Integrating Industry 4.0 and circular economy principles in emergency third-party logistics service provider selection: a new multi-criteria decision-making approach

Humanitarian logistics is a type of logistics whose purpose is to provide effective and timely relief to people in crisis to reduce their suffering by applying an efficient and integrated planning, control and management process of relief goods and information from origin to destination (Heaslip et al., 2018). Because of the lack of use of advanced technologies and ignoring environmental concerns, traditional humanitarian logistics is unable to deal with the uncertainty of crises and environmental challenges caused by them (Ülkü et al., 2024). Recently, with the emergence of circular economy (CE) and Industry 4.0 concepts, a revolutionary transformation has taken place in the logistics field (Khan et al., 2023). CE is a closed-loop system for effectively managing waste to reduce resource consumption. It uses strategies such as recycling, remanufacturing, repair, reuse and refurbishment (Tavana et al., 2024a). On the other hand, Industry 4.0 refers to the use of technologies such as the Internet of Things (IoT), artificial intelligence (AI), big data analytics and so forth (Misra et al., 2020). The literature review demonstrates that the integration of CE practices and Industry 4.0 technologies leads to synergy and improves the economic, environmental and social performance of supply chains (Sorooshian et al., 2024). The humanitarian supply chain is not exempt from this issue and the application of these two paradigms in humanitarian logistics activities can effectively affect its sustainability and resilience (Ülkü et al., 2024).

In humanitarian logistics, several factors contribute to improving relief performance and accelerating service processes. These factors include real-time tracking of relief goods, continuous monitoring of logistics activities, optimization of the flow of relief goods, sharing of data resulting from the assessment of crisis-affected zones, inventory management and integration of operations (Niyazi and Behnamian, 2023; Habibi Rad et al., 2021). Industry 4.0 technologies, such as IoT, global navigation satellite system (GNSS) and platforms for data sharing, provide effective strategies for humanitarian logistics (Govindan et al., 2024a; Govindan et al., 2024b).

In addition, CE practices in humanitarian logistics lead to efficient resource management, waste management and overall increased sustainability (Larson, 2021). A major challenge in relief supply chains is waste management. For example, in crisis-affected zones, a large amount of waste (water bottles, cans and infectious waste such as used syringes and sanitary napkins) is produced daily and, if not properly managed, pollutes the environment and endangers the lives of living organisms and local residents (Corbett et al., 2022; Karl and Scholz Karl, 2022). Hence, in addition to the quality and speed of relief services, decision-makers in the field of disaster management must also emphasize sustainability (Bag et al., 2022). The implementation of CE practices in logistics operations is a suitable solution for moving humanitarian supply chains toward sustainability.

The core question is how can aid agencies or disaster management organizations best provide sustainable, fast and quality services? Public–private partnership is a viable solution to achieve this goal. As the public sector that provides the project budget, the disaster management organization entrusts humanitarian logistics activities in the form of a project to a project implementer, a private sector representative that has the necessary expertise, technical knowledge and equipment. In this vein, a contract is formed between the private and public sectors in the pre-disaster phase, and the private sector is committed to carry out the logistics activities assigned to it in the post-disaster phase. For example, quantity flexibility contract is one of the most well-known and widely used contracts for relief conditions. This contract, on the one hand, reduces public sector costs in the pre-disaster phase by reducing the need for inventory pre-positioning. On the other hand, by assigning logistics activities to the private sector, it decreases the risk of implementing post-disaster activities (Nikkhoo et al., 2018). It is noteworthy that this contract could even incorporate public monetary donations and NGO activities, delegating their management to the private sector (Modarresi and Maleki, 2023).

One effective element in humanitarian logistics is outsourcing; it contributes greatly to the implementation of efficient operations (Falagara Sigala and Wakolbinger, 2019). In this regard, aid agencies allot considerable sums – amounting to numerous US dollars – to logistics services in which providers act as important cogs in the machinery of disaster relief delivery (Gossler et al., 2020). Although they play a pivotal role, the joint partnership of aid agencies and Logistics Service Providers cannot meet expectations (Bealt et al., 2016). Problematic issues that exist in the humanitarian sector of outsourcing has led to dissatisfaction in lots of companies regarding final performance and costs (Nurmala et al., 2017). Different features in place among the aid agencies, including unpredictable funding streams, varying work styles, high staff turnover and cultural differences, may hinder the complex workings of these partnerships and collaborations (Gossler et al., 2020).

