumptions. Creating additional scenarios by excluding remaining lower-weighted criteria would not result in significant differences, as their weights are only marginally different from that of the last excluded criterion (C8).

The ranking results for all scenarios are shown in

Table 5. Sensitivity analysis.

		V1	V2	V3	V4	V5	V6	V7
Sc.0	$\mathrm{Crisp}({\tilde{V}}_v^C)$	0.0379041	0.03029671	0.02033713	0.03406055	0.02350591	0.02543129	0.0206865
	Rank	1	3	7	2	5	4	6
Sc.1	$\mathrm{Crisp}({\tilde{V}}_v^C)$	0.039003	0.02823691	0.0212043	0.03667701	0.02483911	0.02794861	0.0215789
	Rank	1	3	7	2	5	4	6
Sc.2	$\mathrm{Crisp}({\tilde{V}}_v^C)$	0.0488774	0.03914673	0.02622194	0.04403966	0.03020977	0.03248755	0.0266641
	Rank	1	3	7	2	5	4	6
Sc.3	$\mathrm{Crisp}({\tilde{V}}_v^C)$	0.0423105	0.0327353	0.02284778	0.03845477	0.02652585	0.02922316	0.0230245
	Rank	1	3	7	2	5	4	6
Sc.4	$\mathrm{Crisp}({\tilde{V}}_v^C)$	0.0427448	0.03326357	0.02227732	0.03836125	0.02710393	0.02956387	0.0237018
	Rank	1	3	7	2	5	4	6
Sc.5	$\mathrm{Crisp}({\tilde{V}}_v^C)$	0.0416709	0.03337892	0.02219471	0.03679301	0.02585045	0.02826521	0.0227531
	Rank	1	3	7	2	5	4	6
Sc.6	$\mathrm{Crisp}({\tilde{V}}_v^C)$	0.0406814	0.03182263	0.02326518	0.03852483	0.02509651	0.02778457	0.0223227
	Rank	1	3	6	2	5	4	7
Sc.7	$\mathrm{Crisp}({\tilde{V}}_v^C)$	0.0477814	0.03467057	0.02602121	0.04434733	0.03063773	0.03493466	0.026193
	Rank	1	4	7	2	5	3	6
Sc.8	$\mathrm{Crisp}({\tilde{V}}_v^C)$	0.0558396	0.04365349	0.02912636	0.05021655	0.03524607	0.03818339	0.0309203
	Rank	1	3	7	2	5	4	6
Sc.9	$\mathrm{Crisp}({\tilde{V}}_v^C)$	0.043076	0.03057257	0.02246388	0.03976259	0.02835456	0.03253212	0.0246705
	Rank	1	4	7	2	5	3	6
Sc.10	$\mathrm{Crisp}({\tilde{V}}_v^C)$	0.0492539	0.0321673	0.02582189	0.04651964	0.03331346	0.03959723	0.0285517
	Rank	1	5	7	2	4	3	6
Sc.11	$\mathrm{Crisp}({\tilde{V}}_v^C)$	0.0473629	0.02946753	0.02672566	0.04561727	0.031645	0.03987571	0.0277901
	Rank	1	5	7	2	4	3	6
Sc.12	$\mathrm{Crisp}({\tilde{V}}_v^C)$	0.0428989	0.02070884	0.03184847	0.0526532	0.02799942	0.0389028	0.026064
	Rank	2	7	4	1	5	3	6

, while a graphical representation is provided in Figure 9. A high degree of ranking stability was observed—Variant V1 (intermodal route from the Port of Hamburg) retained the first place in eleven out of the twelve scenarios. In Sc.12, after the exclusion of five key criteria, V1 dropped to second place, while V4 (the route from Vienna), which had previously held second position, emerged as the top-ranked option. Variant 3 was ranked lowest in most scenarios, except in Sc.6 and Sc.12. Based on the conducted analysis, it can be concluded that the solution from the baseline scenario demonstrates satisfactory stability and robustness, and can therefore be considered reliable for final decision-making.

