Deploying Industry 5.0 drivers to enhance sustainable supply chain risk resilience

ABSTRACT

With global supply chains facing increasing complexity and uncertainty, mitigating sustainable supply chain risks (SSCRs) has become an important research topic. Industry 5.0 (I5.0) is a supplement to, and re-development of, Industry 4.0, but no research on improving supply chain resilience (SCRE) through I5.0 drivers to mitigate SSCRs currently exists. Therefore, we developed a QFD approach to investigate the relationship between SSCRs, SCRE, and I5.0 drivers. This study utilises the FDM-FDISM-ANP-TOPSIS method, considers China’s largest private logistics enterprises as the object of empirical analysis, and establishes a sustainable supply chain identification system and SCRE evaluation index system for logistics enterprises through two-stage House of Quality. The findings reveal that 20 key I5.0 significantly bolster 17 SCRE, thereby mitigating 13 SSCRs by QFD-MCDM framework. These I5.0 drivers play a pivotal role in enhancing SCRE of logistics enterprises. This study contributes to the field of supply chain risk resilience in sustainable engineering by integrating the perspectives of SSCRs, SCRE and I5.0. It underscores the necessity for managers in logistics enterprises to develop SCRE solutions that incorporate both SSCRs and I5.0 considerations. Furthermore, our findings offer valuable insights for other industries aiming to bolster their supply chain competitive advantages in an increasingly complex global landscape.

KEYWORDS:

• Industry 5.0 driver
• sustainable supply chain risk
• supply chain resilience
• quality function deployment

1. Introduction

In the present-day world, the escalating prominence of social issues stemming from air pollution, resource scarcity, and environmental degradation has piqued growing interest in sustainable development (Khan, Zhang, Kumar, et al. 2020; Syed et al. 2019). Scholars and researchers have witnessed a growing interest in sustainable research pertaining to supply chains (Khan, Zhang, Kumar, et al. 2020; X. Y. Xu et al. 2022). Integrating sustainability into supply chain operation management can help enterprises in realising market competitive advantages and planning key business strategies (Khan, Zhang, Kumar, et al. 2020; Koberg and Longoni 2018; H. C. Liu et al. 2019; Raut et al. 2019). Therefore, the implementation of sustainable supply chain management can bolster the competitive advantage of enterprises, improve their responsiveness to social and environmental uncertainties, and enhance both financial performance and economic benefits (Govindan and Gholizadeh 2021; Nikkhil, Manas, and Dibyajyoti 2021).

Improving supply chain resilience (SCRE) is an effective method of mitigating sustainable supply chain risks (SSCRs) (Hosseini, Ivanov, and Dolgui 2019; Kwak, Seo, and Mason 2018; J. Yang et al. 2021). As a dynamic capability, SCRE allows for a quick response to uncertain problems to gain market competitive advantages, thus effectively coping with supply chain risks in an uncertain environment (Ozdemir et al. 2022; N. Singh 2020; J. Um and Han 2021). Because frequent emergencies may damage the sustainable performance of the supply chain, effective SCRE can mitigate the detrimental effects of SSCRs (He et al. 2020). Several studies indicate that Industry 4.0 can improve SCRE and grant enterprises sustainable competitive advantages (Ralston and Blackhurst 2020). However, existing literature predominantly focuses on Industry 4.0, often overlooking the integration of Industry 5.0 (I5.0) in the study of supply chain resilience (SCRE). As a vital progression and enhancement of Industry 4.0, Industry 5.0 is instrumental in augmenting SCRE (Sindhwani et al. 2022). Notably, I5.0, representing the forefront of industrial research, is dedicated to catalysing industrial transformation towards sustainability, human-centricity, and resilience. This shift provides pivotal solutions for the efficient interaction within supply chains (Breque, Nul, and Petridis 2021). Mehdiabadi et al. (2022) underscore that I5.0 prioritises the sustainability of SCRE, which is key to reducing production costs and invigorating supply chains. Consequently, to effectively navigate the risks and dynamics inherent in sustainable supply chains, it is imperative for logistics enterprises to devise SCRE strategies that incorporate both sustainable supply chain risks (SSCRs) and the principles of I5.0.

In recent years, quality function deploy-multiattribute decision-making has been successfully applied to solving the problem of green supply-chain selection (Yazdani et al. 2017), electronic manufacturing supply chain risks (Ralston and Blackhurst 2020), and supply chain traceability issues (Erol et al. 2022). Therefore, we developed a QFD approach to investigate the interrelationships among three groups of variables: sustainable supply chain risks (SSCRs), supply chain resilience (SCRE), and Industry 5.0 (I5.0) drivers. This study aims to establish an integrated QFD- MCDM framework. Utilizing two stages of the House of Quality (HoQ), we connect SSCRs, SCRE, and I5.0 drivers, ultimately prioritising sustainable supply chain risk resilience solutions for logistics enterprises. This study offers practical contributions by enabling logistics enterprises to implement flexible and sustainable supply chain risk resilience strategies. Additionally, stakeholders and participants involved in sustainable supply chain risk resilience are likely to find the outcomes of this research of significant interest. Therefore, the following issues will be discussed in this study:

1. What are the key drivers of SSCRs, SCRE, and I5.0 in the sustainable supply chain of a logistics enterprise?

2. How can the QFD-MCDM framework be integrated to connect three groups of variables and provide decision support for a logistics enterprise’s sustainable supply chain?

3. How can a logistics enterprise integrate I5.0 key drivers into SCRE and mitigate SSCRs?

The remainder of this paper is organised as follows: Section 2 presents a literature review of SSCRs, SCRE, and I5.0. Section 3 introduces the construction of a QFD-MCDM integration framework. Section 4 conducts empirical research and analysis on the logistics enterprise. Finally, Section 5 presents the results, contributions, and future research directions.

2. Literature review

2.1. Sustainable supply chain risks

The investigation of SSCRs offers valuable insights for the sustainable development of enterprises (Abdel-Basset and Mohamed 2019), and researchers have pursued various research directions in the context of sustainability (Amin et al. 2022). Lee and Vachon (2016) proposed that logistics and transportation connections, natural disasters, packaging waste, and other environmental issues disrupt sustainable supply chains. To address these disruptions, Valinejad and Rahmani (2018) proposed a five-dimensional approach for evaluating SSCRs: encompassing the aspects of economy, society, environment, technology, and institutions. Zulfiquar et al. (2020) proposed a fuzzy SWARA-fuzzy COPRAS framework to effectively identify and rank solutions aimed at mitigating risks in sustainable remanufacturing supply chains. Additionally, Moktadir et al. (2021) applied this five-dimensional approach to examine 30 potential sustainable supply chain risk factors in the Bangladesh leather industry. Notably, the study of SSCRs has gained significant attention, and an increasing number of scholars and practitioners are devoted to addressing SSCRs.

The development of an integrated framework to address SSCRs is crucial for achieving sustainable supply chains (M. Xu et al. 2019). Enterprises with a risk identification and assessment framework can effectively prevent and respond to various risks. Therefore, we aim to build a comprehensive framework to achieve a sustainable supply chain. Accordingly, we initially discuss the classification of risk types as classified by various scholars in the recent five years. Liu et al. (2018) divided risk into environment, quality, and operation dimensions. Ganguly and Kumar (2019) divided them into supply, information, financial, and manufacturing risks. Babu et al. (2021) considered nine categories: environmental, information technology, market, supply, delay, transportation, demand, delay, financial, and process risks. Tavakoli Haji Abadi and Avakh Darestani (2021) considered eight types: economic, environmental, organisational, sustainable supply, sustainable production, sustainable distribution, sustainable recovery, and IT-related risks. Alshehri et al. (2022) categorised SSCRs into seven categories: economic, manufacturing and operation, environmental, management policy, quality, market, and social risks.

From the aforementioned literature, 30 SSCRs factors in Appendix A were selected and classified into internal, cooperation, and external risks. Key SSCRs will be evaluated in Section 4 after fuzzy Delphi method (FDM) screening.

2.2. Supply chain resilience

SCRE is crucial for supply chain risk management and refers to the supply chain’s ability to restore its original state or transition to a more favourable state after disruptions (Glickmann and White 2006; Ponomarov and Holcomb 2009; Sheffi and Rice 2005). According to Hosseini et al. (2019), SCRE is defined as an enterprise’s capacity to promptly respond to supply chain disruptions in order to meet customers’ evolving demands. Accordingly, enterprises can quickly respond to disturbances and maintain high perception through SCRE to realise market competitive advantages (Bui et al. 2020; Sá et al. 2020; Stone and Rahimifard 2018). Um and Kim (2019) proposed the significance of SCRE in mitigating supply chain risks, enabling enterprises to sustain their competitiveness, enhance customer service benefits, and foster sustainable development.

Accordingly, over the past five years, researchers have examined various SCRE indicators. Stone et al. (2018) proposed five core SCRE indicators: flexibility, redundancy, risk awareness culture, early warning, and safety. Singh et al. (2019) categorised SCRE into multiple dimensions, including collaboration, sustainability, agility, redundancy, flexibility, visibility, information sharing, robustness, sensitivity, supply chain risk management culture, security, speed, adaptability, market position, risk control, public-private partnership, and supply chain network design. Cerchioneet al. (2020) examined seven SCRE indicators: flexibility, redundancy, visibility, disaster preparedness, market capability, and financial capability. Bayramova et al. (2021) divided SCRE into various components, including redundancy, collaboration, agility, flexibility, human resource management, risk aversion, and culture. Chrisandina et al. (2022) proposed five crucial factors to enhance SCRE: rapid detection, response and recovery, end-to-end visibility and data-driven control, downsizing, collaboration among stakeholders in the supply chain, and effective demand planning. The aforementioned suggests that scholars regard SCRE as having multidimensional capabilities.

From the above literature review, 34 SCRE indicators in Appendix A are selected for this study, and the key SCRE indicators will be evaluated after FDM screening in Section 4.

2.3. Industry 5.0

According to Ostergaard (2016), I5.0 enables workers to have a clearer understanding of their roles within the factory and promotes collaboration between humans and machines. In contrast to Industry 4.0, which emphasises device interconnection, I5.0 focuses on human – machine collaboration at different plant levels (Johansson 2017). Doyle-Kent and Kopacek (2020) proposed that I5.0 is the lifeline of the manufacturing industry and an important source of industrial development mode transformation. Demir et al. (2019) contend that I5.0 can promote worker benefits, investment and consumers, and maximise corporate profits. Motienko (2020) proposed that I5.0 can combine human operational thinking and critical thinking to assign unified and repetitive tasks to robots, thereby improving production efficiency and product quality. Hence, numerous technology visionaries expect that I5.0 will humanise manufacturing (Tange et al. 2020).

I5.0 has been a hot topic among researchers since 2017 (Demir, Döven, and Sezen 2019; Nahavandi 2019; Paschek, Mocan, and Draghici 2019). I5.0 complements the existing Industry 4.0 paradigm by catering to workers’ well-being, focusing on sustainability, and being a resilient facilitator of prosperity (Breque, Nul, and Petridis 2021; Ghobakhloo et al. 2021). Nahavandi (2019) believe that I5.0 can revolutionise the manufacturing system, penetrate into the supply chain and production shop processes, improve production and operation efficiency, and shorten production time and cycle time. Motienko (2020) assessed the potential and advantages of I5.0 in supply chain management and development, and smart healthcare. Frederico (2021) studied the relationship between I5.0 and the supply chain, and proposed that I5.0 could enhance the strategic effect of supply chain through large-scale personalisation of products and services in pursuit of a more intelligent and sustainable supply chain. Zhang et al. (2022) found that the integration of I5.0 into supply chain management helps to enhance resilience. Ghobakhloo et al. (2022) investigated the relationship between I5.0 and sustainable manufacturing, developing a strategic roadmap to elucidate how Industry 5.0 can enhance sustainable manufacturing practices. Nayeri et al. (2023) introduced the concept of Responsive Supply Chain 5.0 and developed an innovative method to integrate the principles of I5.0 into supply chain management, aiming to enhance business responsiveness. Masoomi et al. (2023) posited that Industry 5.0-based solution methodologies have the potential to address sustainable development challenges within the renewable energy supply chain, particularly in solar energy. Ghobakhloo et al. (2023) developed the strategy roadmap for enabling Industry 5.0 transformation and explained how Industry 5.0 transformation should be managed to facilitate sustainable development. Furthermore, Dwivedi et al. (2023) and Sami et al. (2023) have conducted studies on the interactions between Industry 5.0 and circular supply chains, providing insights that inform supply chain management practices.

Based on the aforementioned literature review, this study summarises 123 I5.0 drivers in Appendix A, which are classified into people-oriented, sustainable, resilient, and technological categories. Key I5.0 drivers after FDM screening are evaluated in Section 4.

2.4. SSCRs, SCRE and industry 5.0

At present, global supply chain risk and resilience has been extensively studied. However, SSCRs and SCRE have been little researched (Chen et al. 2022). Moreover, only four studies on the aforementioned have been identified. Sadeghi and Boyer Hasani (2019) developed a multi-objective optimisation model to combine the advantages of supply chain resilience and maximally reduce SSCRs and vulnerability. Lotfi et al. (2021) designed a closed-loop supply chain network for sustainable resilience and studied the interaction between a sustainable supply chain and resilience, considering conditional risk factors. Bhatti and Nawaz (2020) conducted a survey on a sustainable tourism supply chain in Maldives and concluded that risk and resilience management of the tourism supply chain are positively correlated with sustainable supply chain performance. Hsu et al. (2021) conducted an empirical study on the elevator manufacturing industry, demonstrating that key SCRE can be identified and mitigated. Therefore, while studies related to SSCRs and SCRE are crucial, there has been no research conducted from the perspective of logistics.

Studies on the use of I5.0 to improve SCRE are lacking; however, many researchers have found that industry 4.0 can improve SCRE. I5.0, as the supplement and redevelopment of the Industry 4.0 paradigm, is of great research significance for improving SCRE through I5.0. In this study, we discuss the development trend of SCRE and I5.0 by considering the related research on Industry 4.0 and SCRE. Ralston et al. (2020) believe that Industry 4.0 improves supply chain resilience and can facilitate supply chain competitive advantages and improve the business performance of enterprises. Spieske et al. (2021) proposed an Industry 4.0-SCRE framework, emphasising its pivotal role in enhancing SCRE. Mubarik et al. (2021) by utilising supply chain mapping, found that industry 4.0 could improve supply chain resilience, thereby improving enterprise operating efficiency. Tortorella et al. (2022) conducted a review on the contribution of Industry 4.0 to SCRE. The results revealed that Industry 4.0 has a positive impact on SCRE, emphasising its importance in promoting supply chain risk identification. Therefore, the above studies have shown that Industry 4.0 is critical to promote SCRE, and the application of I5.0 to improving SCRE is worthy of further research.

The aforementioned studies reveal that SCRE plays a crucial role in the study of SSCRs, and several researchers have demonstrated the feasibility and effectiveness of utilising SCRE to address SSCRs (Lina et al. 2021, Hsu et al. 2022). In addition, although a small number of I5.0 studies considered the supply chain, no empirical studies have been conducted on both SSCRs and SCRE in the context of I5.0. Through a literature search in databases, Mehdiabadi et al. (2022) identified a significant research gap in I5.0 concerning supply chain topics, particularly the lack of consideration for sustainable supply chain risk. However, to the best of our knowledge, there have been no studies exploring the interrelationship among SSCRs, SCRE, and I5.0. Therefore, this study constructs a two-stage HoQ to link SSCRs, SCRE, and I5.0; determines the key I5.0 drivers to improve SCRE and mitigate SSCRs; and provides a reference for logistics service providers to make sustainable supply-chain decisions.