Proper decision-making is fundamental to selecting emergency third-party logistics service providers (E3PLSP). It is probable that a number of advantages like cost reduction, improved operational efficiency and higher effectiveness appear when this process is accomplished suitably (Kim et al., 2019). Effective E3PLSP selections can also help overcome critical obstacles, including operational disparities, financial uncertainties and cultural diversity in humanitarian logistics (Kucukaltan et al., 2022).

In summary, the integration of CE and Industry 4.0 concepts in the E3PLSP selection problem marks a vital step forward in the field of sustainable humanitarian logistics. Cooperation with sustainable E3PLSPs that effectively apply Industry 4.0 technologies in logistics operations minimizes adverse environmental effects, optimizes logistics operations, improves resource management efficiently, manages waste and generally makes logistics performance more sustainable (Kannan et al., 2024; Matarneh et al., 2024).

Two factors including “identifying appropriate evaluation criteria” and “developing an efficient evaluation approach” have a significant impact on the selection of an appropriate E3PLSP. The main contributions of this article are based on these two factors. This is the first article that considers circular and Industry 4.0 criteria in evaluating E3PLSPs. Therefore, the first contribution of this research is to identify a comprehensive set of criteria for evaluating sustainable E3PLSPs from the perspective of Industry 4.0 and circularity. Another contribution of this article is the development of a novel MCDM approach for evaluating E3PLSPs. In this vein, we introduce a hybrid MCDM approach by combining fuzzy theory, stepwise weight assessment ratio analysis (SWARA) technique and weighted influence non-linear gauge system (WINGS) method. Fuzzy SWARA method is used to calculate independent weights of evaluation criteria and sub-criteria, and WINGS technique is used to rank the E3PLSPs because it considers the interdependencies among criteria. In general, the objective of this article is to answer the following questions:

The rest of this article is organized as follows. Related literature is included in Section 2. Section 3 deals with proposed approach. The implementation of the proposed approach in the real world is placed in Section 4. Comparative analysis and discussion are presented in Section 5. Managerial implications and research implications are provided in Sections 6 and 7, respectively. Section 8 is devoted to the conclusion.

E3PLSP selection problem is among decision-making problems. MCDM tool is a powerful tool to solve this problem (Kim et al., 2019). Hence, many articles have used MCDM methods to solve this problem. In decision-making problems where MCDM methods are used, two factors are important:

1. identifying the proper criteria; and

2. developing an efficient MCDM approach (Alavi et al., 2021).

Therefore, in this section, with the objective of identifying evaluation criteria and MCDM methods, we review the relevant literature. To achieve these objectives, we review the articles in the field of logistics service provider selection that use MCDM methods. Note that we applied keywords such as “Industry 4.0,” “CE,” “digitalization,” “disaster,” “crisis,” “humanitarian” and “relief,” to make our searches more targeted. Following a comprehensive review, we present the appropriate criteria customized based on expert opinions in Table 1. We also provide the MCDM methods used in the field of 3PLSP selection in Table 2.

The literature review revealed that regardless of the industry under study, some articles such as Liao et al. (2020), Khan et al. (2022), Mohammadkhani and Mousavi (2022), Yuan et al. (2022), Boakai and Samanlioglu (2023), Kalaivani et al. (2023), Cheng et al. (2024) and Sarwar et al. (2024) only used economic criteria to evaluate 3PLSPs. The purpose of these articles is to select cost-effective 3PLSPs with high service quality (Boakai and Samanlioglu, 2023). Because some environmental and social criteria are qualitative and subjective, considering them along with economic criteria leads to an increase in the complexity of the 3PLSP selection problem.

Another reason for ignoring social and environmental criteria by some articles is the lack of legal obligation and external pressures to include these criteria in the evaluation process. For example, in some developing countries, there are still no legal requirements to implement sustainable practices in the daily activities of industries. The noteworthy point is that almost all articles in the field of 3PLSP selection use economic criteria in the evaluation process. Although the selected criteria strongly depend on the studied industry, some economic criteria such as cost, delivery (service), capabilities and geographical location are common in most industries (Sarwar et al., 2024).

