Research on Critical Factors Influencing Organizational Resilience of Major Transportation Infrastructure Projects: A Hybrid Fuzzy DEMATEL-ISM-MICMAC Approach

Abstract

To strengthen major transportation infrastructure projects’ (MTIPs’) organizational resilience and fortify their capacity for crisis management and project risk prevention. In this paper, based on the resilience theory development process, the connotation of organizational resilience of MTIPs is defined, and 20 influencing factors of organizational resilience of MTIPs are extracted from four categories of stability, redundancy, adaptability, and rapidity according to the literature analysis and case study method. The significance, causality, and multilevel recursive order structure of the influencing factors were investigated by the fuzzy DEMATEL-ISM approach, and their driving and dependent characteristics were analyzed through MICMAC. The results indicate that risk warning and prediction, human resources management, inter-organizational synergies, resource reserve situations, organizational leadership, and organizational learning are the crucial factors of organizational resilience in MTIPs. There are three levels and five ranks in the multilevel recursive rank structure of the factors affecting MTIPs’ organizational resilience. Among them, risk warning and prediction, equipment condition and performance, human resources management, and organizational leadership have the deepest impact on organizational resilience in MTIPs. The findings can clarify ideas for subsequent research on organizational resilience in this area and inform project decision-makers in developing strategies for optimizing organizational resilience.

1. Introduction

With the state’s encouragement and support, China’s transportation infrastructure construction has expanded quickly in recent years. Numerous significant transportation infrastructure projects, including the Qinghai–Tibet Railway, the Sichuan–Tibet Railway, and the Hong Kong–Zhuhai–Macao Bridge, have been completed [1]. MTIPs usually refer to large-scale infrastructure led by the government, featuring a huge project scale, complex engineering in the natural environment, a long project life cycle, and multi-body participation, which is a category of fundamental projects with far-reaching impacts on social and economic development [2]. MTIPs are mainly related to railway projects, metro projects, other urban rail transit projects, highway projects, nationally and provincially established trunk highway and water transport projects, civil airport projects, and comprehensive transport hub projects. Most of these projects are linear projects or large-scale hub projects, which are characterized by huge investment scales and long construction cycles. The Twentieth Congress Report has clearly proposed “accelerating the construction of a strong transportation nation” and “building resilient cities”. Over the past decade, China’s transport fixed asset investment has continued to grow, with total investment increasing from about 2.05 trillion yuan in 2011 to 3.9 trillion yuan in 2023. Large-scale investment in transport fixed assets has strongly supported the rapid improvement of the transport infrastructure network, laying a solid foundation for socio-economic development and the improvement of people’s livelihoods. Major transportation infrastructure is an important part of the urban infrastructure system, and vigorously promoting its construction can greatly drive regional economic development.

Project organization, as the core of MTIP management, is of great significance to the achievement of quality, safety, schedule, and other objectives of the project [3]. It can not only provide professional technical support and coordination management but also promote the sustainable development of the project. In conditions of turbulent alterations in the global natural world, transportation infrastructure projects are all exposed to the environment and tend to be at high risk of disasters [4]. The following common problems currently exist with MTIPs. First, the lack of predictive capacity to confront severe weather and catastrophes leads to the inability of the project to respond quickly when it encounters emergencies [5,6]. Second, transportation infrastructure projects are constructed without adequate consideration of the need for reliability and redundancy, resulting in projects facing paralysis in the event of disruptions and crises. Thirdly, the resilience of transportation infrastructure projects is insufficiently constructed for natural disasters and other emergencies, and the ability to recover is weak. Fourth, the organization’s coordination and communication are poor, resulting in untimely information transfer and inefficient decision making, thus affecting the overall project resilience [7]. In response to the above issues, MTIPs need to be equipped with the critical skills of coping and adapting to external changes, that is, the process capabilities within the organization of absorbing, adapting, and responding in the face of crises and risks, and being able to ensure that the project organization recovers in an orderly manner to its normal operating state after a shock. These capabilities are referred to as organizational resilience for MTIPs. MTIPs are often highly complex and risky, requiring cross-regional and cross-sectoral coordination, as well as unforeseen challenges and changes. In this context, there is an urgent need to improve project organizational resilience. Organizational resilience can help project teams cope with adverse events, improve project stability, foster innovation and flexibility, and enhance team cohesion and collaboration. Identifying the factors influencing organizational resilience in MTIPs can help project managers develop organizational resilience, which in turn can enhance the success of MTIPs.