3. Methodology

A QFD model, drawing on the frameworks established by Bottani (2009) and Lam and Bai (2016), is employed to elucidate the interconnections among SSCRs, SCRE, and I5.0 drivers. The primary advantage of this model lies in its ability to construct relationships between these three variable groups. Specifically, this study incorporates two HoQ within the QFD framework to connect SSCRs with SCRE, and SCRE with I5.0. The first HoQ phase focuses on linking SSCRs and SCRE, primarily to identify which SCRE indicators effectively respond to SSCRs. The second HoQ phase establishes a connection between SCRE and I5.0, aiming to determine which I5.0 drivers significantly enhance SCRE. Furthermore, the SCRE weights calculated in the first HoQ phase serve as a basis for developing the second phase HoQ, influencing the importance weights in this subsequent stage. Figure 1 presents a detailed structure of these two HoQ.

Figure 1. Two-stage HoQ. (a) The first HoQ; (b) the second HoQ.

This study employs FDM, fuzzy interpretative structural modelling (FISM), analysis network procedure (ANP), and TOPSIS method are used to screen out SSCR factors and SCRE indicators (see Appendix B for the specific operation steps). The results reveal the different risk factors and resilience indicators that are adapted to different enterprises. The combination method is used for the following reasons: 1. This study adopts a questionnaire survey, and FDM is currently the mainstream analysis method applied to questionnaire surveys: it can avoid the influence of extreme data. 2. FISM can avoid subjectivity in the experts’ scoring to a certain extent, clearly reflecting the role strength of various factors, and is realistic. 3. Applying ANP and TOPSIS can transform the subjective and qualitative evaluation into a quantitative expression. Therefore, the combination of qualitative and quantitative evaluations can overcome the personal preferences of decision-makers, improve the effectiveness of decision-making, and ensure that the evaluation results are practicable. Therefore, the above methods are adopted in this study to improve data objectivity. In particular, ANP-TOPSIS ensures the data are consistent and not over-processed. The research methodology in this study is shown in Figures 1 and 2. The symbols numbered from (1) to (11) in the text correspond to the relevant steps in Figures 1 and 2.

Figure 2. Research flow chart.

3.1. HoQ1: linking SSCRs and SCRE

First, the initial SSCR factors were identified, and industry experts then assessed the relationship between, and causes of, these factors to design the first round of the FDM questionnaire. Second, the obtained FDM questionnaire values were calculated, and a reasonable threshold was set for evaluation. SSCRs① and SCRE② were selected. Furthermore, the selected SSCRs were evaluated using the FISM-ANP analysis method to obtain the interaction matrix ⑥, and the weight value ⑦ was calculated. The SCRE correlation matrix ② and the SSCRs and SCRE correlation matrix ⑤ are then calculated. Finally, the key SCRE ranking ⑧was obtained based on the interaction between the two variables using the TOPSIS method.

3.2. HoQ2: linking SCRE and I5.0

After determining the initial I5.0 drivers, FDM is used to screen out I5.0 drivers ③. Second, the toughness matrix ④, I5.0, and the toughness correlation matrix are calculated. Finally, the weighting of I5.0 drivers and their priorities are obtained via the TOPSIS method.

4. Empirical study

This study takes the largest private logistics enterprise in China as the case object to validate the proposed framework. The company is a prominent leader in China’s logistics industry and a well-known integrated logistics service provider in China, covering a series of supply chain-related services, including procurement, production, sales, transportation, and after sales. Presently, the supply chain business of the case enterprise is divided into three parts: logistics, finance, and commerce, among which logistics is the main business. The case enterprise is chosen as the research object because of (1) typicality. As a rapidly developing and large-scale logistics enterprise in China, it can well represent the development of China’s logistics industry and supply chain. (2) Sustainability. The case enterprise takes sustainability as a basic goal of its development, embodies the idea of sustainability in the operation of the supply chain, and strives to enhance and improve the operation level of the supply chain. (3) The case company faces severe competition in the external market. In recent years, some deficits in performance have also been observed, and the resilience of the supply chain is being challenged. However, as the enterprise supply chain system becomes large and complex and the supply chain risks become increasingly severe, enterprise managers find it difficult to analyze the supply chain risk factors. Thus, they require appropriate I5.0 drivers to improve supply chain resilience. Furthermore, managers often face challenges in prioritizing SSCRs, SCRE, and I5.0 components. To address this, our study engaged eight experienced supply chain experts from the case enterprise for in-depth, structured interviews. These experts, selected for their extensive knowledge and proficiency in relevant areas, provided invaluable insights for the analysis of SSCRs. Based on their collective input, we formed an overarching judgment. The data gathered and the subsequent analysis were then applied to develop two-stage HoQs, thereby validating the proposed framework. An overview of the experts interviewed is presented in Table 1.

Table 1. Survey of interviewees.

Respondent	Years of experience
Chief engineer	19 years
Production manager	16 years
Supply chain manager	13 years
Materials supervisor	14years
Quality manager	10 years
Inventory control supervisor	15 years
IE (Industrial engineer) manager	9 years
Logistics coordinator	7 years

4.1. First HoQ linking SSCRs and SCRE

4.1.1. Stage I: confirmation of important SSCRs and SCRE using FDM

The FDM questionnaire was designed and distributed based on the initial 30 SSCRs and 34 SCRE, which were calculated using Equation (1)(1) ${\mathrm{G}}^{\mathrm{i}}=\frac{\left[\left({\mathrm{C}}_{\mathrm{U}}^{\mathrm{i}}\times {\mathrm{O}}_{\mathrm{M}}^{\mathrm{i}}\right)-\left({\mathrm{O}}_{\mathrm{L}}^{\mathrm{i}}\times {\mathrm{C}}_{\mathrm{M}}^{\mathrm{i}}\right)\right]}{\left[\left({\mathrm{C}}_{\mathrm{U}}^{\mathrm{i}}-{\mathrm{C}}_{\mathrm{M}}^{\mathrm{i}}\right)+\left({\mathrm{O}}_{\mathrm{M}}^{\mathrm{i}}-{\mathrm{O}}_{\mathrm{L}}^{\mathrm{i}}\right)\right]}$(1) in Appendix B. Following expert discussions, the SSCRs and SCRE thresholds were set as 5.0 and 5.48, respectively, leading to the elimination of factors and indicators below the threshold. The initial 13 factors are identified as the key SSCRs, and the corresponding calculation results are presented in Table 2. Similarly, the first 17 factors are regarded as the key SCRE, as shown in Table 3.

Table 2. Confirmation of important SSCRs using FDM.

NO.	Sustainable supply chain risks	G^i
A1	Communication risk	6.38
A2	Management policy and marketing strategy	5.92
A3	Technical risk	5.86
A4	Risks posed by suppliers	5.76
A5	Demand fluctuation	5.74
A6	The threat of business competition	5.58
A7	Lack of high quality staff	5.47
A8	The financial management ability of a partner or owner	5.37
A9	Tax changes or trade barriers	5.35
A10	Transport disruption risk	5.15
A11	IT infrastructure, Communications and information efficiency	5.11
A12	Environmental pollution	5.02
A13	Local events have disrupted supplies	5.02

Table 3. Important SCRE selected by FDM.

NO.	Supply chain resilience index	G^i
B1	Good customer service	7.25
B2	Efficiency	6.84
B3	Agility	6.69
B4	Predictive ability	6.33
B5	Human resource management	6.29
B6	Market positioning	6.27
B7	Analytical ability	6.25
B8	Sustainability	5.94
B9	Communication	5.83
B10	Supply chain structure	5.76
B11	Consciousness of risk	5.74
B12	Innovation ability	5.68
B13	Flexibility	5.58
B14	Confidence	5.57
B15	Production capacity	5.56
B16	Financial capacity	5.52
B17	Culture	5.48

Following the screening process, the original 30 SSCRs were reduced to 13 via FDM, while the original 34 SCRE were screened to 17. Table 2 presents the three most significant factors: communication risk, management policy and marketing strategy, and technical risk. The top-three factors in Table 3 are good customer service, efficiency, and predictive ability. These key SSCRs and SCRE are analysed in sections ① and ② of the first HoQ, as shown in Figures 1–3.

Figure 3. The first HoQ between key SSCRs and SCRE.

4.1.2. Stage II: verification of the interaction between key SCRE

To explore the interaction relationship between key SCRE by the screening results of FDM, the 17 SCRE matrices were further analysed. The expert group evaluated the interaction matrix to determine the interaction between each SCRE. As the upper triangle of the autocorrelation matrix of SCRE is equal to the lower triangle, the arithmetical average of the eight questionnaires recovered is calculated as its autocorrelation matrix. This matrix is part ④ of the first HoQ, and its autocorrelation matrix is presented in Table 4.

Table 4. Autocorrelation matrix of key SCRE indexes.

	B1	B2	B3	B4	B5	B6	B7	B8	B9	B10	B11	B12	B13	B14	B15	B16	B17
B1	1.00	3.50	3.38	1.63	3.75	6.88	3.13	3.00	5.25	1.50	2.88	1.75	2.50	4.13	3.63	2.63	3.38
B2	3.50	1.00	2.75	2.63	2.88	2.13	2.25	2.50	2.50	3.13	3.25	2.63	5.63	1.63	3.13	3.00	3.75
B3	3.38	2.75	1.00	2.38	2.38	2.50	2.25	3.50	1.75	2.50	3.50	2.25	5.38	2.38	3.88	3.38	2.13
B4	1.63	2.63	2.38	1.00	1.75	2.38	2.50	3.25	2.50	2.25	2.75	2.63	2.63	2.00	2.13	2.88	1.38
B5	3.75	2.88	2.38	1.75	1.00	3.00	5.50	5.00	5.00	4.75	5.00	5.75	5.00	2.13	5.63	6.75	7.00
B6	6.88	2.13	2.50	2.38	3.00	1.00	6.25	5.00	4.25	6.00	6.13	5.75	1.88	2.50	2.88	5.50	4.88
B7	3.13	2.25	2.25	2.50	5.50	6.25	1.00	3.75	2.50	3.88	4.13	4.38	4.13	2.13	2.00	3.25	2.00
B8	3.00	2.50	3.50	3.25	5.00	5.00	3.75	1.00	2.88	7.25	6.50	5.13	3.88	3.25	3.63	6.75	6.00
B9	5.25	2.50	1.75	2.50	5.00	4.25	2.50	2.88	1.00	5.75	2.25	3.75	4.00	5.75	2.88	2.50	6.50
B10	1.50	3.13	2.50	2.25	4.75	6.00	3.88	7.25	5.75	1.00	4.13	2.63	5.75	3.13	2.13	2.75	1.75
B11	2.88	3.25	3.50	2.75	5.00	6.13	4.13	6.50	2.25	4.13	1.00	2.13	6.50	2.75	3.00	2.75	6.13
B12	1.75	2.63	2.25	2.63	5.75	5.75	4.38	5.13	3.75	2.63	2.13	1.00	2.50	3.13	7.00	4.25	2.88
B13	2.50	5.63	5.38	2.63	5.00	1.88	4.13	3.88	4.00	5.75	6.50	2.50	1.00	3.38	3.50	3.88	1.63
B14	4.13	1.63	2.38	2.00	2.13	2.50	2.13	3.25	5.75	3.13	2.75	3.13	3.38	1.00	3.75	3.00	2.38
B15	3.63	3.13	3.88	2.13	5.63	2.88	2.00	3.63	2.88	2.13	3.00	7.00	3.50	3.75	1.00	2.88	1.50
B16	2.63	3.00	3.38	2.88	6.75	5.50	3.25	6.75	2.50	2.75	2.75	4.25	3.88	3.00	2.88	1.00	1.63
B17	3.38	3.75	2.13	1.38	7.00	4.88	2.00	6.00	6.50	1.75	6.13	2.88	1.63	2.38	1.50	1.63	1.00

4.1.3. Stage 3: construct SSCRs and SCRE association matrix

SSCRs and SCRE have a mutual influence relationship, but the influence relationship between the pairwise and degree of influence differs from the autocorrelation toughness matrix: a pairwise influence relationship exists. Therefore, the arithmetic mean of SSCRs and SCRE is calculated to form the correlation matrix of SSCRs and SCRE, which is the fifth part of the first HoQ. The matrix is shown in Table 5.

Table 5. Construct SSCRs and SCRE association matrix.

AB	B1	B2	B3	B4	B5	B6	B7	B8	B9	B10	B11	B12	B13	B14	B15	B16	B17
A1	6.25	2.75	5.50	5.25	7.25	2.00	6.00	5.00	7.50	5.75	6.75	3.25	5.00	6.50	5.75	4.00	6.75
A2	6.25	2.00	7.25	7.50	6.00	7.25	7.00	7.00	6.50	6.25	7.00	6.50	2.50	6.75	6.50	6.00	6.50
A3	2.50	4.75	7.25	6.75	5.50	5.50	6.00	7.00	3.75	5.75	6.50	7.00	5.25	3.50	5.00	5.50	4.00
A4	3.25	2.25	5.25	6.00	3.25	6.00	6.25	3.75	5.50	7.00	6.50	6.50	5.50	6.50	6.00	4.00	1.75
A5	4.50	6.50	5.50	7.00	1.63	6.50	7.50	6.25	6.00	6.50	6.75	2.75	6.75	6.00	6.75	5.00	1.25
A6	4.25	5.25	2.75	5.00	4.75	6.50	6.50	5.50	2.75	6.25	6.25	6.75	5.00	3.00	6.25	5.00	3.00
A7	4.25	2.50	3.75	4.75	7.75	4.13	6.50	4.00	2.50	3.75	4.75	7.50	4.00	1.50	5.25	4.75	6.25
A8	3.25	2.25	2.13	1.75	4.75	2.00	5.25	5.75	2.25	5.00	4.50	5.75	5.25	3.00	4.75	5.00	5.25
A9	2.50	3.00	1.38	2.25	1.75	3.75	2.50	4.75	1.25	2.25	4.25	1.50	1.63	2.00	2.25	2.75	1.00
A10	2.25	5.50	5.25	5.75	2.13	3.50	6.75	6.50	1.63	5.25	6.50	1.63	2.75	3.25	5.00	5.00	1.13
A11	1.00	5.75	6.25	6.50	4.00	4.50	5.75	6.50	2.50	5.75	5.25	6.75	4.75	3.75	6.00	6.00	0.50
A12	0.75	4.50	3.00	3.00	1.63	4.00	1.75	7.00	1.50	3.25	4.50	3.25	3.63	2.75	4.00	2.25	3.50
A13	3.25	5.25	5.75	5.75	2.50	3.75	2.38	4.75	5.63	5.25	6.00	1.50	3.75	4.75	6.25	4.00	1.75

4.1.4. Stage 4: construct SSCRs matrix and ranking

FISM can be used to explore the relationship between any two elements. The fuzzy association strength matrix of the SSCRs was constructed using Equation (2)(2) $\mathrm{B}={\mathrm{b}}_{\mathrm{ij}}={\mathrm{f}}_{\mathrm{ij}}/{\mathrm{f}}_{\mathrm{i}}+{\mathrm{f}}_{\mathrm{j}}\text{---}{\mathrm{f}}_{\mathrm{ij}}$(2) in Appendix B, as listed in Table 6.