Some researchers believe that for implementing the 3PLSP selection problem in polluting industries, environmental criteria should be considered along with economic criteria. For example, Zheng et al. (2023) investigated the 3PLSP selection problem in the battery industry by considering economic and environmental criteria. But some researchers ignored environmental criteria and used economic and social criteria to evaluate 3PLSPs. For instance, Kim et al. (2019) believed that environmental criteria can be ignored in the 3PLSP selection problem in emergency situations. They introduced a total of 14 economic, technical and social criteria to evaluate E3PLSPs. Using decision-making trial and evaluation laboratory (DEMATEL) method, they assessed the relationships among the criteria and weighted the criteria by means of analytic network process (ANP) technique. Finally, they made use of grey complex proportional assessment of alternatives (COPRAS) method to prioritize E3PLSPs.

Some articles go further and consider all three economic, social and environmental criteria in the 3PLSP selection problem. They believe that although social and environmental criteria may not be as important as economic criteria, they should not be left out in the evaluation of 3PLSPs, because considering these criteria leads supply chains and industries toward sustainability (Zarbakhshnia et al., 2018). There are a significant number of articles for the 3PLSP selection problem considering all three economic, social and environmental criteria in various industries such as automotive industry (Mishra et al., 2021; Zarbakhshnia et al., 2020; Zarbakhshnia et al., 2018), food industry (Nila and Roy, 2023; Dadashpour and Bozorgi-Amiri, 2020; Roy et al., 2020), electronics industry (Wang et al., 2024; Mishra et al., 2022) and so forth, which indicates the importance of sustainability in this field.

With the emergence of the fourth industrial revolution, researchers in the fields of supply chain and logistics combined the concept of sustainability with Industry 4.0-based technologies. For example, Tavana et al. (2023b) used sustainability and Industry 4.0 criteria to evaluate suppliers. Also, Sorooshian et al. (2024) presented a mathematical model for designing a supply chain network in the battery industry by integrating the concepts of sustainability and Industry 4.0. But a search in the literature of 3PLSP selection reveals that the application of Industry 4.0 in this field has been less considered. To the best of our knowledge, only one paper has investigated the sustainable 3PLSP selection problem from the perspective of Industry 4.0. Wang et al. (2024) dealt with the sustainable 3PLSP selection problem in the realm of Industry 4.0. They identified the economic, environmental and social criteria under Industry 4.0 perspective. Fuzzy AHP and fuzzy measurement alternatives and ranking according to compromise solution (MARCOS) are used to weight the criteria and prioritize 3PLSPs, respectively. They investigated the effectiveness of their approach by ranking 3PLSPs of a Vietnamese electronics manufacturing firm.

Another difference between this article and previous studies is that instead of green criteria, this article uses circular criteria to evaluate E3PLSPs. The literature review shows that green criteria are focused on reducing the environmental impact of products and processes and are defined based on the principles of environmental sustainability and pollution reduction (Mina et al., 2021). But the circular criteria are defined based on the principles of CE and have a concept beyond the reduction of environmental effects (Govindan et al., 2020). These criteria are defined based on closed-loop, reuse and recycling designs, and their main focus is on waste management, minimizing disposal and generally remaining materials and products in the production and consumption cycle (Tavana et al., 2024b). Therefore, in this research, for the first time, the concept of CE is considered in the E3PLSP selection problem.

In general, previous studies have used various criteria in the 3PLSP selection problem, which can be categorized into three economic, social and environmental dimensions. However, there is a significant gap in the application of CE and Industry 4.0 criteria in the 3PLSP selection problem, especially in the field of disaster management. In other words, although previous studies are a solid foundation for identifying the evaluation criteria of E3PLSPs, they have neglected the benefits of integrating Industry 4.0 and CE in this area. With the aim of bridging this gap, this article develops an effective evaluation approach to evaluate E3PLSPs using MCDM methods. In Table 2, a summary of the articles related to the field of 3PLSP selection that have used MCDM tools is presented to show the research gaps and the innovations of this article.

As Table 2 shows, so far, many articles have used MCDM techniques in the field of 3PLSP selection, and most of them have considered all three aspects of stability in the evaluation process of 3PLSPs. But to the best of our knowledge, only Wang et al. (2024) included Industry 4.0 criteria in this area in addition to sustainability criteria. On the other hand, the literature review revealed that the issue of 3PLSP selection has been studied very little in the field of disaster management. Therefore, in this article, for the first time, the 3PLSP selection problem is examined by considering the criteria of sustainability and Industry 4.0.