The enhancement of organizational resilience in major transport infrastructure projects occupies an important position in the successful implementation and stable development of the projects. However, the current research on organizational resilience in major transport infrastructure projects is still unclear. Therefore, based on the harsh environment in which the construction industry has developed in recent years and the deficiencies exposed by engineering project organizations in the face of crises, an in-depth analysis of organizational resilience in engineering projects is conducive to enhancing the ability of engineering project organizations and inter-organizational systems to withstand external risks. This research aims to:

(1)

Investigate and figure out the relevant factors affecting the organizational resilience construction of MTIPs.

(2)

Adopt a hybrid fuzzy DEMATEL-ISM-MICMAC methodology to analyze the importance level of influencing factors, and to study in depth the hierarchical relationship, the path of action, and the driving dependency attributes among the influencing factors.

This paper’s research is as follows. A review of the literature and an analysis of the connotations and influencing factors of organizational resilience are presented in Section 2. Section 3 first puts forward the definition of organizational resilience of major transportation infrastructure projects and identifies the indicators of organizational resilience affecting factors of MTIPs from the four characteristic elements of organizational resilience, namely stability, redundancy, rapidity, and adaptability, by combining the approaches of literature analysis and case study. The research methodology of this paper is also presented. Section 4 identifies the critical influencing factors affecting the organizational resilience of MTIPs based on fuzzy DEMATEL. A multilayer recursive structural model of organizational resilience was created using the ISM, and it was found that the organizational resilience indicator system for MTIPs is a multilevel recursive system with 3 levels and 5 ranks. Finally, a MICMAC analysis was conducted to quadrant the attributes of organizational resilience-influencing factors for MTIPs. Section 5 provides an in-depth discussion of the findings of the study. Section 6 summarizes the main research conclusions, highlighting the research implications of the article, its current limitations, and future research directions. This study further enriches and develops the theoretical investigation of the organizational resilience of MTIPs and also provides references to risk prevention and crisis response during the construction period of the projects.

2. Literature Review

2.1. Connotations of Organizational Resilience

The concept of “resilience” is derived from the Latin word “Resilere”, which means “to rejuvenate”, while resilience is also a physical concept, indicating the capacity of a material to absorb energy during deformation or fracture. The notion of resilience was first introduced into modern science by Holling [8], an ecologist, to evaluate the capability of ecosystems to absorb interferences and repair them quickly. Since the proposed resilience theory, it has undergone the first cognitive shift from the physical to the ecological field, and its connotation has been gradually refined and extended. In the 1980s, resilience was applied to the field of child psychology [9]. Meyer [10] was the first to introduce resilience into the management field, creating a new chapter in academic research on organizational resilience [11,12,13]. The notion of resilience has since been extended to various domains in social-ecological systems, such as urban resilience [14], disaster resilience [15], and economic resilience [16]. Resilience theory has experienced a secondary development from ecology to a multidisciplinary field. Figure 1 illustrates the development of resilience theory.

It is widely recognized by domestic and international academics that organizational resilience is the capacity of a project organization to recover and adapt when facing uncertain situations [17]. According to the research, the nature of organizational resilience is the capacity to maintain its original function and re-establish its strategy through active mobilization and resource allocation in the face of crisis and turmoil [18]. Resilient organizations thrive through unexpected events and uncertain conditions. There are two viewpoints on the conceptual study of organizational resilience, as shown in Figure 2. According to the first perspective, organizational resilience is the capacity of an organization to swiftly adapt to adversity or a crisis, act promptly, and return to its pre-adversity state. This idea is familiar to the idea of plasticity in the natural sciences, where an elastic object can recover its original state and properties after bending or folding. The research of Lengnick-Hall et al. [19], organizational resilience is a combination of situational, behavioral, and cognitive traits that can enable enterprises to cope with significant shocks. The authors also suggest that organizations can strengthen their organizational resilience by strategically managing their human resources to advance the competencies of their core workforce. Aguila and ElMaraghy [20] consider organizational resilience as the supply chain’s capacity to resume normal operations after a disruption and propose an assessment framework to quantify supply chain resilience and network topology. Resilience in organizations has been defined by Garg et al. [21] as an organization’s capacity to handle a range of difficulties and disasters. Organizational resilience, according to Sajko et al. [22], is the power of an organization to recognize, prevent, and respond to shocks from the outside world. Xu et al. [23] elaborated organizational resilience as the resilience, buffer, and adaptability against external shocks formed by organizations relying on the linkage dependencies within the system.