Table 6. Construct SSCRs fuzzy association strength matrix.

	A1	A2	A3	A4	A5	A6	A7	A8	A9	A10	A11	A12	A13
A1	0.00	2.44	1.56	2.06	2.39	1.44	1.11	1.39	1.11	1.22	1.06	1.17	1.44
A2	1.67	0.00	2.11	1.89	1.78	2.11	2.11	2.28	1.17	1.89	2.06	1.50	1.44
A3	1.44	2.11	0.00	1.44	1.56	2.06	1.56	1.50	1.50	1.44	2.50	1.39	1.50
A4	2.22	1.61	0.61	0.00	2.39	2.00	0.56	0.61	0.56	1.50	0.56	1.56	1.44
A5	1.44	2.56	1.94	2.06	0.00	2.50	1.00	2.06	1.33	1.61	1.44	1.44	2.50
A6	1.44	2.44	2.11	2.39	2.39	0.00	1.61	1.50	1.44	1.61	1.39	1.06	2.11
A7	1.44	2.06	1.50	1.44	1.06	1.50	0.00	1.50	1.06	1.39	1.50	1.33	1.50
A8	1.56	1.83	1.22	2.00	2.06	2.50	1.83	0.00	2.50	1.61	2.17	1.94	2.22
A9	1.06	2.11	1.50	2.39	1.39	2.39	1.39	2.28	0.00	1.61	2.44	1.11	1.00
A10	1.39	0.89	1.06	2.50	2.17	1.50	1.33	1.44	2.00	0.00	1.44	0.56	2.11
A11	1.72	1.44	2.61	1.00	1.06	1.50	1.39	1.94	1.06	0.89	0.00	0.94	1.06
A12	0.56	1.39	1.06	0.61	1.06	1.50	1.00	1.50	2.00	0.94	2.06	0.00	1.28
A13	1.50	2.33	1.06	1.89	2.22	1.44	1.33	1.72	1.11	2.11	1.89	1.06	0.00

When the threshold value of the fuzzy adjacency matrix of the SSCRs key factors is set to 2.3, Equation (3)(3) $a_{\mathit{ij}}=\left\{\begin{array}{c}1,a_{\mathit{ij}}\ge \mathit{\lambda }\\ 0,b_{\mathit{ij}}<\mathit{\lambda }\end{array}\right.$(3) in Appendix B is used to transform Table 6 into an adjacency matrix, where 1 indicates that the factors in the left column are directly related to the uplink factors, and 0 indicates that they are unrelated. The transformation results of the adjacency matrix of the key risk factors for a sustainable supply chain are listed in Table 7.

Table 7. Construct adjacency matrix of key SSCRs.

	A1	A2	A3	A4	A5	A6	A7	A8	A9	A10	A11	A12	A13
A1	0	1	0	0	1	0	0	0	0	0	0	0	0
A2	0	0	0	0	0	0	0	0	0	0	0	0	0
A3	0	0	0	0	0	0	0	0	0	0	1	0	0
A4	0	0	0	0	1	0	0	0	0	0	0	0	0
A5	0	1	0	0	0	1	0	0	0	0	0	0	1
A6	0	1	0	1	1	0	0	0	0	0	0	0	0
A7	0	0	0	0	0	0	0	0	0	0	0	0	0
A8	0	0	0	0	0	1	0	0	1	0	0	0	0
A9	0	0	0	1	0	1	0	0	0	0	1	0	0
A10	0	0	0	1	0	0	0	0	0	0	0	0	0
A11	0	0	1	0	0	0	0	0	0	0	0	0	0
A12	0	0	0	0	0	0	0	0	0	0	0	0	0
A13	0	1	0	0	0	0	0	0	0	0	0	0	0

In the adjacency matrix, 1 indicates a strong relationship between the two elements, and 0 indicates a weak relationship. An important relationship exists between the management policy and marketing strategy (A2) and communication risk (A1), demand fluctuation (A5), enterprise competitiveness (A6), and local events (A13). Technology risk (A3) is significant for IT infrastructure, communications and information efficiency (A11). The risk (A4) introduced by suppliers is significantly related to the competitive threat (A6), financial risk (A9), and transportation interruption risk (A10) of enterprises. Demand fluctuation (A5) is significantly related to communication risk (A1), supplier risk (A4), and enterprise competitive threat (A6). Enterprise competitive threat (A6) has a significant relationship with demand fluctuation (A5), financial management ability of partners or owners (A8), and financial risk (A9). Financial risk (A9) has an important relationship with financial management ability (A8). IT infrastructure, communications, and information efficiency (A11) poses significant risks to technical risk (A3) and financial risk (A9). Local events (A13) has a significant impact on demand fluctuations (A5).

From the above calculation of the fuzzy interpretation structure model, the structure analysed by FISM is used as the basis of ANP analysis. In this study, a 0–9 scoring method is adopted. The higher the score is, the greater the impact of one element on another element will be. The relationship between various elements can be realised through expert scoring. The data of Expert 1 is taken as an example, and the weights are calculated using Equation (5)(5) $W_{\mathrm{i}}=\frac{1}{\mathrm{n}}\sum _{\mathrm{j}=1}^{\mathrm{n}}\frac{{\mathrm{a}}_{\mathrm{ij}}}{\sum _{\mathrm{k}=1}^{\mathrm{n}}{\mathrm{a}}_{\mathrm{kj}}}$(5) in Appendix B. The pairwise comparison results of each risk are shown in Table 8.

Table 8. Pairwise comparison judgment matrix of original key SSCRs by expert 1.

	A1	A2	A3	A4	A5	A6	A7	A8	A9	A10	A11	A12	A13	Weingt
A1	1.00	0.50	0.50	0.50	0.50	1.00	0.50	0.50	0.33	0.50	1.00	0.50	2.00	0.05
A2	2.00	1.00	0.50	0.50	1.00	2.00	2.00	1.00	1.00	1.00	2.00	1.00	0.33	0.08
A3	2.00	2.00	1.00	1.00	1.00	1.00	2.00	0.50	0.50	2.00	1.00	0.50	3.00	0.09
A4	2.00	2.00	1.00	1.00	1.00	1.00	1.00	2.00	0.50	1.00	1.00	2.00	2.00	0.09
A5	2.00	1.00	1.00	1.00	1.00	2.00	2.00	1.00	1.00	1.00	1.00	1.00	2.00	0.09
A6	1.00	0.50	1.00	1.00	0.50	1.00	2.00	0.50	0.50	0.50	0.50	3.00	2.00	0.07
A7	2.00	2.00	0.50	1.00	0.50	0.50	1.00	0.50	0.50	0.50	0.50	2.00	1.00	0.06
A8	2.00	1.00	2.00	0.50	1.00	2.00	2.00	1.00	0.50	1.00	0.33	1.00	0.50	0.07
A9	3.00	1.00	2.00	2.00	1.00	2.00	2.00	2.00	1.00	1.00	1.00	2.00	2.00	0.11
A10	2.00	1.00	0.50	1.00	1.00	2.00	2.00	1.00	1.00	1.00	1.00	2.00	2.00	0.09
A11	1.00	0.50	1.00	1.00	1.00	2.00	2.00	3.00	1.00	1.00	1.00	1.00	2.00	0.09
A12	2.00	1.00	2.00	0.50	1.00	0.33	0.50	1.00	0.50	0.50	1.00	1.00	0.50	0.06
A13	0.50	3.00	0.33	0.50	0.50	0.50	1.00	2.00	0.50	0.50	0.50	2.00	1.00	0.06

The SSCR weights can be calculated using Equations (5)(5) $W_{\mathrm{i}}=\frac{1}{\mathrm{n}}\sum _{\mathrm{j}=1}^{\mathrm{n}}\frac{{\mathrm{a}}_{\mathrm{ij}}}{\sum _{\mathrm{k}=1}^{\mathrm{n}}{\mathrm{a}}_{\mathrm{kj}}}$(5) and (6(6) $W_{\mathrm{i}}^0=\frac{{\mathit{W}}_{\mathrm{i}}}{\sum _{\mathrm{i}}{W}_{\mathrm{i}}}$(6) ) in Appendix B: the calculation results are listed in Tables 9–16.

Table 9. Weight of each factor of expert 1 under the condition of A2.

A2	A2	A1	A5	A6	A13	$W_i^o$
A2	0.25	0.44	0.20	0.16	0.20	0.25
A1	0.13	0.22	0.20	0.32	0.40	0.25
A5	0.13	0.11	0.10	0.05	0.10	0.10
A6	0.25	0.11	0.30	0.16	0.10	0.18
A13	0.25	0.11	0.20	0.32	0.20	0.22

Table 10. Weight of each factor of expert 1 under the condition of A3.

A3	A3	A11	$W_i^o$
A3	0.75	0.75	0.75
A11	0.25	0.25	0.25

CR = 0.0000 ＜ 0.1.

Table 11. Weight of each factor of expert 1 under the condition of A4.

A4	A4	A6	A9	A10	$W_i^o$
A4	0.17	0.08	0.19	0.16	0.15
A6	0.17	0.08	0.11	0.03	0.10
A9	0.50	0.42	0.56	0.65	0.53
A10	0.17	0.42	0.14	0.16	0.22

CR = 0.0686 ＜ 0.1.

Table 12. Weight of each factor of expert 1 under the condition of A5.

A5	A5	A1	A4	A6	$W_i^o$
A5	0.35	0.30	0.43	0.18	0.32
A1	0.12	0.10	0.09	0.09	0.10
A4	0.35	0.50	0.43	0.64	0.48
A6	0.18	0.10	0.06	0.09	0.11

CR = 0.0288 ＜ 0.1.

Table 13. Weight of each factor of expert 1 under the condition of A6.

A6	A6	A5	A8	A9	$W_i^o$
A6	0.24	0.08	0.19	0.63	0.28
A5	0.24	0.08	0.07	0.13	0.13
A8	0.48	0.42	0.37	0.13	0.35
A9	0.05	0.42	0.37	0.13	0.24

CR = 0.0941 ＜ 0.1.

Table 14. Weight of each factor of expert 1 under the condition of A9.

A9	A9	A8	$W_i^o$
A9	0.50	0.50	0.50
A8	0.50	0.50	0.50

CR = 0.0000 ＜ 0.1.

Table 15. Weight of each factor of expert 1 under the condition of A11.

A11	A11	A3	A9	$W_i^o$
A11	0.29	0.40	0.25	0.31
A3	0.14	0.20	0.25	0.20
A9	0.57	0.40	0.50	0.49

CR = 0.0170 ＜ 0.1.

Table 16. Weight of each factor of expert 1 under the condition of A13.

A13	A13	A5	$W_i^o$
A13	0.50	0.50	0.50
A5	0.50	0.50	0.50

CR = 0.0000 ＜ 0.1.

The data from Expert 1 passed the consistency test. The ANP analysis results evaluated by eight experts are summarised using the above methods, and the average value is the weight of the ANP analysis. The results are presented in Tables 17–25.

Table 17. Importance vector calculation results of key SSCRs.

Factors	Expert 1	Expert 2	Expert 3	Expert 4	Expert 5	Expert 6	Expert 7	Expert 8	Average
A1	0.047	0.058	0.048	0.063	0.047	0.071	0.064	0.065	0.058
A2	0.080	0.079	0.079	0.060	0.073	0.070	0.077	0.050	0.071
A3	0.088	0.074	0.072	0.084	0.071	0.069	0.111	0.091	0.082
A4	0.087	0.086	0.084	0.090	0.049	0.071	0.090	0.072	0.079
A5	0.086	0.084	0.069	0.082	0.125	0.087	0.106	0.068	0.088
A6	0.067	0.083	0.081	0.058	0.051	0.062	0.069	0.075	0.068
A7	0.061	0.061	0.060	0.066	0.108	0.072	0.047	0.082	0.069
A8	0.074	0.088	0.122	0.069	0.094	0.048	0.052	0.097	0.080
A9	0.111	0.085	0.121	0.103	0.091	0.115	0.095	0.087	0.101
A10	0.087	0.084	0.089	0.111	0.048	0.079	0.072	0.082	0.081
A11	0.090	0.073	0.074	0.093	0.172	0.063	0.077	0.098	0.092
A12	0.062	0.058	0.046	0.064	0.038	0.108	0.077	0.069	0.065
A13	0.06	0.09	0.06	0.06	0.03	0.09	0.06	0.07	0.64

Table 18. Importance vector calculation results of eight experts under A2.

A2	Expert 1	Expert 2	Expert 3	Expert 4	Expert 5	Expert 6	Expert 7	Expert 8	Average
A2	0.250	0.348	0.384	0.272	0.269	0.207	0.265	0.279	0.284
A1	0.253	0.177	0.279	0.309	0.322	0.137	0.119	0.167	0.220
A5	0.098	0.111	0.197	0.080	0.090	0.256	0.196	0.183	0.151
A6	0.184	0.187	0.069	0.140	0.178	0.179	0.171	0.148	0.157
A13	0.215	0.177	0.071	0.198	0.142	0.222	0.249	0.223	0.187

Table 19. Importance vector calculation results of eight experts under A3.

A3	Expert 1	Expert 2	Expert 3	Expert 4	Expert 5	Expert 6	Expert 7	Expert 8	Average
A3	0.750	0.889	0.500	0.889	0.750	0.750	0.667	0.333	0.691
A11	0.250	0.111	0.500	0.111	0.250	0.250	0.333	0.667	0.309

Table 20. Importance vector calculation results of eight experts under A4.

A4	Expert 1	Expert 2	Expert 3	Expert 4	Expert 5	Expert 6	Expert 7	Expert 8	Average
A4	0.150	0.288	0.210	0.150	0.156	0.175	0.282	0.233	0.205
A6	0.099	0.205	0.296	0.099	0.183	0.131	0.177	0.281	0.184
A9	0.531	0.252	0.269	0.531	0.397	0.416	0.296	0.291	0.373
A10	0.221	0.255	0.224	0.221	0.263	0.278	0.246	0.195	0.238

Table 21. Importance vector calculation results of eight experts under A5.

A5	Expert 1	Expert 2	Expert 3	Expert 4	Expert 5	Expert 6	Expert 7	Expert 8	Average
A5	0.315	0.240	0.258	0.315	0.364	0.367	0.338	0.289	0.311
A1	0.098	0.105	0.170	0.098	0.315	0.107	0.257	0.247	0.175
A4	0.479	0.349	0.389	0.479	0.225	0.342	0.244	0.289	0.350
A6	0.107	0.306	0.183	0.107	0.096	0.184	0.162	0.176	0.165

Table 22. Importance vector calculation results of eight experts under A6.

A6	Expert 1	Expert 2	Expert 3	Expert 4	Expert 5	Expert 6	Expert 7	Expert 8	Average
A6	0.274	0.272	0.354	0.283	0.211	0.251	0.142	0.254	0.255
A5	0.191	0.272	0.389	0.130	0.464	0.252	0.200	0.239	0.267
A8	0.340	0.294	0.175	0.347	0.221	0.154	0.329	0.295	0.269
A9	0.194	0.162	0.082	0.240	0.104	0.343	0.329	0.212	0.208

Table 23. Importance vector calculation results of eight experts under A9.

A9	Expert 1	Expert 2	Expert 3	Expert 4	Expert 5	Expert 6	Expert 7	Expert 8	Average
A9	0.500	0.500	0.500	0.500	0.417	0.750	0.500	0.800	0.558
A8	0.500	0.500	0.500	0.500	0.583	0.250	0.500	0.200	0.442

Table 24. Importance vector calculation results of eight experts under A11.