AHP method is one of the most well-known MCDM methods that is used in decision-making problems. In this method, to calculate the weight of n criteria, $n(n−1)2$ pairwise comparisons must be made. This means that this method requires a large number of pairwise comparisons. BWM was introduced to meet this challenge (Rezaei, 2015). BWM requires 2n-3 pairwise comparisons to calculate the weights of n criteria. The SWARA method has reduced the number of pairwise comparisons to the minimum possible number, that is, N − 1. On the other hand, unlike BWM, the SWARA method does not require modeling and the use of an optimization software to calculate the weights of the criteria. This means that the SWARA method has no computational complexity and is very user-friendly. Hence, in this article, we applied the fuzzy SWARA method to calculate the independent weights of criteria and sub-criteria. On the other hand, because of the dependency between criteria, we needed a method that calculates the dependency between criteria, easily integrates with weighting methods and does not have computational complexity. The literature review demonstrates that the WINGS method has all these features (Tavana et al., 2023a; Tavana et al., 2022). Hence, in this section, we present a novel approach for ranking E3PLSPs by combining fuzzy SWARA and WINGS methods. The overall structure of the proposed approach is summarized in Figure 1. The developed approach consists of two stages, which are given below:

Stage 1: Calculating the independent weights of criteria and sub-criteria using fuzzy SWARA method

In this stage, the fuzzy SWARA method applied by Agarwal et al. (2020) is used to determine the independent weights of criteria and sub-criteria for evaluating E3PLSPs. In the following, this method is organized in seven steps:

Step 1.1: By reviewing and studying the literature in depth, evaluation sub-criteria are identified from economic, social, circular and Industry 4.0 facets, and the most efficient sub-criteria are extracted in consultation with DMO experts.

Step 1.2: In this step, the evaluation criteria are first sorted from the most important to the least important. Then this operation is implemented on the sub-criteria of each criterion separately.

Step 1.3: In this step, to determine the relative fuzzy importance of criteria (sub-criteria), criterion (sub-criterion) j is compared with criterion (sub-criterion) j − 1 using the linguistic expressions displayed in Table 3. Note that the comparison process starts from the second criterion. $S˜j$ denotes the relative importance of criterion (sub-criterion) j.

Step 1.4: In this step, $K˜j$ is calculated through equation 1. Note that $K˜j$ must be calculated for all criteria and sub-criteria:

Step 1.5: In this step, recalculated fuzzy weight of criterion (sub-criterion) j (⁠$Q˜j$⁠) is determined with the help of equation 2:

Step 1.6: In this step, independent relative fuzzy weight of criterion (sub-criterion) j (⁠$W˜j$⁠) is calculated through equation 3:

where n denotes the number of criteria (sub-criteria).

Step 1.7: In this step, the independent relative fuzzy weights obtained from the previous step are defuzzified via center-of-area technique (Sharma et al., 2022) presented in equation 4:

where $(Wjl,Wjm,Wju)$ represent the independent relative fuzzy weight of criterion (sub-criterion) j, and $Wj$ is the independent relative defuzzified weight of criterion (sub-criterion) j.

Stage 2: Ranking alternatives considering interdependencies among criteria using WINGS technique

In this stage, the WINGS technique used by Tavana et al. (2023a) is used to rank the alternatives considering interdependencies among criteria. This technique consists of five steps as follows:

Step 2.1: In this step, the experts depict the influence of the criteria on each other.

Step 2.2: In this step, by applying the independent relative defuzzified weights calculated in Step 1.7, we calculate a proper range to specify the influence intensity of criteria on each other and to score the alternatives for each sub-criterion. In this vein, equation 5 is first used to rescale the weights obtained from Step 1.7. Then, we use equation 6 to determine internal strength of criteria and their sub-criteria. Finally, based on the minimum and maximum internal strength values, a proper range is provided to determine the influence intensity of the criteria on each other and to score the alternatives for each sub-criterion:

where $ωi$ and $ϕii$ represent the rescaled weight and internal strength of criterion (sub-criterion) j.