To validate the results, the same problem was solved with an additional five highly applied MCDM methods, namely fuzzy TOPSIS (FTOPSIS), fuzzy VIKOR (FVIKOR), fuzzy SWARA (FSWARA), fuzzy CODAS (FCODAS), and fuzzy comprehensive distance-based ranking (FCOBRA). Similarity of the obtained results has been assessed by calculating the Spearman correlation coefficient (SCC). The average value of 0.957 indicates a very high level of conformity of the results with the reference methods. The results are validation results are presented in

Table 6. Results validation.

Method		V1	V2	V3	V4	V5	V6	V7	SCC
FADAM	Result	0.038	0.031	0.020	0.034	0.024	0.026	0.021
	Rank	1	3	7	2	5	4	6	/
FTOPSIS	Result	0.963	0.828	0.061	0.820	0.146	0.137	0.066
	Rank	1	2	7	3	4	5	6	0.92857
FVIKOR	Result	−0.378	−0.312	−0.181	−0.332	−0.238	−0.224	−0.216
	Rank	1	3	7	2	4	5	6	0.96429
FSWARA	Result	0.768	0.658	0.485	0.722	0.566	0.563	0.525
	Rank	1	3	7	2	4	5	6	0.96429
FCODAS	Result	1.428	1.163	−0.123	1.279	0.006	−0.007	−0.078
	Rank	1	3	7	2	4	5	6	0.96429
FCOBRA	Result	0.000	0.139	0.357	0.059	0.254	0.259	0.305
	Rank	1	3	7	2	4	5	6	0.96429

.

The SCC values indicate that there are no significant differences between the final rankings, and the computational times are very similar. However, although the procedural steps of the ADAM method application are somewhat more complex, it provides more reliable results. ADAM avoids reliance on artificial ideal or anti-ideal solutions, unlike TOPSIS, VIKOR, CODAS, or COBRA, which can be distorted by outliers. Instead, it ranks alternatives based on the actual spatial configuration of their performance across criteria, using volume calculations. This makes ADAM more robust to extreme values, less prone to rank reversal, and better suited for problems with subjective or uncertain data, such as, among others, sustainability and transport planning.

6. Discussion

Based on the evaluation of IT routes in the context of connecting B&H with key European ports and logistics hubs, it was determined that Variant V1 (the route via the Port of Hamburg) represents the most favorable solution. This variant stood out as the most efficient in the majority of scenarios, taking into account the defined criteria and the weights assigned through the involvement of representatives of relevant stakeholder groups. In the baseline scenario and most of the alternative sensitivity scenarios, V1 retained the leading position, confirming its robustness under different circumstances. The sensitivity analysis, conducted through twelve scenarios involving changes in criteria weights, showed that significant changes in ranking occurred only when five criteria with the highest weights were excluded from the model. Even in that case, the shift occurred only between the first and second variants, indicating a stable preference structure within the model. The dominant position of Variant V1 is the result of its outstanding performance in the key criteria that carry the greatest weight in the model. Primarily, V1 achieved the highest score in criterion C6—the potential for achieving economies of scale—indicating high efficiency and cost-effectiveness through cargo consolidation and larger transport capacities. In addition, the variant proved highly competitive in criteria C2 and C3, which refer to infrastructure readiness and the potential for developing new markets.

Although environmental criteria (C9 and C10) received relatively lower weights in the evaluation, this outcome reflects the prevailing priorities of stakeholders in the given context, where legal, technical, and economic constraints are perceived as more pressing. Nonetheless, the selected criteria are aligned with the sustainability development goals (SDGs) and collectively reflect a strong awareness of sustainability principles.