A11	Expert 1	Expert 2	Expert 3	Expert 4	Expert 5	Expert 6	Expert 7	Expert 8	Average
A11	0.312	0.388	0.389	0.297	0.321	0.315	0.200	0.321	0.318
A3	0.198	0.443	0.202	0.164	0.454	0.481	0.400	0.454	0.349
A9	0.490	0.169	0.409	0.539	0.225	0.204	0.400	0.225	0.333

Table 25. Importance vector calculation results of eight experts under A13.

A13	Expert 1	Expert 2	Expert 3	Expert 4	Expert 5	Expert 6	Expert 7	Expert 8	Average
A13	0.500	0.750	0.833	0.250	0.333	0.667	0.250	0.458	0.505
A5	0.500	0.250	0.167	0.750	0.667	0.333	0.750	0.542	0.495

The key factors affecting sustainable supply chain risks, A2, A3, A4, A5, A6, A9, A11 and A13, are determined through FISM analysis in the previous section, and the average value of the importance vector of each factor is obtained, which is connected with the adjacency matrix in Table 8 to form a supermatrix with key risks. This matrix is part ⑥ of the first HoQ, and its autocorrelation matrix is presented in Table 26.

Table 26. Key SSCRs autocorrelation matrix.

	A1	A2	A3	A4	A5	A6	A7	A8	A9	A10	A11	A12	A13
A1	1.000	0.220	0	0	0.175	0	O	0	0	0	0	0	0
A2	0	0.284	0	0	0	0	0	0	O	0	0	0	0
A3	0	0	0.691	0	0	0	0	0	0	0	0.349	0	0
A4	0	0	0	0.205	0.350	0	0	0	0	0	0	0	0
A5	0	0.151	0	0.00	0.311	0.267	0	0	0	0	0	0	0.495
A6	0	0.157	0	0.184	0.165	0.255	0	0	0	0	0	0	0
A7	0	0	0	0	0	0	1.000	0	0	0	0	0	0
A8	0	0	0	0	0	0.269	0	1.000	0.442	0	0	0	0
A9	0	0	0	0.373	0	0.208	0	0	0.558	0.00	0.333	0	0
A10	0	0	0	0.238	0	0	0	0	0	1.000	0	0.00	0
A11	0	0	0.309	0	0	0	0	0	0	0	0.318	0	0
A12	0	0	0	0	0	0	0	0	0	0	0	1.000	0
A13	0	0.187	0	0	0	0	0	0	0	0	0	0.00	0.505

By multiplying the ANP weight by the key SSCRs autocorrelation matrix, the integrated weight of SSCRs can be obtained and sorted according to the integrated weight. The SSCRs factor weight and ranking result is listed in Table 27 and can be used as part ⑦ of the first HoQ.

Table 27. Weights ranking results of key SSCRs.

Factors	ANP weights	Integrated weights	Ranking
A1	0.058	0.088761	5
A2	0.071	0.020106	13
A3	0.082	0.089254	4
A4	0.079	0.047054	11
A5	0.088	0.088149	6
A6	0.068	0.057506	9
A7	0.069	0.069467	7
A8	0.080	0.143302	1
A9	0.101	0.130605	2
A10	0.081	0.100152	3
A11	0.092	0.054842	10
A12	0.065	0.065127	8
A13	0.064	0.045676	12

Table 27 reveals that the top-five key SSCRs factors in terms of weight rank are financial management ability (A8), changes in tax collection and barriers (A9), transportation interruption risk (A10), technology risk (A3), and communication risk (A1).

4.1.5. Stage 5: TOPSIS ranking between SSCRs and SCRE

The key SSCRs autocorrelation matrix, SSCRs and SCRE autocorrelation matrix, and SCRE autocorrelation matrix inTables 27, 7, and 6, respectively, were multiplied to obtain a comprehensive initial matrix for TOPSIS analysis of SSCRs and SCRE, as presented in Table 28.

Table 28. TOPSIS initial matrix of supply chain resilience index analysis.

	A1	A2	A3	A4	A5	A6	A7	A8	A9	A10	A11	A12	A13
B1	828.47	195.72	604.55	325.24	668.67	440.37	497.28	698.52	654.15	589.20	357.75	364.28	365.64
B2	828.42	197.65	575.20	316.32	647.52	434.06	499.94	688.47	634.68	564.74	339.06	328.53	355.11
B3	809.73	188.70	564.73	311.70	643.16	428.20	496.47	700.50	630.88	555.13	333.94	338.34	346.93
B4	785.11	183.75	553.97	303.55	625.40	414.87	473.75	676.36	613.60	543.68	327.13	335.25	339.58
B5	747.95	175.29	525.99	291.16	597.65	391.93	435.50	627.00	575.64	521.50	311.24	316.25	327.85
B6	746.06	171.78	520.50	286.63	590.10	386.59	436.53	620.92	569.30	520.17	308.07	306.56	322.88
B7	772.18	182.19	550.44	298.73	626.54	406.73	456.28	641.36	602.00	532.01	325.13	335.28	347.08
B8	817.29	191.50	565.57	314.51	646.81	428.42	492.25	667.34	623.35	551.48	333.88	318.81	354.92
B9	755.50	181.39	553.79	293.65	609.18	405.47	473.19	658.33	603.21	541.89	327.38	332.34	330.55
B10	776.84	182.75	552.70	299.24	620.63	406.61	469.31	650.77	601.79	541.34	326.49	330.88	339.35
B11	721.01	171.26	525.29	284.62	584.94	386.60	441.70	612.65	570.27	507.49	311.14	309.81	317.17
B12	806.29	187.79	555.17	308.05	642.28	415.08	462.81	654.08	610.45	562.81	327.85	333.38	355.40
B13	759.10	182.41	535.61	289.38	606.19	398.24	454.06	629.61	588.27	535.82	316.82	312.25	335.59
B14	723.37	170.58	517.96	272.52	570.62	379.05	458.97	619.96	558.84	494.47	305.10	301.84	311.10
B15	806.12	190.76	583.32	310.06	639.33	424.33	491.22	684.02	629.71	561.89	343.90	340.00	348.26
B16	794.55	184.99	565.05	305.86	628.86	417.72	483.00	668.76	616.29	542.86	333.68	339.59	342.75
B17	769.88	181.31	552.24	303.75	622.59	410.36	456.00	642.58	610.49	544.10	327.72	315.72	339.85

Communication (A1), technical risks (A3), the risks of suppliers (A4), demand fluctuations (A5), enterprise competitive threat (A6), lack of high-quality talent (A7), the tax change or barriers (A9), transportation interruption risk (A10), environmental pollution (A12), and local event type (A13) belong to the cost index. Specifically, a small value of this type of indicator is desirable. Therefore, the reciprocal method is adopted to transform it into a beneficial indicator to ensure the whole matrix exhibits the same trend, as presented in Table 29.

Table 29. Syntrended results of SSCRs and SCRE.

	A1	A2	A3	A4	A5	A6	A7	A8	A9	A10	A11	A12	A13
B1	0.0012	195.7215	0.0017	0.0031	0.0015	0.0023	0.0020	698.5154	0.0015	0.0017	357.7481	0.0027	0.0027
B2	0.0012	197.6498	0.0017	0.0032	0.0015	0.0023	0.0020	688.4686	0.0016	0.0018	339.0604	0.0030	0.0028
B3	0.0012	188.6997	0.0018	0.0032	0.0016	0.0023	0.0020	700.4978	0.0016	0.0018	333.9374	0.0030	0.0029
B4	0.0013	183.7516	0.0018	0.0033	0.0016	0.0024	0.0021	676.3611	0.0016	0.0018	327.1274	0.0030	0.0029
B5	0.0013	175.2926	0.0019	0.0034	0.0017	0.0026	0.0023	627.0005	0.0017	0.0019	311.2427	0.0032	0.0031
B6	0.0013	171.7816	0.0019	0.0035	0.0017	0.0026	0.0023	620.9182	0.0018	0.0019	308.0652	0.0033	0.0031
B7	0.0013	182.1871	0.0018	0.0033	0.0016	0.0025	0.0022	641.3648	0.0017	0.0019	325.1350	0.0030	0.0029
B8	0.0012	191.5011	0.0018	0.0032	0.0015	0.0023	0.0020	667.3388	0.0016	0.0018	333.8754	0.0031	0.0028
B9	0.0013	181.3867	0.0018	0.0034	0.0016	0.0025	0.0021	658.3347	0.0017	0.0018	327.3772	0.0030	0.0030
B10	0.0013	182.7511	0.0018	0.0033	0.0016	0.0025	0.0021	650.7674	0.0017	0.0018	326.4862	0.0030	0.0029
B11	0.0014	171.2632	0.0019	0.0035	0.0017	0.0026	0.0023	612.6462	0.0018	0.0020	311.1383	0.0032	0.0032
B12	0.0012	187.7901	0.0018	0.0032	0.0016	0.0024	0.0022	654.0776	0.0016	0.0018	327.8524	0.0030	0.0028
B13	0.0013	182.4054	0.0019	0.0035	0.0016	0.0025	0.0022	629.6105	0.0017	0.0019	316.8160	0.0032	0.0030
B14	0.0014	170.5810	0.0019	0.0037	0.0018	0.0026	0.0022	619.9636	0.0018	0.0020	305.0980	0.0033	0.0032
B15	0.0012	190.7553	0.0017	0.0032	0.0016	0.0024	0.0020	684.0166	0.0016	0.0018	343.8959	0.0029	0.0029
B16	0.0013	184.9886	0.0018	0.0033	0.0016	0.0024	0.0021	668.7581	0.0016	0.0018	333.6844	0.0029	0.0029
B17	0.0013	181.3139	0.0018	0.0033	0.0016	0.0024	0.0022	642.5768	0.0016	0.0018	327.7231	0.0032	0.0029

The dimensionality of the matrix is processed using Equation (10)(10) $r_{\mathit{ij}}=\frac{{\mathit{x}}_{\mathit{ij}}}{\sqrt{{\sum }_{i=1}^m{x}_{ij}^2}}(\mathit{i}=1,2\dots \mathit{m},\mathit{j}=1,2\dots \mathit{n}$(10) in Appendix B, and the normalised initial matrix is obtained, as listed in Table 30.

Table 30. Key SCRE formalization initial matrix in TOPSIS.

	A1	A2	A3	A4	A5	A6	A7	A8	A9	A10	A11	A12	A13
B1	0.2275	0.2584	0.2213	0.2237	0.2250	0.2254	0.2282	0.2583	0.2239	0.2225	0.2653	0.2170	0.2249
B2	0.2275	0.2610	0.2326	0.2300	0.2323	0.2286	0.2270	0.2546	0.2308	0.2321	0.2514	0.2406	0.2316
B3	0.2328	0.2492	0.2369	0.2335	0.2339	0.2318	0.2286	0.2590	0.2321	0.2361	0.2476	0.2337	0.2371
B4	0.2401	0.2426	0.2415	0.2397	0.2405	0.2392	0.2395	0.2501	0.2387	0.2411	0.2426	0.2358	0.2422
B5	0.2520	0.2315	0.2544	0.2499	0.2517	0.2532	0.2606	0.2318	0.2544	0.2513	0.2308	0.2500	0.2509
B6	0.2526	0.2268	0.2571	0.2539	0.2549	0.2567	0.2600	0.2296	0.2573	0.2520	0.2284	0.2579	0.2547
B7	0.2441	0.2406	0.2431	0.2436	0.2401	0.2440	0.2487	0.2371	0.2433	0.2464	0.2411	0.2358	0.2370
B8	0.2306	0.2529	0.2366	0.2314	0.2326	0.2316	0.2305	0.2468	0.2350	0.2377	0.2476	0.2480	0.2317
B9	0.2495	0.2395	0.2416	0.2478	0.2469	0.2448	0.2398	0.2434	0.2428	0.2419	0.2427	0.2379	0.2488
B10	0.2426	0.2413	0.2421	0.2432	0.2424	0.2441	0.2418	0.2406	0.2434	0.2421	0.2421	0.2389	0.2424
B11	0.2614	0.2261	0.2547	0.2557	0.2572	0.2567	0.2569	0.2265	0.2568	0.2583	0.2307	0.2552	0.2593
B12	0.2338	0.2480	0.2410	0.2362	0.2342	0.2391	0.2452	0.2418	0.2399	0.2329	0.2431	0.2371	0.2314
B13	0.2483	0.2408	0.2498	0.2515	0.2482	0.2492	0.2499	0.2328	0.2490	0.2446	0.2349	0.2532	0.2451
B14	0.2606	0.2252	0.2583	0.2670	0.2636	0.2618	0.2473	0.2292	0.2621	0.2651	0.2262	0.2619	0.2644
B15	0.2338	0.2519	0.2294	0.2347	0.2353	0.2339	0.2310	0.2529	0.2326	0.2333	0.2550	0.2325	0.2362
B16	0.2372	0.2443	0.2368	0.2379	0.2392	0.2376	0.2350	0.2473	0.2376	0.2414	0.2474	0.2328	0.2399
B17	0.2448	0.2394	0.2423	0.2396	0.2416	0.2418	0.2489	0.2376	0.2399	0.2409	0.2430	0.2504	0.2420

The matrices of the positive and negative ideal solutions were obtained by processing Equations (12)(12) $\mathit{V}={\left[{\mathit{v}}_{\mathit{ij}}\right]}_{\mathit{m}\times \mathit{n}}={\begin{array}{c}\begin{array}{cccc}{C}_1& C_2& \cdots & C_n\end{array}\\ \begin{array}{c}{A}_1\\ A_2\\ \mathit{M}\\ A_m\end{array}\left[\begin{array}{cccc}{W}_1{r}_{11}& W_2{r}_{12}& \cdots & W_i{r}_{1\mathit{n}}\\ W_1{r}_{21}& W_2{r}_{22}& \cdots & W_i{r}_{2\mathit{n}}\\ \cdots & \cdots & \cdots & \cdots \\ W_1{r}_{\mathit{m}1}& W_2{r}_{\mathit{m}2}& \cdots & W_i{r}_{\mathit{mn}}\end{array}\right]\end{array}}_{\mathit{m}\times \mathit{n}}$(12) , (13(13) $e_j=-\frac{1}{ln\mathit{m}}\sum _{\mathit{i}=1}^m{r}_{\mathit{ij}}ln{r}_{\mathit{ij}}$(13) ), (14(14) $w_j=-\frac{1-e_j}{{\sum }_{j=1}^n\left(1-e_j\right)}$(14) ), (15(15) $\mathit{W}=\left(w_1,w_2,\dots ,w_n\right)$(15) ), and (16(16) $v_{\mathit{ij}}=r_{\mathit{ij}}\times w_j$(16) ) in Appendix B, and the positive and negative ideal solutions, Euclidean distance, and relative proximity and their ranking were then obtained according to Equations (18), (19), (20(22) $C_i=\frac{S_i^{-}}{S_i^{+}+S_i^{-}}$(22) ), (21), and (22), as presented in Table 31.

Table 31. Construct SCRE positive and negative ideal solutions, Euclidean distance, relative proximity and ranking.