Step 2.3: This step deals with forming the initial strength-influence matrix (A). In this vein, a matrix $m×m$ is formed, where m is the sum of the number of criteria, sub-criteria and alternatives. Internal strength values should be placed on the main diameter of matrix A. Furthermore, the influence intensity of criterion i on criterion j is put in row i and column j of matrix A. Also, the score of alternative k in sub-criterion i will be located at the intersection of alternative k and sub-criterion i in matrix A. Finally, the other cells of matrix A will be filled with zeros.

Step 2.4: In this step, the matrix formed in is normalized through the formulas provided in equation 7. Note that B represents the normalized matrix:

where $η$ is the value placed in row i and column j of matrix A.

Step 2.5: In this step, equation 8 is used to calculate the total strength-influence matrix (C):

where I is $m×m$ identity matrix.

Step 2.6: In this step, the total impact indicator is used to rank the alternatives. How to calculate this index is given in equation 9:

where $cij$ denotes the value placed in raw i and column j of the matrix C.

Today, many government organizations have realized that to increase productivity and to improve service quality and affordability, they must outsource some of their projects and organizational tasks to the private sectors. The cooperation between the public and private sectors, called “public-private partnerships,” provides the possibility for the public sector to use technical expertise, operational capabilities and innovation of the private sector to implement projects effectively and enhance provision of services to the public (Tavana et al., 2022). DMO is one of the types of government organizations that collaborate with private sectors to increase its productivity. One of the tasks of this organization is to provide relief to the damaged areas in natural and man-made crises. In this vein, the DMO should identify the most appropriate E3PLSP in the pre-disaster phase and conclude the activities that the private sector should implement in the post-disaster phase in the form of a public–private partnership contract. In this section, the applicability of the developed approach is examined using the data and experts’ knowledge of the DMO in one Iranian city. The demographic information of the studied DMO experts is given in Table 4. The studied DMO intends to select the most appropriate E3PLSP company from the perspective of the triple bottom line and Industry 4.0 and to assign all logistics activities related to the post-disaster phase to it. The overall structure of the studied problem is depicted in Figure 2. Below is the process of choosing the best E3PLSP using the proposed approach:

Stage 1: Applying fuzzy SWARA method to calculate the independent weights of criteria and sub-criteria

In this stage, evaluation sub-criteria from economic, social, circular and Industry 4.0 facets are identified and weighted by fuzzy SWARA method.

Step 1.1: By deeply studying the literature and consulting with the studied DMO experts, we identified 27 sub-criteria for ranking E3PLSPs, which are listed in Table 1.

Step 1.2: In this step, we sort the criteria and sub-criteria according to their importance with the help of experts’ knowledge and experience. The sorted criteria and sub-criteria are given in Table 5.

Step 1.3: In this step, experts compared criterion (sub-criterion) j with criterion (sub-criterion) j − 1 using the language expressions displayed in Table 3. The results achieved from these operations are presented in Table 6.

Step 1.4: In this step, we calculate coefficient $K˜j$ for each criterion and sub-criterion using equation 1, which are presented in Table 7.

Step 1.5: In this step, recalculated fuzzy weights for criteria and sub-criteria are calculated through equation 2, which is denoted in Table 8.

Step 1.6: In this step, independent relative fuzzy weights of criteria and sub-criteria are calculated through equation 3. Table 9 shows these weights.

Step 1.7: In this step, by using equation 4, independent relative defuzzified weights of criteria and sub-criteria are calculated, which is displayed in Table 10.

Stage 2: Applying WINGS technique to rank alternatives

In this stage, taking into account the interdependencies between the criteria, the alternatives are ranked. The ranking process using WINGS technique is as follows:

Step 2.1: In this step, the cause-and-effect diagram related to the criteria was drawn by experts, which is given in Figure 3.

Step 2.2: In this step, we calculate the rescaled weight and internal strength of each criterion and sub-criterion through equations 5 and 6, respectively. Table 11 reports the results obtained from these calculations. Finally, a proper range is defined based on the internal strength values calculated for criteria and sub-criteria. As can be seen in Table 11, the smallest and largest values of internal strength are 0.091 and 0.486, respectively. Therefore, [0,0.5] is a suitable range, where 0 means no influence (strength) and 0.5 means very high influence (strength).

Step 2.3: In this step, the experts determine the influence intensity of the criteria on each other through the extracted range. Then, experts score the five studied E3PLSPs for each sub-criterion using the mentioned range. Finally, the initial strength-influence matrix (A) is structured as given in Figure 4.