The listed criteria align with several sustainable development goals (SDGs). C1 (Trade Volume) and C3 (New Market Development) support SDG 8 (Decent Work and Economic Growth) and SDG 9 (Industry, Innovation and Infrastructure) by enhancing economic integration and access to markets. C2 (Infrastructure Readiness), C4 (Infrastructure Investment), and C5 (Maintenance Costs) directly relate to SDG 9, promoting resilient infrastructure and sustainable industrialization. C6 (Economies of Scale) contributes to SDG 12 (Responsible Consumption and Production) by improving resource efficiency in logistics. C7 (Legal Regulation) and C8 (Customs Efficiency) align with SDG 16 (Peace, Justice and Strong Institutions) through regulatory coherence and institutional effectiveness. C9 (Energy Consumption) links to SDG 7 (Affordable and Clean Energy) and SDG 13 (Climate Action) by addressing energy efficiency in transport. C10 (Environmental Impact) supports SDG 13, SDG 11 (Sustainable Cities and Communities), and SDG 3 (Good Health and Well-being) by reducing pollution and enhancing environmental and public health outcomes. These linkages reflect the multidimensional contribution of intermodal transport to sustainable development. Therefore, even though C9 and C10 were weighted lower, other criteria implicitly support the environmental, social, and economic dimensions of sustainability. The inclusion of diverse indicators from economic, technical-legal, and environmental domains ensures a balanced and holistic evaluation framework.

Several of the evaluated intermodal routes align with key sections of the EU TEN-T network, particularly the Mediterranean and Atlantic Corridors. For example, Variant V1 corresponds closely with the Atlantic Corridor through its connection to the Port of Hamburg, while Variants V4 and V5 intersect with the Mediterranean Corridor via major hubs such as Vienna, Ljubljana, and Zagreb. This alignment reinforces the regional relevance of the selected routes, highlighting their potential to enhance BiH’s integration into the broader European transport system and support EU connectivity, trade facilitation, and cohesion policy objectives.

Regional political dynamics, customs procedures, and the ongoing EU accession process are important contextual factors that influence the stability and feasibility of IT routes in B&H. While these aspects were beyond the scope of this study, future research could incorporate geopolitical risk indicators and cross-border coordination metrics into the evaluation framework.

These results confirm the validity and consistency of the applied MCDM model, which combines the FDELPHI-FFARE and FADAM methods. Compared to previous research, this paper contributes by applying a hybrid fuzzy MCDM model for evaluating IT routes in a regional context. Furthermore, the study supports the establishment of a criteria-based foundation for structured decision-making in the planning and development of transport routes. Identifying routes that provide greater flexibility, better environmental sustainability, and integration into broader European logistics flows forms the basis for defining concrete measures in the process of modernizing and enhancing the competitiveness of the transport system. One limitation of this study relates to the number of defined IT route variants. Although the seven selected variants represent the most relevant and strategically significant routes connecting B&H with major European ports and logistics centers, the inclusion of additional variants could provide a broader perspective. Given the anticipated infrastructure development and changing geopolitical and trade conditions in the region, future research could include a larger number of routes to obtain a more comprehensive picture. A second limitation concerns the number of stakeholder groups involved in the evaluation process. This study includes three key groups—operators, service users, and government representatives—whose perspectives are essential for transport planning. However, involving additional actors such as representatives of non-governmental organizations, academia, or international investors could further enrich the analysis and provide a deeper understanding of the complexity of the issue. The theoretical implications of the study lie in its contribution to the existing literature on IT and MCDM through the development and application of an innovative hybrid model that integrates FDELPHI-FFARE and FADAM methods. By combining these approaches into a single model, the paper enables more accurate evaluation of complex problems under uncertainty and limited data, which is a common occurrence in transport planning. This enhances the methodological framework for research in logistics and opens new possibilities for its application in related areas. Additionally, the paper emphasizes the importance of involving diverse stakeholder groups in the decision-making process, further underlining the need for integrating multiple perspectives in the development of sustainable transport solutions. The practical implications include providing concrete recommendations for improving the IT network in B&H. The identification and ranking of IT routes enable decision-makers to direct available resources toward priority projects, which can contribute to more efficient integration of B&H with European markets. The model can also be adapted for other sectors where there is a need to make decisions based on multiple conflicting criteria under conditions of uncertainty.