Object	${\mathrm{V}}^{+}$	${\mathrm{V}}^{-}$	${\mathrm{S}}^{+}$	${\mathrm{S}}^{+}$	Degree of relative proximity	Ranking
B1	0.020	0.018	0.009	0.005	0.329	17
B2	0.020	0.017	0.008	0.005	0.375	16
B3	0.020	0.017	0.007	0.004	0.376	14
B4	0.021	0.017	0.006	0.005	0.438	10
B5	0.020	0.017	0.005	0.007	0.595	4
B6	0.020	0.017	0.005	0.008	0.615	3
B7	0.020	0.017	0.006	0.005	0.462	9
B8	0.020	0.017	0.007	0.004	0.384	13
B9	0.020	0.017	0.005	0.005	0.507	6
B10	0.020	0.017	0.006	0.005	0.466	8
B11	0.020	0.017	0.005	0.008	0.631	2
B12	0.020	0.017	0.007	0.004	0.384	12
B13	0.020	0.017	0.005	0.006	0.565	5
B14	0.020	0.018	0.005	0.009	0.659	1
B15	0.020	0.017	0.007	0.004	0.376	15
B16	0.020	0.017	0.006	0.004	0.403	11
B17	0.021	0.017	0.006	0.005	0.480	7

According to the relative proximity ranking of SCRE analysed by TOPSIS, the top-five indicators are confidence (B14), risk awareness (B11), market positioning (B6), human resource management (B5), and flexibility (B13), which are indices closest to the optimal solution. This section is part ⑧ of the first HoQ, as shown in Figures 1–3.

In the first HoQ, FDM, FISM, ANP, and TOPSIS were combined to obtain the key SSCRs and SCRE rankings, and the results are shown in Figure 3.

4.2. Second HoQ linking SCRE and I5.0 drivers

4.2.1. Stage I: selection of key I5.0 drivers using FDM

Enterprises face gradually increasing uncertainty due to fierce market competition. Logistics enterprises have realised the need to create a supply chain for the logistics service system to quickly respond to uncertainty and deal with risk factors. In this study, SSCRs, SCRE, and I5.0 are considered to broaden the perspective of supply chain management for logistics enterprises.

In this study, FDM and TOPSIS were used to bridge key SCRE and I5.0 drivers. From a literature review and expert analysis, 123 initial drivers were selected. From these, the FDM questionnaire was designed: experts were asked to evaluate the maximum and minimum values of each driver on a scale between 0 and 10. After the questionnaire was recovered, the extreme values (more than two times the standard deviation) were removed, and the values were calculated.

Considering the expert discussion and actual case enterprises, the number of I5.0 drivers was reduced from 123 to 20 through screening and subdivided into four dimensions: people first, sustainability, resilience, and technology. The calculations presented in Table 32 will be included in section ③ of the second HoQ, along with Figures 1, 2, and Figure 4.

Figure 4. Key SCRE and I5.0 drivers in the second HoQ.

Table 32. Key I5.0 drivers selected by FDM.

Dimension	I5.0 Drivers	Number	${\mathrm{G}}_{\mathrm{i}}$
People-oriented	Pay attention to human needs	C1	7.02
	Focus on the collaboration between artificial intelligence and cognition	C2	6.35
	Motivating workers	C3	6.34
	Organizational justice, trust and innovation	C4	6.16
	Pay attention to the different values and needs of society	C5	6.12
Sustainable	Advanced digital technology reduces product waste and costs	C6	6.40
	Responsible consumption and production	C7	6.39
	Industrialization and infrastructure innovation	C8	5.92
	Customize products and services to customers	C9	5.86
	Energy conservation and emission reduction	C10	5.80
Resilience	Adaptability (coping with changes in the outside world)	C11	5.09
	Improve production flexibility and increase control redundancy	C12	5.18
	Raise the level of autonomy and control	C13	5.26
	New operation management mode	C14	5.27
	Improve supply chain recovery and risk mitigation mechanisms	C15	5.35
Technology	Renewable energy and trusted indigenous technologies	C16	6.66
	Financial resources	C17	6.53
	Pressure of market competition	C18	6.48
	Fully integrate enterprise resource planning	C19	6.45
	Cost and expense management	C20	6.44

4.2.2. Stage II: the autocorrelation matrix of I5.0 drivers was constructed

The autocorrelation questionnaire on I5.0 drivers was designed using the 20 I5.0 key driving factors selected via FDM, and the average value of each index was calculated to construct the autocorrelation matrix. The value of is equal to, that is, the upper triangle and the lower triangle of the matrix are the same, and the matrix construction results are listed in Table 33. This comprises part ⑨ of the second HoQ.

Table 33. Autocorrelation matrix of key I5.0 drivers.

	C1	C2	C3	C4	C5	C6	C7	C8	C9	C10	C11	C12	C13	C14	C15	C16	C17	C18	C19	C20
C1	1.000	3.750	2.375	4.125	3.875	2.250	2.375	1.875	3.625	3.625	2.250	2.875	2.500	2.750	2.000	2.125	6.375	5.625	3.750	3.625
C2	3.750	1.000	6.750	2.750	1.375	2.750	2.125	7.375	2.750	4.500	2.250	6.125	3.375	2.875	6.375	3.125	2.500	3.000	3.125	2.500
C3	2.375	6.750	1.000	2.750	2.125	3.500	3.500	4.250	2.875	4.875	2.750	4.875	3.250	3.375	3.750	3.875	1.875	2.500	2.500	3.875
C4	4.125	2.750	2.750	1.000	4.625	3.875	4.625	3.875	6.125	4.875	6.500	2.875	3.875	4.125	3.750	4.125	4.125	5.250	4.750	5.750
C5	3.875	1.375	2.125	4.625	1.000	3.750	5.375	2.125	4.250	4.000	3.125	2.250	4.500	3.125	4.875	3.375	2.875	3.625	3.375	3.000
C6	2.250	2.750	3.500	3.875	3.750	1.000	5.375	4.125	3.250	5.250	2.875	4.125	5.625	3.125	2.375	4.625	1.875	2.750	3.750	3.375
C7	2.375	2.125	3.500	4.625	5.375	5.375	1.000	3.500	5.250	6.125	5.375	4.125	5.000	3.250	4.875	4.625	3.625	4.750	4.875	4.750
C8	1.875	7.375	4.250	3.875	2.125	4.125	3.500	1.000	2.125	4.750	2.875	6.750	3.375	3.000	6.000	3.375	2.500	2.625	2.750	4.375
C9	3.625	2.750	2.875	6.125	4.250	3.250	5.250	2.125	1.000	4.250	5.125	3.000	3.125	6.125	3.000	3.875	5.125	5.500	6.125	4.875
C10	3.625	4.500	4.875	4.875	4.000	5.250	6.125	4.750	4.250	1.000	4.500	4.750	4.500	3.375	3.375	4.000	4.000	4.250	3.625	3.625
C11	2.250	2.250	2.750	6.500	3.125	2.875	5.375	2.875	5.125	4.500	1.000	2.625	4.875	5.000	2.375	3.375	4.250	5.375	4.125	5.000
C12	2.875	6.125	4.875	2.875	2.250	4.125	4.125	6.750	3.000	4.750	2.625	1.000	4.625	3.125	3.625	3.625	1.875	2.250	2.500	3.000
C13	2.500	3.375	3.250	3.875	4.500	5.625	5.000	3.375	3.125	4.500	4.875	4.625	1.000	2.750	2.875	5.125	2.375	3.125	4.875	4.750
C14	2.750	2.875	3.375	4.125	3.125	3.125	3.250	3.000	6.125	3.375	5.000	3.125	2.750	1.000	5.000	3.875	3.750	4.875	4.500	4.250
C15	2.000	6.375	3.750	3.750	4.875	2.375	4.875	6.000	3.000	3.375	2.375	3.625	2.875	5.000	1.000	2.750	2.250	3.000	2.875	3.750
C16	2.125	3.125	3.875	4.125	3.375	4.625	4.625	3.375	3.875	4.000	3.375	3.625	5.125	3.875	2.750	1.000	4.500	3.125	5.250	5.375
C17	6.375	2.500	1.875	4.125	2.875	1.875	3.625	2.500	5.125	4.000	4.250	1.875	2.375	3.750	2.250	4.500	1.000	6.250	3.375	3.750
C18	5.625	3.000	2.500	5.250	3.625	2.750	4.750	2.625	5.500	4.250	5.375	2.250	3.125	4.875	3.000	3.125	6.250	1.000	4.625	4.250
C19	3.750	3.125	2.500	4.750	3.375	3.750	4.875	2.750	6.125	3.625	4.125	2.500	4.875	4.500	2.875	5.250	3.375	4.625	1.000	5.875
C20	3.625	2.500	3.875	5.750	3.000	3.375	4.750	4.375	4.875	3.625	5.000	3.000	4.750	4.250	3.750	5.375	3.750	4.250	5.875	1.000

4.2.3. Stage III: construction the association matrix between SCRE and I5.0 drivers

An influential relationship exists between SCRE and I5.0 drivers. Key I5.0 drivers can effectively improve the SCRE. Therefore, the association matrix between SCRE and I5.0 drivers is herein constructed, which is the primary part of the second HoQ. The matrix is listed in Table 34.

Table 34. Correlation matrix of SCRE and I5.0 drivers.

	B1	B2	B3	B4	B5	B6	B7	B8	B9	B10	B11	B12	B13	B14	B15	B16	B17
C1	1.63	2.88	3.13	3.50	3.13	4.25	3.00	7.13	2.25	4.00	3.25	4.00	2.88	2.50	2.75	4.25	2.63
C2	7.50	4.13	2.63	4.00	7.50	4.38	3.88	4.50	5.13	3.00	2.63	3.13	2.13	4.50	3.75	3.88	5.75
C3	8.25	6.63	2.63	4.13	4.88	5.88	4.75	3.88	6.50	3.13	2.75	5.25	2.50	3.00	4.00	3.00	4.13
C4	3.75	4.00	4.75	3.13	4.25	4.13	3.75	4.88	2.50	4.75	3.88	4.50	5.25	2.50	5.75	8.75	2.75
C5	3.00	6.38	4.88	5.75	2.63	3.50	4.63	5.13	2.50	5.25	5.63	3.38	5.13	3.13	5.50	4.63	2.25
C6	2.63	4.38	4.13	5.50	3.63	5.88	6.13	3.63	3.50	6.13	5.00	4.13	3.50	3.50	3.25	3.25	3.38
C7	3.25	4.13	5.88	3.50	3.13	3.38	3.75	4.75	2.63	3.75	4.38	4.38	4.50	2.13	6.25	4.25	3.13
C8	5.75	2.13	4.00	2.88	7.88	3.38	3.00	4.00	5.00	3.25	4.00	4.63	3.50	2.88	3.38	4.13	7.00
C9	3.13	3.75	5.75	3.13	2.13	4.00	3.25	4.38	2.88	4.38	3.38	7.00	3.50	2.38	5.38	5.13	3.13
C10	4.88	3.75	5.13	3.38	3.25	4.63	4.75	5.75	3.00	5.50	5.38	3.25	4.13	4.00	7.38	5.38	4.38
C11	2.88	3.75	5.50	3.13	2.50	3.38	5.00	5.63	3.13	4.13	4.13	4.75	4.00	2.88	4.63	6.25	2.75
C12	4.25	3.13	3.88	3.38	4.88	3.00	4.38	3.38	4.75	3.50	2.63	6.13	2.88	6.00	3.88	3.63	6.63
C13	3.38	4.75	5.50	3.25	4.00	4.88	7.25	4.25	3.00	4.50	3.63	4.50	4.00	3.63	4.63	4.25	3.88
C14	3.38	5.38	4.88	4.63	3.25	4.25	5.13	3.00	3.00	2.75	2.50	5.00	3.00	2.75	5.00	4.00	3.13
C15	3.00	4.25	6.00	2.38	6.00	2.75	3.25	3.13	4.00	3.13	3.50	4.25	4.38	3.50	4.88	3.25	3.75
C16	4.00	5.25	5.25	4.00	4.63	5.25	4.63	4.88	2.75	4.88	2.88	5.25	3.25	3.50	4.13	4.25	4.13
C17	2.13	2.88	3.50	2.38	2.00	2.75	2.38	7.38	2.25	3.00	3.88	5.25	2.88	2.13	4.88	4.25	3.75
C18	2.13	2.75	3.25	2.63	2.13	3.25	2.63	6.75	2.00	4.00	3.25	5.13	2.63	2.50	5.50	4.63	3.13
C19	2.75	5.13	5.50	4.38	3.13	3.88	4.88	3.75	2.00	5.38	5.25	4.63	4.75	2.75	5.63	3.88	2.50
C20	4.50	5.13	5.38	4.63	3.25	5.75	3.75	3.75	3.63	3.25	5.38	5.63	3.75	4.25	5.50	4.13	3.25

4.2.4. Stage Ⅳ: TOPSIS ranking of SCRE and I5.0 key drivers

The initial matrix for TOPSIS analysis was obtained by multiplying the I5.0 key driver autocorrelation matrix and key SCRE index with the I5.0 key driver correlation matrix and key SCRE autocorrelation matrix, as presented in Table 35.

Table 35. TOPSIS initial matrix of key I5.0 driver analysis.

	B1	B2	B3	B4	B5	B6	B7	B8	B9	B10	B11	B12	B13	B14	B15	B16	B17
C1	27726.8	26923.4	26902.3	26363.2	25050.4	24514.1	26047.4	26855.9	25913.7	26099.2	24653.7	26254.4	25249.8	24393.2	27000.2	26354.1	25611.3
C2	31350.4	30531.0	30516.9	29897.2	28210.4	27670.2	29485.6	30611.3	29241.2	29584.8	27915.3	29735.8	28567.2	27730.2	30700.4	29945.3	28987.8
C3	29891.1	29108.3	29043.6	28424.8	26887.9	26351.4	28082.2	29142.0	27918.7	28172.6	26595.6	28418.8	27256.6	26353.7	29216.2	28476.9	27621.6
C4	37144.9	36015.5	35863.7	35252.5	33458.2	32756.4	34793.5	35957.8	34642.5	34880.0	32966.3	35079.0	33735.8	32600.8	36152.3	35300.2	34258.6
C5	29542.0	28640.4	28494.1	28044.4	26570.8	25988.9	27635.1	28600.5	27539.6	27716.4	26210.5	27897.1	26804.2	25931.2	28738.6	28042.2	27239.3
C6	31386.7	30513.4	30441.6	29875.6	28306.0	27724.3	29469.7	30560.6	29336.0	29579.4	27943.2	29784.1	28584.6	27660.3	30632.5	29885.2	29016.6
C7	37968.1	36826.0	36712.9	36039.9	34184.5	33465.0	35548.1	36844.0	35399.6	35631.5	33705.8	35905.8	34492.5	33317.6	36983.0	36046.4	35026.7
C8	32654.3	31757.8	31779.7	31081.5	29402.6	28826.4	30699.0	31888.5	30479.5	30823.7	29105.6	31050.2	29788.6	28778.2	31920.0	31130.1	30228.7
C9	36095.5	34968.1	34902.5	34232.3	32516.0	31833.4	33813.7	34929.0	33686.6	33927.8	32063.0	34159.5	32791.5	31681.0	35121.4	34234.3	33316.9
C10	36964.2	35897.1	35869.9	35116.8	33260.7	32606.6	34684.9	35922.1	34490.2	34804.6	32901.9	34999.9	33627.6	32581.6	36138.4	35160.0	34154.4
C11	33672.7	32683.0	32589.8	32001.4	30375.8	29707.4	31600.2	32682.8	31464.4	31671.8	29947.0	31857.9	30628.2	29616.9	32778.7	31990.7	31100.9
C12	31451.0	30680.6	30664.5	29985.4	28265.9	27766.0	29628.6	30730.6	29417.4	29727.2	28050.1	29977.5	28732.8	27872.0	30839.2	30023.1	29155.8
C13	33945.1	32965.4	32884.0	32216.7	30560.5	29946.6	31866.4	33008.3	31678.3	31901.3	30170.4	32167.7	30876.3	29844.4	33116.2	32272.1	31354.8
C14	32524.9	31580.1	31477.0	30931.6	29327.3	28701.4	30536.9	31554.8	30354.5	30578.5	28897.5	30755.3	29577.6	28617.4	31716.0	30928.0	30051.7
C15	31314.6	30518.3	30544.0	29816.7	28235.6	27711.2	29523.7	30593.6	29305.1	29618.8	27951.5	29831.2	28613.2	27720.8	30679.6	29863.0	29020.8
C16	33606.5	32638.6	32556.3	31941.1	30279.2	29665.9	31512.3	32675.2	31353.3	31623.1	29863.5	31818.3	30564.2	29556.7	32755.8	31959.7	31042.3
C17	30082.3	29175.6	29080.0	28533.8	27096.1	26494.0	28203.5	29067.3	28099.0	28243.3	26739.7	28436.3	27344.6	26422.2	29306.8	28525.0	27771.3
C18	34402.8	33357.3	33246.4	32654.6	31009.7	30343.0	32251.0	33289.4	32110.1	32350.8	30560.8	32536.0	31258.4	30233.8	33528.6	32634.3	31761.2
C19	34543.8	33507.3	33405.1	32800.2	31122.8	30435.2	32386.8	33487.8	32250.6	32487.7	30723.9	32661.7	31412.9	30360.9	33678.6	32788.1	31895.0
C20	35863.8	34805.1	34712.2	34070.3	32291.2	31622.3	33600.3	34780.2	33481.3	33704.0	31894.6	33924.1	32595.8	31576.4	34990.0	34069.0	33111.6