Step 2.4: In this step, equation 7 is applied to normalize the matrix A presented in Figure 4. The normalized matrix (B) is given in Figure A1.

Step 2.5: In this step, equation 8 is used to calculate the total strength-influence matrix (C). This matrix is represented in Figure A2.

Step 2.6: In this step, the total impact indicator is calculated using equation 9 for alternatives, then the alternatives are ranked based on this indicator. Table 12 provides the total impact and ranking of alternatives.

The results reported in Table 12 show that E3PLSP 5 (PRV5) is the best E3PLSP compared to other E3PLSPs. By comparing the evaluation scores of E3PLSPs with each other in Table 12, it is concluded that PRV 2 and PRV 5 generally perform better than other E3PLSPs in more than 80% of the sub-criteria, especially the circular and Industry 4.0 sub-criteria. Therefore, assigning the top rank to these two E3PLSPs is in line with reasonable expectations. But the prioritization of PRV2 and PRV5 is not possible except by using an efficient approach, because their performance is very close to each other. Our proposed approach enables the ranking of such E3PLSPs.

In this section, we intend to discuss the effect of dependency between criteria in the ranking of alternatives. While we applied the proposed approach to prioritize alternatives, this approach is also able to calculate the dependent weights of criteria and their sub-criteria. In the following, we will examine this feature of the proposed approach. The process of calculating the dependent weight of the criteria (sub-criteria) using the proposed approach is as follows:

• We calculate the total impact indicator for criteria (sub-criteria) using equation 9.

• Then, we calculate the dependent weights of criteria (sub-criteria) using equation 10.

where $DWi$ shows the dependent weight of criterion (sub-criterion) i.

Therefore, by using the introduced structure, it is possible to calculate the dependent weights of criteria and sub-criteria, which are given in Table 13.

Now the question will be answered, how much does the consideration of dependency between criteria affect the independent weights of criteria and sub-criteria? To answer this question, we must compare the final independent weights of criteria and sub-criteria with their dependent weights. Table 10 shows the independent weights of the criteria and sub-criteria. To calculate the final independent weights of the criteria, we should divide the independent weights of each criterion by the sum of the independent weights of the criteria. Also, for calculating the final independent weights of the sub-criteria, we should first multiply the independent weights of the criteria by the independent weights of their sub-criteria. Then, the value calculated for each sub-criterion must be divided by the sum of the values to obtain the final independent weights of the sub-criteria. Table 14 denotes the final independent weights of criteria and sub-criteria. A comparison between the dependent and independent weights of the criteria and their sub-criteria is provided in Figures 5 and 6. Note that $IWi$ represents the final independent weight of criterion (sub-criterion) i.

The comparison of independent and dependent weights of criteria and sub-criteria in Table 14 indicates that dependency is an important item that should be considered in weighting intertwined criteria. For example, as seen in Table 14, there is a difference of more than 229% between the independent and dependent weights of the sub-criterion INDS3. This difference in weights may affect the ranking of alternatives. To investigate this issue, we rank the E3PLSPs using final independent weights and compare the results with the results obtained from our proposed approach. To rank E3PLSPs without considering the dependency between criteria, we must calculate the sum of the product of the independent weights of sub-criteria in the scores evaluated for E3PLSPs. The results of this operation are presented in Table 15.

The comparison of the results presented in Tables 12 and 15 reveals that by considering dependency, Alternatives 5 and 2 are ranked 1 and 2, respectively, and by ignoring dependency, the rank of these two alternatives is changed and Alternative 2 is selected as the best E3PLSP. Therefore, it can be concluded that in problems where the criteria are intertwined, the dependency between the criteria should not be ignored, because it may lead to incorrect outputs.

There is a consensus among the experts and specialists in this area that the emergency of some service delivery occasions lessens the significance of circularity and sustainability. As a result, this perception may come into play that the efficiency of emergency services may be limited. In this study, it is claimed that the integration of circular economy into the E3PLSP selection problem is both relevant and necessary for the realization of lasting resilience and resource optimization even in the case of emergency situations. The related literature asserts that although the main emphasis in emergency contexts like disasters is placed upon quick response and cost management, the ignorance of circularity may bring about adverse consequences such as environmental degradation, logistical inefficiencies and resource wastage in the long run. Without sacrificing the pace and quality of service delivery in emergency situations, it is possible to liven up sustainability by means of such practices as efficient resource recovery and reuse of materials. However, the incorporation of Industry 4.0 technologies like IoT and blockchain can also help the accomplishment of the above-mentioned practices when resource optimization, enhanced transparency and real-time tracking come into play. In this way, not only does emergency service delivery remain strongly at play but also circularity will be of help to such services. It appears necessary to do more research in this area on advanced technologies to come to a deeper understanding on how to integrate circular principles into the requirements of disaster management so that a framework can be produced wherein immediate needs and sustainable practices are balanced with each other.