The second view is the capability standpoint, which argues that organizational resilience has the ability to transcend the initial state. That is to say that organizational resilience can open up fresh possibilities for the organization to grow and reach an optimal system state after the adversity or crisis is over. Bothello and Salles-Djelic [24] define resilience as a competence in a social-ecological system that not only carries shocks but also learns and adapts over time. From the perspective of organizational adaptation theory, Li and Huang [25] found that organizational resilience helps firms adapt to uncertain changes in the outside world and toward sustainable growth. The findings enrich the research perspectives, ideas, and methods of organizational resilience and provide a theoretical basis for entrepreneurial enterprises to practice the adaptive concept of organizational resilience cultivation. In the words of Tang et al. [26], organizational resilience has a major benefit in improving enterprise vitality and flexibility as a type of enterprise dynamic capability. This study argues that project organizations’ ability to recover or bounce back from a crisis or disruptive event is the foundational capability of organizational resilience that can help organizations survive adversity. And it is the ability to bounce back that organizational resilience has that is key to the development and survival of project organizations in VUCA (volatile, uncertain, complex, ambiguous) situations.

As can be seen through the combing of organizational resilience literature, the majority of current research on organizational resilience centers on the fields of business, enterprise, and manufacturing, with relatively little research on organizational resilience in the engineering field, particularly on organizational resilience in MTIPs. With the goal of improving the current theory of organizational resilience and expanding the scope of the study environment for organizational resilience, this work focuses on MTIPs. At the same time, it can help MTIPs cope with sudden crises or accumulated pressure, which is crucial for facilitating the consistent development of the projects.

2.2. Organizational Resilience Influencing Factors

In recent decades, there has been a notable growth in the academic community’s interest in organizational resilience, and a considerable body of study has emerged on the factors that affect organizational resilience [27]. Multiple variables influence organizational resilience; therefore, it is possible to discover and analyze the main nodes and influencing factors of organizational resilience through the research of organizational resilience factors. Additionally, the path to organizational resilience may be established and optimized. Nevertheless, there is not much study on the factors affecting organizational resilience from the standpoint of engineering projects. The majority of organizational resilience research has been conducted at the enterprise level and has generally looked at the mechanisms underlying the impact of a single factor on organizational resilience.

In relation to relevant studies, developing resilient human resource practices inside an organization is the most effective way to successfully execute technological change. Human resource management is widely acknowledged as a key component in attaining organizational resilience [28,29]. The concept of organizational resilience is multifaceted, multidimensional, and multi-connotated. Studies have indicated that a high degree of diversity within an organization can positively influence organizational resilience [30]. Organizational leadership also positively impacts business organizational resilience. On the one hand, resilient leaders are adept at quickly recovering and resilience bouncing back to normalcy under stress, from which they can improve themselves and lead their followers [31], and on the other hand, the decision-making ability of organizational leaders in times of crises is crucial [32]. Studies have also found that the level of implementation of relevant key technologies positively affects firms’ organizational resilience and perceived performance [33]. Furthermore, the ability to innovate continuously and the efficient use of organizational resources both significantly enhance organizational resilience. Based on the analytical framework of “resource-capability” theory, organizational resilience can help enterprises cope with crises and take advantage of crises to achieve counter-trend growth by means of restructuring organizational resources and processes [34]. Some scholars also delineate the influencing factor dimensions of organizational resilience from the standpoint of organizational resilience evaluation. In order to foster a positive feedback loop and mutual advancement between theoretical research and the real-world application of organizational resilience, Zhang et al. [35] built an “I-P-O” model that covered the situational picture and factors that affect organizational resilience. They did this by starting from the perspectives of competency, process, trait, and results of organizational resilience. Wang and Cai [36] established an organizational resilience scale and conducted exploratory factor assessment, validation factor analysis, and credibility tests on the sample data, respectively. The results clarified that organizational resilience includes four dimensional elements: Resilience, planning ability, situational awareness, and resilience commitment. Based on organizational information processing theory, Ma et al. [37] found that network market orientation exerts a progressive influence on the organizational resilience of start-ups.