Because all the factors exhibit a positive trend, co-trending is unnecessary. The initial matrix was standardised using Equation (10)(10) $r_{\mathit{ij}}=\frac{{\mathit{x}}_{\mathit{ij}}}{\sqrt{{\sum }_{i=1}^m{x}_{ij}^2}}(\mathit{i}=1,2\dots \mathit{m},\mathit{j}=1,2\dots \mathit{n}$(10) in Appendix B, and the positive and negative ideal solutions, Euclidean distance, and relative proximity were obtained using Equations (17)(17) $\mathit{V}={\left[{\mathit{v}}_{\mathit{ij}}\right]}_{\mathit{m}\times \mathit{n}}$(17) , (18), and (19), respectively. The calculation results are presented in Table 36.

Table 36. Euclidean distance and relative proximity.

Object	${\mathrm{V}}^{+}$	${\mathrm{V}}^{-}$	${\mathrm{S}}^{+}$	${\mathrm{S}}^{+}$	Degree of relative proximity	Ranking
C1	0.015	0.011	0.017	0.000	0.000	20
C2	0.015	0.011	0.011	0.006	0.362	16
C3	0.015	0.011	0.013	0.004	0.216	18
C4	0.015	0.011	0.001	0.015	0.918	2
C5	0.015	0.011	0.014	0.003	0.171	19
C6	0.015	0.011	0.011	0.006	0.362	15
C7	0.015	0.011	0.000	0.004	1.000	1
C8	0.015	0.011	0.008	0.008	0.490	11
C9	0.015	0.011	0.003	0.014	0.816	4
C10	0.015	0.011	0.002	0.015	0.908	3
C11	0.015	0.011	0.007	0.010	0.582	9
C12	0.015	0.011	0.010	0.006	0.376	13
C13	0.015	0.011	0.006	0.010	0.610	8
C14	0.015	0.011	0.009	0.008	0.470	12
C15	0.015	0.011	0.011	0.006	0.363	14
C16	0.015	0.011	0.007	0.010	0.576	10
C17	0.015	0.011	0.013	0.004	0.226	17
C18	0.015	0.011	0.006	0.011	0.651	7
C19	0.015	0.011	0.006	0.011	0.666	6
C20	0.015	0.011	0.003	0.013	0.797	5

The top-five indicators according to the I5.0 relative closeness ranking of TOPSIS analysis are responsible consumption and production (C7), organisational justice, trust and innovation (C4), energy conservation and emission reduction (C10), customising products and services to customers (C9), and cost and expense management (C20).

4.3. Results and discussion

Generally, enterprise resources are limited, making it challenging to enhance SCRE in all aspects simultaneously. To effectively cope with SSCRs, enterprise managers should prioritise SCRE and gradually invest in maximising production and operation benefits because the Pareto effect can be utilised using a few key factors when resource utilisation is effective. Therefore, we applied the Pareto effect to the QFD-MCDM framework in this study. Moreover, we discuss the importance ranking of these three variables with the results of the two HoQ as a reference to improve SCRE and mitigate the SSCRs of a logistics enterprise using key I5.0 drivers, as shown in Figure 5. If all key SCRE can be improved, resistance to SSCRs can be greatly improved.

Figure 5. I5.0 drivers for improving SCRE to mitigate SSCRs in logistics enterprise.

4.3.1. HoQ1: connecting SSCRs and SCRE to get the most critical ranking

In this study, SSCRs and SCRE were connected by the first HoQ. As listed in Table 27, the top-five key SSCRs are enterprise’s financial management capability (A8), changes and barriers in tax collection (A9), transportation disruption risk (A10), technology risk (A3), and communication risk (A1). Table 4 lists the top-five key SCRE indicators as confidence (B14), risk awareness (B11), market positioning (B6), human resource management (B5), and flexibility (B13). The above key SSCRs and SCRE should be the focus for enterprise managers.

As shown in Figures 3 and 5, the key SCRE priorities calculated using TOPSIS can significantly mitigate SSCRs. The TOPSIS ranking results show that confidence (B14) ranks first in SCRE, indicating that confidence is the basis of cooperation in the face of SSCRs. A good confidence relationship can enhance stability upstream and downstream of the supply chain and increase resistance to fierce market competition. If the financial management of the enterprise is poor, for example, a shortage of the capital chain will cause abnormal changes in supply chain orders, it will lead to a crisis of confidence. Christopher et al.(2004) proposed that enhancing confidence can reduce supply chain risks. Therefore, confidence (B14) is helpful for enterprises seeking to mitigate SSCRs, which can be achieved by establishing a sound credit risk management and early warning mechanism.

The second-ranking SCRE is risk awareness (B11), which indicates that enterprises should improve their awareness of risk response as the risks related to the supply chain become increasingly prominent. Two examples can be used to illustrate this. In 2011, due to the tsunami and earthquake in Japan, suppliers of parts to Toyota had an insufficient estimate of natural risks, resulting in the risk of transportation disruption (A10), which resulted in missed delivery schedules (Hosseini, Ivanov, and Dolgui 2019). Another example is the outbreak of COVID-19 in early 2020, which caused unprecedented supply chain disruptions, halted the production and flow of materials, affected the demand and supply side of logistics infrastructure, directly affected personnel and operational needs, and even threatened corporate finances. Logistics service providers must perform operational changes to survive (Wackett 2020). Therefore, establishing risk awareness (B11) and a complete-risk early warning system is crucial, as is presetting different possible risks in advance and formulating corresponding measures. Therefore, increasing risk awareness can mitigate SSCRs, especially for the top-five key SSCRs.

In third place in SCRE is market positioning (B6). Evaluating the enterprise’s financial management ability (A8); improving the service quality to win the trust of customers; strategically adopting a differentiation competitive strategy to break trade barriers; improving the competitiveness of the different business modules; and avoiding transportation disruptions (A10), technical risks (A3), SSCRs (A1), communication, risk among others is necessary for establishing an accurate market position and improving the quality of enterprise service. For example, DHL, the world’s leading postal and logistics group, precisely positioned the market through logistics innovation of emerging green technologies, green production, reducing carbon emissions and energy consumption, and increasing the use of renewable energy to maintain a competitive advantage in the market (Cosimato et al. 2015). Tesla replaced gasoline and diesel with lithium batteries, developed low-carbon vehicles to meet consumers’ demand preferences for low-carbon products, obtained government tax breaks, and avoided traditional automotive industry barriers. In 2020, their total sales volume reached 500,000 units, indicating that market positioning is crucial to the development of enterprises (Ma et al. 2021). Therefore, market positioning (B6) is related to the competitive advantage of enterprises in the market and has a positive role in mitigating SSCRs.

Human resource management (B5) ranked fourth in SCRE. Effective human resource management can avoid risks, enhance risk awareness, and improves the financial management ability of enterprises (A8). It includes paying attention to the satisfaction of employees’ self-development needs, conducting professional skills training, improving the adaptability and application ability of digital technology, and promoting all-round employee development. Humayun (2021) believed that HRM can effectively improve the sustainable development needs of supply chain members. Evidently, human resource management plays an indispensable role in the supply chain. Therefore, improving human resource management (B5) can mitigate the occurrence of SSCRs.

Flexibility (B13) ranks fifth among the SCRE factors. Flexibility plays an important mediating role between strategic partnership and business performance, and can effectively avoid risks such as communication risk (A1) and transportation disruption risk (A10). Yang et al. (2021) demonstrated that flexibility can promote information transmission among suppliers, manufacturers, and retailers; improve the supply chain transparency; and enhance the market risk response ability of enterprises. Therefore, to maintain a flexible supply chain, the relationship with suppliers must be coordinated and a stable supply chain structure established. Second, the flexible use of human resources, funds, and equipment helps to improve the overall financial management ability of enterprises (A8). Finally, building an information resource cooperation platform is necessary to establish a good partnership with suppliers to reduce communication risks (A1). Therefore, flexibility (B13) is critical to mitigating SSCRs.

Therefore, in situations with limited resources, enterprises should focus on the above five most-critical SCRE to mitigate SSCRs and maximise production and operation efficiency. Taking into account the long-term development of enterprises, the other key SCRE identified in this study can also be used as an important reference for enterprises to mitigate SSCRs and gradually applied and promoted.

4.3.2. HoQ2: connecting SCRE and I5.0 to obtain the most critical I5.0 ranking

Among the driving I5.0 factors, 20 important factors are selected from the four dimensions in Table 32, indicating that all the important factors deployed from SCRE in this study can cover the entirety of I5.0.

The human dimension comprises the following: attaching importance to human needs (C1); focusing on collaboration between artificial intelligence and cognition (C2); motivating workers (C3); organisational justice, trust, and innovation (C4); and paying attention to different values and needs of society (C5).

The sustainable dimension comprises the following: advanced digital technologies to reduce product waste and cost (C6), responsible consumption and production (C7), industrialisation, innovation in infrastructure (C8), customising products and services to customers (C9), energy saving and emission reduction (C10).

The ductile dimension comprises the following: adaptability (coping with external changes) (C11), improve production flexibility and increase control redundancy (C12), improve the level of autonomy and control (C13), new operation management model (C14), and improve supply chain recovery and mitigation mechanism (C15).

The technology dimension comprises the following: renewable energy storage and trusted autonomous technology (C16), financial resources, financial resources (C17), market competitive pressure (C18), fully integrate enterprise resource planning (C19), and cost and expense management (C20)

In the second HoQ, SCRE acts as an intermediate bridge connecting the SSCRs and I5.0 drivers. Table 36 reveals that the top-five key I5.0 drivers are responsible consumption and production (C7), organisational justice, trust and innovation (C4), energy saving and emission reduction (C10), customising products and services to customers (C9), and cost and expense management (C20). Being the five most critical I5.0 drivers of the case enterprise, we analyse them in order of rank below.

Responsible consumption and production (C7) enables the case company to align with sustainable and intelligent practices, achieve precise market positioning (B6), establish a perfect privacy protection system to protect the information security of upstream and downstream users of the supply chain as the goal, and improve the confidence (B14) between enterprises. According to Juniati et al. (2020) reviewed and analysed 44 sustainability reports and 453 annual reports of Indonesian listed enterprises from 2014 to 2016 and pointed out that enterprises tend to support responsible consumption and production to realise sustainable development goals (Juniati, Paulina, and Carol 2020). Within the entire supply chain, it is essential to consider not only economic benefits, such as reinforcing product quality and safety oversight to improve enterprise risk awareness (B11) and enhance enterprise flexibility (B13), but also consider human resource management (B5) to form a good working environment and meet the development needs of employees.

Organizational justice, trust, and innovation (C4) can promote confidence (B14) and enhance risk awareness (B11) upstream and downstream of the enterprise, which is conducive to establishing a good organisational atmosphere, improving the job satisfaction and happiness of employees, and making human resource management (B5) more harmonious. Cultivating employees’ innovation ability to grasp the market positioning (B6) can facilitate the sustainable development of the enterprise and make the enterprise flexible. For example, Akram et al. (2020) by investigating the innovative work behaviours of 345 employees in China’s telecom industry, found that organisational justice and trust can promote positive work behaviours in employees, inculcate their willingness to express their ideas and implement them in the organisation, and create a good innovation atmosphere.

Energy conservation and emissions reduction (C10) is an effective path towards sustainable development. Green packing, green transportation, green marketing, and green office imply grasping the risk awareness (B11) to ensure production is in line with the market positioning (B6) product sustainable supply chain, making the enterprise more flexible. For example, DHL, a global leader in logistics, focuses on the Paris Agreement and the United Nations Sustainable Development Goals, optimises energy efficiency through sustainable fuels, reduces dependence on fossil fuels and pollutants, and promotes its sustainable development with sustainable carbon management practices to enhance its competitive edge (Deweerdt, Dargusch, and Hill 2022). Therefore, energy saving and emission reduction (C10) promotes the benign development of the supply chain from the perspective of enterprise strategy and benefits enterprise supply chain management.

Customise products and services to customers (C9), by accurately targeting market positioning (B6), provides customers with customised products and services, ensuring risk awareness (B11). This facilitates developing value-added services and providing integrated supply chain solutions to improve customer viscosity and increase enterprise flexibility (Bl3). For example, Maddikunta et al. (2021) pointed out that mass production of customised and personalised products can improve output and customer satisfaction, thereby generating value-added services and helping supply chain management to reduce costs. Therefore, customising products and services to customers (C9) can ensure a diversified development of the supply chain from the perspective of the enterprise.

Cost and expense Management (C20) indicates that enterprises should pay attention to supply chain collaboration and fine management of costs in each link to prevent capital chain shortage to improve the enterprise risk awareness (B11) and enhance upstream and downstream supply chain confidence (B14). Um and Kim (2019) demonstrated that supply chain collaboration can offer transaction cost advantages and improve the performance of enterprises. Furthermore, cost and expense management (C20) plays an important role in promoting the rapid development of enterprises with the improvement of logistics infrastructure and technological progress, thereby promoting the scale effect. Enterprises can transform from labour intensive to technology intensive, thereby reducing costs and increasing efficiency. Finally, cost and expense management (C20) enables enterprises to focus on their own market positioning (B6) to improve core competitiveness, better human resource management (B5) to improve the overall efficiency of the supply chain, and promote enterprise flexibility (B13).

Enterprises should focus on the five most critical I5.0 drivers above to mitigate SSCRs while improving supply chain resilience. Meanwhile, other important I5.0 drivers can be progressively applied to case enterprise decision support, with adequate resources, to enhance SCRE and mitigate SSCRs for supply chain sustainability.