While implementing the intended approach, a number of challenges came into play that can be a motif for doing further research. The reliable and valid gathering of data on all 27 sub-criteria required experts’ deep collaboration, which was a key challenge. There was no inclusive agreement among decision-makers regarding the weight of criteria, especially about their interdependencies, which required a large number of discussion rounds. Another challenging issue turned up when circularity and Industry 4.0 were to get incorporated with the evaluation framework, as stakeholders in the disaster management were not familiar with such concepts. Regardless of these challenges, decision-makers considered the structured and systematic nature of this approach to be effective in addressing the E3PLSP selection problem. Although this approach was found effective, decision-makers did not ignore the complexity of putting such approaches into practice when it comes to urgent situations. Therefore, the trend of future research should emphasize the simplification of applying such approaches in real situations and devising practical methods to easily integrate the new dimensions, including circularity and technology in emergency contexts.

The 3PLSP selection problem in the field of disaster management is different from other fields; in this field, the appropriate reaction in a specific time window is very important and cooperation with ineffective E3PLSPs puts human lives at risk. Disaster management activities are divided into two categories identified as pre- and post-disaster activities. The first phase (pre-disaster phase) includes strategic decisions, while the post-disaster phase focuses on operational decisions. Obviously, the decisions made in the pre-disaster phase directly affect the efficiency of the post-disaster phase activities. Decision-makers should identify and evaluate potential E3PLSPs in the pre-disaster phase and contract with the best ones. Decision-makers will be able to choose the best E3PLSP if they identify appropriate criteria for evaluation and use efficient methods for weighting these criteria and prioritizing E3PLSPs. The results of the literature review prove that economic criteria such as cost, geographical location, technological ability and financial stability are key criteria in evaluating both E3PLSPs and 3PLSPs. The results of the literature review prove that the economic, social and environmental criteria are the key criteria in the 3PLSP/E3PLSP selection problem. An E3PLSP who has a high Industry 4.0 ability can monitor logistics activities continuously, prevent fraud and theft, manage inventory efficiently and improve the overall efficiency of the logistics system by applying technologies such as IoT, GNSS, RFID-tags and big data analytics. Therefore, this article, for the first time, considers a set of economic, circular, Industry 4.0 and social criteria in the E3PLSP selection problem.

Furthermore, as mentioned earlier, developing an efficient evaluation and ranking approach in decision-making problems is of great importance. The developed approach should consider all the needs of the problem and at the same time be user-friendly and simple to implement. Hence, this article combines fuzzy SWARA and WINGS techniques for the first time and develops a practical approach to solve decision problems. Few pairwise comparisons, no need for special software, computational simplicity, user-friendliness and consideration of uncertainty in the weighting process are among the advantages of fuzzy SWARA method. On the other hand, WINGS is easily combined with weighting methods and provides an accurate ranking of alternatives in intertwined networks.

In addition to ranking alternatives, the developed approach is able to calculate the weights of factors (criteria and sub-criteria) in an intertwined network. For this purpose, as explained in the previous section, first the total impact should be calculated for each factor, then the total impact value of each factor should be divided by the sum of the total impact values of the factors to calculate the dependent weight of the factors. There may be other solutions to calculate the dependent weight of the factors, which are suggested to be focused on in future research.