7. Conclusions

In the modern context of transport and logistics infrastructure planning, IT is increasingly recognized as a key instrument for achieving a sustainable, efficient, and integrated transport system. The development of IT routes serves as a sustainable alternative that simultaneously reduces negative environmental impacts and increases economic efficiency. Although IT is widely implemented in developed transport systems across Europe and the world, its full integration in transitional and infrastructurally less developed countries such as B&H still presents a challenge. Infrastructure fragmentation, lack of alignment in the regulatory framework, outdated technical and technological capacities, a shortage of terminals, and insufficient coordination among various actors in the transport chain limit the effective application of IT. On the other hand, B&H’s geostrategic position, which connects Central and Southeastern Europe, offers significant potential for the development of an IT network, provided that key routes are properly identified and adequately evaluated. The subject of this research was to define a framework for the evaluation and selection of the most favorable IT routes. The framework and methodology encompassed in this study were tested and validated through the case of selecting the most favorable IT route variant for B&H, considering a complex set of criteria and the inclusion of diverse stakeholder groups. To address this problem, a hybrid MCDM model was applied, integrating the FDELPHI-FFARE and FADAM methods. This approach enabled the structured identification and evaluation of relevant criteria, among which the most significant were capacity and infrastructure readiness, potential for developing new markets, and the possibility of achieving economies of scale. Based on the evaluation of seven defined IT route variants, the most favorable solution was identified as Variant V1—the route via the Port of Hamburg, which achieved the highest overall score in the baseline scenario, as well as in most of the sensitivity analysis scenarios. Conversely, Variant V3 was found to be the least acceptable solution. The sensitivity analysis, conducted through twelve alternative scenarios, further confirmed the robustness and consistency of the results, with significant changes in ranking occurring only after the removal of the five most influential criteria. This confirmed the resilience of the developed model to changes in the decision-making structure.

To support practical implementation, the study’s results can be translated into an actionable decision-support toolkit for policymakers. A priority matrix for infrastructure investment in Bosnia and Herzegovina (B&H) is recommended, categorizing routes by urgency and strategic impact. For example, V1, due to its strong infrastructure readiness and integration with key logistics hubs, should be prioritized for immediate investment to enhance connectivity and competitiveness. Conversely, V3, which offers high environmental performance but currently lacks sufficient infrastructure, is best suited for long-term decarbonization strategies. In parallel, the study highlights the need for enhanced cross-border regulatory coordination, particularly to streamline dual customs procedures affecting routes such as V4. Policymakers should consider establishing bilateral agreements and harmonized customs protocols with neighboring countries to reduce administrative delays and improve the efficiency of intermodal flows. Together, these targeted actions can serve as an integrated toolkit to operationalize sustainable transport planning in B&H.

The main contribution of this paper lies in the development and application of a fuzzy MCDM model for the selection of IT routes in a regional context, as well as in laying the foundation for strategic decision-making in the field of transport policy. The novelty of this paper is twofold. From a methodological perspective, it introduces a new hybrid decision-making framework that integrates FDEPHI, FFARE, and FADAM sequentially, allowing for expert-driven criteria validation, relational weighting, and robust, non-compensatory ranking of alternatives. To the best of our knowledge, this is the first study to apply this specific combination in the domain of transport planning. From an application standpoint, the model is applied to the evaluation of intermodal transport routes in a regional context that has received limited attention in the literature. By incorporating input from key stakeholders and testing multiple real-world scenarios, the model provides a replicable and decision-supportive tool for sustainable transport development in emerging European regions.

Based on the route rankings and sensitivity analysis, the study suggests that infrastructure investments should prioritize corridors with strong regulatory and technical readiness, while simultaneously addressing customs inefficiencies. Policymakers are also encouraged to promote intermodal integration and cross-border coordination, especially in the context of EU transport policy alignment.

Although the model was developed for the case of B&H, its modular structure allows for adaptation to other regional or national settings. By adjusting the evaluation criteria, stakeholder composition, and linguistic scales to reflect local conditions, the FDELPHI–FFARE–FADAM model can be applied in a wide range of IT planning contexts.

Future research directions could include expanding the analysis to additional transport route variants, incorporating a larger number of criteria, and adapting the model for specific regional areas. Additionally, the developed model can serve as a foundation for developing similar MCDM tools applicable to other domains within logistics and transport planning.