Enterprises in different industries can also utilise the proposed framework to reduce risk. However, in constructing a risk list, it should be divided according to the risks of different industries. Therefore, this framework should be applied according to the industry characteristics, making full use of the target company’s supply chain experts to analyse and identify the risks of sustainable supply chain. Moreover, the corresponding resilience improvement measures should be conducted in a gradual cycle according to the planning, execution, inspection, and implementation processes, and its performance should be regularly reviewed.

5. Conclusions and recommendations

In the current landscape, logistics enterprises are confronted with increasingly complex risks and uncertainties. These enterprises recognise that disruptions in the supply chain can have widespread impacts on the global supply chain system. Therefore, it is imperative for these enterprises to swiftly mitigate uncertain risks within the supply chain. In response, we developed a QFD approach to explore the interrelationships among SSCRs, SCRE, and I5.0 drivers. This research expands the scope of sustainable supply chain risk resilience management by incorporating considerations of SSCRs, SCRE, and I5.0 drivers. It is crucial for managers to devise SCRE strategies that account for both SSCRs and I5.0, particularly from the perspective of logistics enterprises. Validation conducted on China’s largest private logistics enterprise reveals the following main findings:

• The top-five SSCRs are the enterprise’s financial management ability, changes and barriers in tax collection, transportation disruption risk, technology risk, and communication risk.

• The top-five SCRE are confidence, risk awareness, market positioning, human resource management, and flexibility.

• The top-five I5.0 drivers are responsible consumption and production, organisational justice, trust and innovation, energy saving and emission reduction, customising products and services to customers, and cost and expense management.

The primary contributions of this study are outlined as follows:

Firstly, this research represents a pioneering effort to integrate I5.0 into the sustainable supply chain risk resilience framework of a logistics enterprise. We propose a QFD integrated with MCDM, utilising a combination of FDM, FISM, ANP, and TOPSIS. This approach is designed to elucidate the relationships among key SSCRs, SCRE, and I5.0 drivers, providing decision support to enhance SCRE in logistics enterprises.

Secondly, this paper offers a reference point for future research, presenting a feasible framework for logistics enterprises seeking to apply I5.0 drivers to improve SCRE and mitigate SSCRs.

Finally, the research framework and data analysis presented here can aid logistics enterprises in effectively adjusting their supply chain strategies, operations, and management directions, particularly when leveraging enterprise resources efficiently.

5.1. Theoretical and practical implications

The significance of this study lies in its pioneering exploration of I5.0 within the realm of SSCRs and SCRE. This novel investigation sheds light on the integration of I5.0 into the mitigation of SSCRs and enhancement of SCRE in logistics enterprises. Our findings reveal that the top five SSCRs pose significant challenges to the implementation of sustainable supply chains in these enterprises. Conversely, the five most crucial SCRE and I5.0 drivers have been identified as effective in addressing these challenges, lending practical managerial significance to our research. We recommend that managers prioritise the evaluation and management of these top five factors in SSCRs, SCRE, and I5.0. Additionally, it is advisable for them to tailor these considerations to the specific contexts of their respective enterprises.

This study contributes to the literature on sustainable supply chain management by empirically examining the interplay between SSCRs, SCRE, and I5.0 in logistics enterprises. The insights provided offer theoretical foundations for managers to leverage I5.0 drivers in mitigating sustainable supply chain risks and bolstering supply chain resilience.

Looking ahead, it is imperative for future research to integrate I5.0 with supply chain considerations across various industries. This integration will enable the identification of industry-specific risk factors and the development of tailored I5.0-driven solutions for sustainable supply chain risk resilience.

5.2. Limitations and future implications

While this research provides valuable insights, it is not without limitations, which present opportunities for future studies.

Firstly, the empirical analysis was conducted solely on China’s largest private logistics enterprise, thus not encompassing enterprises from other countries. This geographical limitation suggests that future research could benefit from a broader, more diverse set of enterprises across different countries to enhance the generalizability of the findings.

Secondly, there is an opportunity to extend the research to include medium and small-sized logistics enterprises. Investigating a wider range of company sizes could yield a more comprehensive understanding of the relationship between SSCRs, SCRE, and I5.0 drivers.

Lastly, future studies should consider including suppliers and customers of logistics enterprises as respondents. This would allow for a more holistic analysis of the interplay between SSCRs, SCRE, and I5.0, providing a more rounded perspective.

Additionally, the focus on a single logistics enterprise within China means the study’s implications are primarily relevant to specific national contexts and related industries. This limitation highlights the necessity for caution when generalising the study’s findings to different countries and industrial settings, due to unique country-specific characteristics and behaviours.

Authors contributions

C.-H.H.: Conceptualization, Methodology, Writing-review and editing

J.Z.W: Methodology, Software, Funding acquisition, Formal analysis, Writing-original draft

T.Y.Z: Data curation, Visualization, Supervision

J.Y.C: Visualization, Project administration, Resources

All authors reviewed the results and approved the final version of the manuscript.

Acknowledgments

The authors of this paper would like to acknowledge greatly to all the authors who have published valuable research articles and books that have been used for this review paper.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

Data sharing is not applicable to this article.

Additional information

Funding

This paper was supported by the Special Fund of Education and Scientific Research of Fujian Provincial Finance Department [Grant No. GY-Z21001] and Surface of the State Natural Science Fund projects (61976055).

Notes on contributors

Chih-Hung Hsu

Chih-Hung Hsu received his bachelor’s degree in Industrial Management from Taiwan University of Science and Technology in 1996, master’s degree in Industrial Management from Taiwan University of Science and Technology in 1999, doctor’s degree in industrial Engineering and Engineering Management from Tsinghua University in Hsinchu, Taiwan in 2005, and master’s degree in International logistics from the Department of Shipping Management, Ocean University of Taiwan in 2011. His current research involves multi-attribute decision making and data mining in resilient, agile, green and sustainable supply chains.

Jin-Zheng Wu

Jin-Zheng Wu received the B.Sc. degree in Transportation from the Fujian University of Technology, Fujian, China, in 2020. He is currently pursuing the M.Sc. degree in Mechanical & Industrial Engineering from the Fujian University of Technology, Fujian, China. His main areas of interest are multi-attribute decision making, supply chains and Industry5.0.

Ting-Yi Zhang

Ting-Yi Zhang served as a associate professor of Institute of Industrial Engineering, College of Management, FuJian University of Technology. His main areas of interest are supply chain management, production system scheduling and optimization.

Jia-Yi Chen

Jia-Yi Chen received the B.Sc. degree in Logistics management from the Fujian University of Technology, Fujian, China, in 2022.

Appendix A

Table A1. 30 SSCRs factors.

NO.	Sustainable supply chain risks
1	Equipment failure
2	Product safety and quality
3	Technical risk
4	Inventory risk
5	Reliance on a single supplier, resulting in unexpected delivery delays
6	Storage space
7	Violation supervision during operation
8	Lack of highly qualified staff with key competencies/skills
9	Supplier selection/outsourcing risk
10	Management policy and marketing strategy
11	IT infrastructure, information systems and communication efficiency
12	Information risk (asymmetric information, information not shared, incorrect integration or accuracy of information)
13	Terrorist attack
14	The impact of local events (Labour strikes, industrial accidents) resulting in supply disruptions or loss of production capacity
15	Delay in delivery
16	Risks posed by suppliers (e.g. unqualified supply quality)
17	Financial management ability of partner or owner
18	Communication risk (resulting in cognitive errors between customers and suppliers)
19	Risk of transport disruption
20	Errors and delays in loading
21	Legislative or regulatory reform
22	Government instability
23	Changes or barriers to taxation arise
24	Natural hazard
25	Environmental pollution
26	Stray animals (birds, insects, etc.)
27	Transport route anomalies (accidents, traffic jams) delay delivery times
28	Economic risk, exposure to adverse market fluctuations (inflation, securities, exchange rates, interest rate spreads)
29	Demand fluctuation (unexpected rise or fall in product demand)
30	The threat of corporate competition

Table A2. 34 SCRE indicators.

NO.	Supply chain resilience index
1	Agility
2	Flexibility (ability to adjust to demand and environmental conditions in the short term)
3	Redundancy (resulting in unnecessary waste of resources, such as inventory)
4	Production capacity
5	velocity
6	Efficiency
7	reactivity
8	Authority (ability to quickly make decisions about product changes in the face of market changes)
9	Visibility (Capture data from multiple sources for transparency in the supply chain system)
10	adaptive capacity
11	Predictive ability
12	recovery capability
13	Decentralization (dispersion of assets, decisions, resources, markets)
14	cooperation
15	Market positioning
16	information sharing
17	Safety assurance (ability to protect product safety and respond to crises)
18	Financial capacity
19	Product management (inventory, quality, etc.)
20	Transport disruption risk
21	Knowledge (internal knowledge of customer relationship, management, technology, etc.)
22	Human resource management
23	Culture (Internal culture construction)
24	Technical competence (adaptation to new technologies and integration of products and technologies)
25	Confidence(confidence in partners and strengthening relationships)
26	Sustainability (long-term economic, social and environmental development)
27	Supply chain structure (connections between upstream and downstream of the supply chain)
28	Good customer service
29	Robustness (ability to withstand failures and disturbances)
30	Analytical ability
31	Dynamic capability (the ability to adjust the system to the uncertainty of the environment)
32	Innovation ability (innovation of technology, product, model, etc.)
33	Communication (ability to communicate with consumers)
34	Sensitivity (the ability of the supply chain to obtain information about actual demand from the end market and to react quickly to it)

Table A3. 123 I5.0 drivers.

Dimension	I5.0 Drivers
People-oriented	Pay attention to human needs(Workers’ well-being, basic rights, etc.)
	Motivating workers
	Focuses on the collaboration between human intelligence and cognitive computing
	Consider ethics in society and morality
	Expanding human capabilities and supporting workers to make the right decisions (skills training, etc.)
	Improve the use of personalised human-computer interaction technology
	Create a safe and inclusive work environment
	Focus on employee safety and management, employee learning/training
	Valuing employees’ emotional abilities (EQ, focus, etc.)
	Organizational justice, job satisfaction, trust and innovation
	Respect and protect privacy
	Empower workers
	New technologies need to be easy to understand and transparent
	Pay attention to the different values and needs of society
	The needs of employees must be combined with the needs of society
	Implement retraining and lifelong learning
	Acceptance and trust in technology
	Adapting technology to humans and training people how to use the new technology
Sustainable	Personalized products and services
	Development of innovation ecosystem
	Advanced digital technology reduces product waste and final costs
	Energy conservation and emission reduction
	Developing new materials
	Developing new energy sources (bioeconomy, etc.)
	Responsible consumption and production
	Industrialization, infrastructure innovation
	Sustainable cities and human settlements
	Increasing availability of traditional values
	Social orientation and values that enhance productivity
	Enhance corporate social responsibility
	Focus on customer and value creation
	Focus on the individual needs of customers
	Mass customisation of products and services to customers
	Leveraging digital technologies to deliver on climate commitments, the right combination of digital and green
	Economic activities designed to reduce the overall carbon and material footprint of the supply chain
	Transition from “Energy Efficiency First” to “Resource Efficiency First”
	More emphasis is placed on policy and measurement of the physical/real economy rather than financial indicators
	Technologies that can measure environmental and social value
	Develop business models that make ecological and social values accessible (sustainable, regenerative and circular economy)
	Integrated industry stands for policies that support the implementation of an entire industry or small and medium enterprises
Resilience	Complete accurate and creative tasks with less time and cost
	A production system that is well adapted to external changes ((geopolitical) political changes and natural emergencies)
	Flexible business processes
	Improve production flexibility and increase control redundancy
	Assess production system risk points
	Establish contingency plans
	Improving the industrial chain and supply chain restoration and the investment mechanism for royal insurance
	Improving the level of autonomy and control
	Focusing on resource scarcity, climate change, skills gaps
	New operational management model
	Self-resilience and systemic resilience
	Wearable devices maintain contactless operations during the pandemic
	A complete survey of all value chains
	Radically redesign the value chain
	Changing the way we think about policy, investment and corporate governance
Technology	Internet of Things system development
	Mobile technology
	Non-contact automation technology
	Simulation, horizontal and vertical system integration
	Digital technology
	The application of intelligent technology
	Integration of virtual and physical Spaces
	Using embedded sensor technology
	Connect and combine the strengths of humans and machines
	Biomimetic technology and smart materials allow materials to be embedded with sensors and enhancements while being recyclable
	Based on real-time digital twinning and simulation of the entire system.
	Cybersecurity data transfer, storage, and analysis technologies that handle data and system interoperability
	Artificial intelligence and energy efficiency
	Renewable energy storage and trusted autonomous technologies
	machine learning
	Industrial edge computing (can be thought of as a miniature data centre that processes or stores critical data locally and pushes all incoming data to a central data centre or cloud storage).
	Federated learning (designed to train machine learning algorithms across multiple distributed devices or servers, each with its own local and private data)
	Integration of all types of applications
	Beyond 5G(B5G) network (B5G is a perfect process to solve some application scenarios and technologies, such as in telemedicine, intelligent transportation, industry 4.0 industry application)
	Beyond 5G(B5G) network (B5G is a perfect process to solve some application scenarios and technologies, such as in telemedicine, intelligent transportation, industry 4.0 industry application)
	Bioeconomy technology (Bioeconomy is an economy based on the production, distribution and use of biotechnological products on the basis of sustainable utilisation of biological resources and biotechnology)
	Energy storage and quantum computing (is a new computing model that follows the laws of quantum mechanics to control quantum information units for calculation.)
	Automatic identification Auto-ID and sensor technology (Automatic identification technology is an important method and means to automatically read and enter information data into the computer, which is a comprehensive science and technology based on computer technology and communication technology)
	Nanotechnology (the science and technology of manufacturing substances from single atoms and molecules, studying the properties and applications of materials with structural dimensions ranging from 0.1 to 100 nanometres). Nanoscience and technology is a science and technology based on many modern advanced science and technology, and it is the product of the combination of dynamic science and modern technology.
	Top management support and lead, change leadership style
	Infrastructure (Internet, CPS, cloud computing, etc.)
	Customer involvement, customer design and manufacturing process integration, customer relationship maintenance
	Data privacy, security and availability
	Financial resources
	Business strategy management, strategic alignment between the adoption of new technologies and expectations
	Organizational culture, digital culture
	Using digital technology for new product innovation and intelligence
	New technologies for security, dealing with insecurity, security vulnerabilities
	Compatibility with existing technologies, integration of technology platforms
	Horizontal and vertical integration of value chain
	Existing technical skill levels within the organisation, IT information technology structure
	Competition and pressure from business partners, market competition pressure
	Good supply chain management and collaboration make the organisation’s goals clear and focused
	Availability of collaboration tools
	Organizational structure changes, organisational digitization
	Global engagement, global connections
	Process modularisation or dynamic design of business processes
	Hardware and software connectivity, Internet and machine production convergence
	IT infrastructure Captures, stores, and analyzes data (Big Data Management)
	Team work and its expertise, lean production experience
	Investing in new equipment, invest in I4.0 facilities
	Support from academic researchers
	Government and policy support
	Direct information sharing and communication among supply chain members
	Supply chain digitization
	Development of data and simulation tools
	Improving IT standards and implementing I4.0 regulations
	Lack of global standards and data-sharing agreements
	Product lifecycle management
	Virtual testing and simulation
	Fully integrated enterprise resource planning
	Real-time tracking of inventory, real-time data collection and analysis
	Raw material and production traceability
	Economics of digital investment (profitability unclear)
	legal question
	State economic security
	Cost and expense management
	Companies and institutions cooperate
	Clean Development Mechanism, low waste
	Housing and service infrastructure maintenance
	Scale of company
	Project management
	Centralized management of products, processes and resources

Appendix B

Fuzzy Delphi Method (FDM)

FDM is an indicator identification method that combines fuzzy theory and the Delphi method proposed by Murray et al. The detailed steps are as follows.