The present research can be useful for both academics and decision-makers in the field of disaster management. The proposed approach can solve the challenge of E3PLSP selection in the pre-disaster phase for decision-makers. The output of the proposed approach provides useful information for decision-makers and enables detailed analysis of E3PLSPs for decision- makers. More precisely, the weights obtained from the proposed approach prepare the conditions for the performance analysis of E3PLSPs. Awareness of the weight of the criteria leads to an increase in the analytical power of the decision-makers and makes the investigated issue more concrete for them. On the other hand, in this research, a new vision has been presented to the E3PLSP selection problem, and for the first time, sustainability criteria and Industry 4.0 have been used to evaluate E3PLSPs. In other words, this article suggests a new research area in the E3PLSP selection problem, which can contribute to the literature related to the field of disaster management and smooth the winding research path of this field. In addition, the approach presented in this article can help the development of MCDM methods. By focusing on the developed approach, researchers can identify its weaknesses and, by addressing them, improve the proposed approach. Also, researchers can examine the flexibility of the proposed approach by applying it in other areas such as sustainable supplier selection, prioritization of risks in supply chains, prioritization of the hotels from the sustainability perspective, ranking of barriers to integrate circular economy and Industry 4.0-based technologies in supply chains and so forth.

The current study can make a number of contributions to applying Industry 4.0 and circular economy principles in decision-making during emergency contexts. The development of a hybrid approach consisting of fuzzy SWARA and WINGS in this research is a new methodology that can mitigate the complexities of interdependencies existing in evaluation criteria. While the majority of traditional methods have put their emphasis on linear or independent evaluation criteria, the current framework can cover this niche in the literature.

Theoretically looking at the issue, this research widens the scope of MCDM as various dimensions such as economic, social, technological and circular ones are integrated with each other. Actually, circular economy principles are systematically ignored in emergency contexts. However, the consideration of these principles emphasizes the need for creating a trade-off between immediate disaster response needs and long-term sustainability goals. Accordingly, one of the contributions of this research is its attempt to upgrade the body of knowledge on the sustainability and resilience of supply chain management.

From a methodological point of view, the integration of circularity and Industry 4.0 technologies into decision-making practices is one of the contributions of this study whereby resource optimization and operational efficiency can be boosted. Future research in this area can put focus on integrating more dimensions like environmental justice, ethical considerations and resilience to improve the approach.

These findings provide food for the conduct of further research in this domain. It is a good idea to replicate this research in other geographic contexts or disaster situations so that a stronger generalizability power can come into existence. Moreover, the conduct of more research on circularity and technological capabilities can shed light on this scope, as these are relatively new concepts in disaster management. It can be clarified the way Industry 4.0 technologies like blockchain and IoT can lead to the improvement of traceability, visibility and efficiency of circular practices in disaster management.

In this article, for the first time, the E3PLSP selection problem considering economic, social, circular and Industry 4.0 criteria was studied, and a decision support system for prioritizing E3PLSPs was introduced by integrating fuzzy SWARA and WINGS techniques. It should be noted that the developed approach, in addition to ranking the alternatives, is able to calculate the dependent weights of the criteria (sub-criteria) in a complex and intertwined network. Because of few pairwise comparisons, user-friendliness, consideration of uncertainty in the weighting process, computational simplicity and no need for special software to calculate weights, the proposed approach uses SWARA fuzzy method to determine the independent weights of criteria and sub-criteria. In addition, the WINGS technique calculates the dependent weights of criteria and sub-criteria and prioritizes alternatives by considering interdependencies among criteria. With the help of experts from a DMO, the applicability of the developed approach was evaluated. For this purpose, five E3PLSPs were ranked using 27 sub-criteria including nine economic sub-criteria, six Industry 4.0 sub-criteria, five social sub-criteria and seven circular sub-criteria. In addition, the dependent weights of criteria and sub-criteria were calculated by developed approach and the findings represented that the most important sub-criteria are related to economic and social sub-criteria, respectively.

Although circular and Industry 4.0 criteria have been considered in E3PLSP selection problem for the first time in this research, this problem has not yet been investigated considering Industry 5.0 criteria, which is suggested for future researchers to focus on. In this study, a strategic approach has been developed for prioritizing E3PLSPs. It is suggested to provide a decision support system to optimize the activities assigned to E3PLSPs in the post-disaster phase by using optimization models in future research. In addition, it is suggested to use the balanced scorecard approach to categorize sub-criteria in future research. For this purpose, the identified economic, social, circular and Industry 4.0 sub-criteria are categorized into four groups including financial, customer, growth and learning and internal process. Then, by applying the proposed approach, the dependent weights of the groups and sub-criteria should be calculated. Finally, it is suggested to compare the dependent weights of the sub-criteria obtained in this way with our proposed approach.