Step A: Design the FDM questionnaire. Set the value range to evaluate the importance of the evaluation target. The most conservative value and optimistic value are evaluated. The larger the value is, the more important the index.

Step B: We calculate the most conservative and optimistic values given by all experts and remove extreme values that more than double the standard deviation. According to fuzzy theory, the minimum $C_L^i$, geometric mean $C_M^i$, and maximum values $C_U^i$ of the conservative values are calculated, and the conservative values of triangular fuzzy numbers $C^i=\left(C_L^i{C}_M^i{C}_U^i\right)$ are obtained. The minimum value $O_L^i$, geometric mean $O_M^i$, and the maximum value of the optimistic value $O^i=\left(O_L^i{O}_M^i{O}_U^i\right)$ are calculated, and the optimistic value of triangular fuzzy number is obtained.

Step C: Calculate the degree of consensus $G^i$. The higher the value of $G^i$, the more important the indicator is. The calculation formula is as follows:

(1) ${\mathrm{G}}^{\mathrm{i}}=\frac{\left[\left({\mathrm{C}}_{\mathrm{U}}^{\mathrm{i}}\times {\mathrm{O}}_{\mathrm{M}}^{\mathrm{i}}\right)-\left({\mathrm{O}}_{\mathrm{L}}^{\mathrm{i}}\times {\mathrm{C}}_{\mathrm{M}}^{\mathrm{i}}\right)\right]}{\left[\left({\mathrm{C}}_{\mathrm{U}}^{\mathrm{i}}-{\mathrm{C}}_{\mathrm{M}}^{\mathrm{i}}\right)+\left({\mathrm{O}}_{\mathrm{M}}^{\mathrm{i}}-{\mathrm{O}}_{\mathrm{L}}^{\mathrm{i}}\right)\right]}$(1)

Step D: Check for convergence. $Z^i=C_u^i-O_L^i$ represents the grey area of fuzzy relation, and $M^i=O_M^i-C_M^i$ represents the range of optimistic value and conservative value. When $Z^i\le M^i$, the expert’s opinion converges; otherwise it diverges, and steps A – C need to be repeated until convergence.

Step E: Set a reasonable threshold $\mathit{\alpha }$. While the value of $G^i$＞$\mathit{\alpha }$ is retained, other factors smaller than the threshold are removed.

Fuzzy Interpretation Structure Model (FISM)

FISM introduces the concept of a fuzzy matrix, which can accurately reflect the relationship between complex systems. The main steps are as follows:

Step A: Select the membership function and obtain the fuzzy association strength function B using Equation (2)(2) $\mathrm{B}={\mathrm{b}}_{\mathrm{ij}}={\mathrm{f}}_{\mathrm{ij}}/{\mathrm{f}}_{\mathrm{i}}+{\mathrm{f}}_{\mathrm{j}}\text{---}{\mathrm{f}}_{\mathrm{ij}}$(2) .

(2) $\mathrm{B}={\mathrm{b}}_{\mathrm{ij}}={\mathrm{f}}_{\mathrm{ij}}/{\mathrm{f}}_{\mathrm{i}}+{\mathrm{f}}_{\mathrm{j}}\text{---}{\mathrm{f}}_{\mathrm{ij}}$(2)

Step B: The Boolean incidence matrix. The threshold value is set to screen the correlation strength of matrix B. The smaller the threshold value is, the larger the scope covered by each element and the larger the system decomposition. According to Equation (3)(3) $a_{\mathit{ij}}=\left\{\begin{array}{c}1,a_{\mathit{ij}}\ge \mathit{\lambda }\\ 0,b_{\mathit{ij}}<\mathit{\lambda }\end{array}\right.$(3) , intensity matrix B is transformed into adjacency matrix A.

(3) $a_{\mathit{ij}}=\left\{\begin{array}{c}1,a_{\mathit{ij}}\ge \mathit{\lambda }\\ 0,b_{\mathit{ij}}<\mathit{\lambda }\end{array}\right.$(3)

Analysis Network Procedure (ANP)

ANP verifies the weight relationship between various factors through analysis among different levels and determining the corresponding analysis matrix. The main steps are as follows:

Step A: Design an ANP questionnaire according to the questionnaire results of the previous stage and allow experts to objectively evaluate each factor through the evaluation matrix scale in Table B1. Assume n elements are present on the same level. Judge the importance degree between the factor of $\mathit{i}$ and factor of $\mathit{j}$: that is, get the value $a_{\mathit{ij}}$.

Table B1. ANP assessment scale.

Assessment Scale	Definition	Instruction
1	Equal important	Both are of equal importance
3	A little important	As a rule of thumb, one indicator is slightly more important.
5	Quite important	As a rule of thumb, one indicator matters.
7	Very important	As it turns out, a certain indicator is very important.
9	Absolutely important	There is ample evidence that one metric is absolutely
2, 4, 6, 8	The median of	A compromise option

Step B: Collect the questionnaire and conduct consistency test to judge the validity of the data. First, according to the sum of eigenvalues of the judgement matrix and according to Equation (4)(4) $\mathit{\lambda }{\sum }_{i=1}^n{\lambda }_i=\mathit{n}$(4) , if the sum of eigenvalues is ${\lambda }_{\mathit{max}}$≥n, the absolute value of each component vector of the orthogonalization vector of the matrix can be obtained as the weight value of each element. Second, the importance vector ${\mathrm{W}}_{\mathrm{i}}$ and ${\mathrm{W}}_{\mathrm{i}}^{\mathrm{o}}$ of each row is calculated according to Equations (5)(5) $W_{\mathrm{i}}=\frac{1}{\mathrm{n}}\sum _{\mathrm{j}=1}^{\mathrm{n}}\frac{{\mathrm{a}}_{\mathrm{ij}}}{\sum _{\mathrm{k}=1}^{\mathrm{n}}{\mathrm{a}}_{\mathrm{kj}}}$(5) and (6(6) $W_{\mathrm{i}}^0=\frac{{\mathit{W}}_{\mathrm{i}}}{\sum _{\mathrm{i}}{W}_{\mathrm{i}}}$(6) ).

(4) $\mathit{\lambda }{\sum }_{i=1}^n{\lambda }_i=\mathit{n}$(4)

(5) $W_{\mathrm{i}}=\frac{1}{\mathrm{n}}\sum _{\mathrm{j}=1}^{\mathrm{n}}\frac{{\mathrm{a}}_{\mathrm{ij}}}{\sum _{\mathrm{k}=1}^{\mathrm{n}}{\mathrm{a}}_{\mathrm{kj}}}$(5)

(6) $W_{\mathrm{i}}^0=\frac{{\mathit{W}}_{\mathrm{i}}}{\sum _{\mathrm{i}}{W}_{\mathrm{i}}}$(6)

Step C: Check the consistency of the results. According to Equation (7)(7) $\mathit{C}.\mathit{I}.=\frac{{\mathit{\lambda }}_{\max }-\mathrm{n}}{\mathrm{n}-1}$(7) , the consistency index (CI) can be calculated to determine whether the matrix is consistent. When =n and CI ≤ 0.1, the judgement matrix passes the consistency test.

(7) $\mathit{C}.\mathit{I}.=\frac{{\mathit{\lambda }}_{\max }-\mathrm{n}}{\mathrm{n}-1}$(7)

(8) $\mathit{C}.\mathit{R}.=\frac{\mathit{C}.\mathit{I}.}{\mathit{R}.\mathit{I}.}$(8)

Table B2. Average random consistency index.

n	1	2	3	4	5	6	7	8	9	10	11	12	13	14
R.I.	0	0	0.52	0.89	1.12	1.26	1.36	1.41	1.46	1.49	1.52	1.54	1.56	1.58

When too many factors are evaluated by the judgement matrix, the calculation to be performed will be heavier. To test the consistency of each matrix, random index (RI) is introduced here to assist in testing, as listed in Table B1.

If CI > 0.1, its consistency ratio (CR) should be calculated according to Equation (8)(8) $\mathit{C}.\mathit{R}.=\frac{\mathit{C}.\mathit{I}.}{\mathit{R}.\mathit{I}.}$(8) . When CR ≤ 0.1, the matrix has passed the consistency test: that is, the expert’s evaluation of the factors and the calculation results are valid and reasonable. Otherwise, it fails. The questionnaire that fails the one-time test is deemed invalid and needs to be resent to the expert for filling to obtain as much valid data as possible and ensure it meets the consistency test.

The fourth step is calculating the average value $W_{\mathrm{i}}$ in the judgement matrix that meets the requirements and obtaining the importance degree and ranking of each factor calculated via ANP.

TOPSIS method

The TOPSIS method takes the maximum positive ideal solution and the minimum negative ideal solution as the optimal decision. The specific steps are as follows:

Step A: Construct the initial matrix D, which contains n evaluation objects and m evaluation indicators.

(9) $\mathit{D}={\left[{\mathit{x}}_{\mathit{ij}}\right]}_{\mathit{m}\times \mathit{n}}={\begin{array}{c}\begin{array}{cccc}{C}_1& C_2& \cdots & C_n\end{array}\\ \begin{array}{c}{A}_1\\ A_2\\ \mathit{M}\\ A_m\end{array}\left[\begin{array}{cccc}{x}_{11}& x_{12}& \cdots & x_{1\mathit{n}}\\ x_{21}& x_{22}& \cdots & x_{2\mathit{n}}\\ \cdots & \cdots & \cdots & \cdots \\ x_{\mathit{m}1}& x_{\mathit{m}2}& \cdots & x_{\mathit{mn}}\end{array}\right]\end{array}}_{\mathit{m}\times \mathit{n}}$(9)

Step B: Normalization. To standardise the calculation of each index, the matrix should first be normalised, where is the normalised evaluation value of the first scheme to the second scheme.

(10) $r_{\mathit{ij}}=\frac{{\mathit{x}}_{\mathit{ij}}}{\sqrt{{\sum }_{i=1}^m{x}_{ij}^2}}(\mathit{i}=1,2\dots \mathit{m},\mathit{j}=1,2\dots \mathit{n}$(10)

(11) $\mathit{R}={\left[{\mathit{r}}_{\mathit{ij}}\right]}_{\mathit{m}\times \mathit{n}}={\begin{array}{c}\begin{array}{cccc}{C}_1& C_2& \cdots & C_n\end{array}\\ \begin{array}{c}{A}_1\\ A_2\\ \mathit{M}\\ A_m\end{array}\left[\begin{array}{cccc}{r}_{11}& r_{12}& \cdots & r_{1\mathit{n}}\\ r_{21}& r_{22}& \cdots & r_{2\mathit{n}}\\ \cdots & \cdots & \cdots & \cdots \\ r_{\mathit{m}1}& r_{\mathit{m}2}& \cdots & r_{\mathit{mn}}\end{array}\right]\end{array}}_{\mathit{m}\times \mathit{n}}$(11)

Step C: Construct the weighted standardised matrix $V_{\mathit{ij}}$ according to the index weights $W_i$ obtained via the ANP analysis method, or calculate objective weight based on the entropy weight.

(12) $\mathit{V}={\left[{\mathit{v}}_{\mathit{ij}}\right]}_{\mathit{m}\times \mathit{n}}={\begin{array}{c}\begin{array}{cccc}{C}_1& C_2& \cdots & C_n\end{array}\\ \begin{array}{c}{A}_1\\ A_2\\ \mathit{M}\\ A_m\end{array}\left[\begin{array}{cccc}{W}_1{r}_{11}& W_2{r}_{12}& \cdots & W_i{r}_{1\mathit{n}}\\ W_1{r}_{21}& W_2{r}_{22}& \cdots & W_i{r}_{2\mathit{n}}\\ \cdots & \cdots & \cdots & \cdots \\ W_1{r}_{\mathit{m}1}& W_2{r}_{\mathit{m}2}& \cdots & W_i{r}_{\mathit{mn}}\end{array}\right]\end{array}}_{\mathit{m}\times \mathit{n}}$(12)

(13) $e_j=-\frac{1}{ln\mathit{m}}\sum _{\mathit{i}=1}^m{r}_{\mathit{ij}}ln{r}_{\mathit{ij}}$(13)

(14) $w_j=-\frac{1-e_j}{{\sum }_{j=1}^n\left(1-e_j\right)}$(14)

(15) $\mathit{W}=\left(w_1,w_2,\dots ,w_n\right)$(15)

(16) $v_{\mathit{ij}}=r_{\mathit{ij}}\times w_j$(16)

(17) $\mathit{V}={\left[{\mathit{v}}_{\mathit{ij}}\right]}_{\mathit{m}\times \mathit{n}}$(17)

Step D: Calculate positive ideal solutions $V^{+}$ and negative ideal solutions $V^{-}$.

$V^{+}=\left\{V_1^{+},V_2^{+},\dots ,V_n^{+}\right\}$

(18) $=(\mathit{max}\left\{V_{11}^{\hspace{0.17em}},V_{21}^{\hspace{0.17em}},\dots ,V_{m1}^{\hspace{0.17em}}\right\}\mathit{max}\left\{V_{12}^{\hspace{0.17em}},V_{22}^{\hspace{0.17em}},\dots ,V_{m2}^{\hspace{0.17em}}\right\}\mathit{max}\left\{V_{1n}^{\hspace{0.17em}},V_{2n}^{\hspace{0.17em}},V_{mn}^{\hspace{0.17em}}\right\})$(18)

(19) $=(\mathit{min}\left\{V_{11}^{\hspace{0.17em}},V_{21}^{\hspace{0.17em}},\dots ,V_{m1}^{\hspace{0.17em}}\right\}\left\{V_{12}^{\hspace{0.17em}},V_{22}^{\hspace{0.17em}},\dots ,V_{m2}^{\hspace{0.17em}}\right\}\left\{V_{1n}^{\hspace{0.17em}},V_{2n}^{\hspace{0.17em}},V_{mn}^{\hspace{0.17em}}\right\})$(19)

Step E: Calculate the distance between positive and negative ideal solutions. The attribute of the index can be divided into the benefit type and achievement type. The benefit type $S_i^{+}$ should be as large as possible, whereas the cost type $S_i^{-}$ should be as small as possible. $S_i^{\hspace{0.17em}}$is used to represent the Euclidean distance, $C_i$ represents the relative proximity, and the value range of $C_i$ is (0,1). A large value implies that the decision is closer to the ideal solution, and the decision is more optimal if it is further away from the negative ideal solution.

(20) $S_i^{+}=\sqrt{\sum _{\mathit{j}=1}^n{\left({\mathit{V}}_{\mathit{ij}}-V_j^{+}\right)}^2}$(20)

(21) $S_i^{-}=\sqrt{\sum _{\mathit{j}=1}^n{\left({\mathit{V}}_{\mathit{ij}}-V_j^{-}\right)}^2}$(21)

(22) $C_i=\frac{S_i^{-}}{S_i^{+}+S_i^{-}}$(22)