In conclusion, it is found that the current domestic and international research related to organizational resilience is mainly concentrated on the general field of enterprise management, with relatively limited resilience research from an organizational perspective in the project management field. The engineering project organization system is different from the enterprise organization. In order to achieve the set goal and establish a temporary organization, engineering project organization resilience is often affected by the project’s external natural environment and its own management capabilities. This paper constructs a system of factors influencing organizational resilience in MTIPs through a literature analysis method and a case study method to provide references for the improvement of organizational resilience in the engineering field.

2.3. Research Methodology for Influencing Factors

Organizational resilience of MTIPs has complex characteristics and involves multiple stakeholders, and the mechanism of organizational resilience influencing factors is often complicated [38]. Consequently, to effectively recognize the key variables affecting organizational resilience in MTIPs and to define the interaction paths between the various variables, it is necessary to seek methods that can address the identification of the key variables and the clarity of the interrelationships.

In the research method of resilience-influencing factors, scholars have carried out a large number of studies. Among them, the DEMATEL-ISM method has been widely used. Liu et al. [28] analyzed the intrinsic logical relationships and hierarchical structure of factors influencing the organizational resilience of firms based on ISM. Zhao et al. [39] used DEMATEL-ISM to study the institutional resilience impact mechanism of large railway construction projects. Wang and Yu [40] investigated the importance, degree, and role mechanism of factors influencing the toughness of the high-tech ship industry chain based on fuzzy DEMATEL-ISM. Sujan et al. [41] analyzed the drivers of supply chain resilience during the COVID-19 pandemic using an integrated fuzzy DEMATEL-ISM approach. However, the DEMATEL method also has certain limitations, and there are defects brought about by the uncertainty of the environment and the fuzzy semantic representation. Fuzzy DEMATEL integrates triangular fuzzy numbers into traditional DEMATEL, converting expert semantics into corresponding triangular fuzzy numbers, which can solve the above defects and reduce expert subjectivity. The use of the fuzzy DEMATEL approach to analyze the factors affecting the organizational resilience of MTIPs, which can sort out the interrelationships between the various elements and determine the importance of the factors in a scientific and rational way. ISM discovers the primary factors and their intrinsic connections by constructing a multilevel recursive directed topology graph and using directed graphs and matrices. Therefore, combining DEMATEL with ISM can both clarify the critical influencing factors in the system and elucidate the logical relationships and hierarchies among the factors. In addition, MICMAC analyses can better reflect the interdependencies and driving forces between factors [42]. Therefore, mixing the above methods can help analyze the research questions more accurately. The hybrid method has now been applied in various studies. Feng et al. [43] used fuzzy DEMATEL-ISM-MICMAC to analyze the factors affecting the green behavior of employees. Alshahrani et al. [44] conducted a study on the factors affecting the implementation of AI-enabled sustainable cloud systems in the IT industry based on this hybrid approach. Shanker and Brave [45], on the other hand, used this hybrid approach to study the factors affecting the sustainability of the diamond supply chain.

As MTIPs belong to a complex system involving a large number of influencing factors and cumbersome organizational relationships, the influencing factors of project organizational resilience are not a purely hierarchical structure, and different factors in different dimensions can influence or impede the development of other factors. Therefore, this hybrid approach has significant advantages in dealing with the factors affecting the organizational resilience of MTIPs, which can reveal the causal relationship between the factors, structured analysis, classification of priorities, and formulation of strategies to analyze the current obstacles to the organizational resilience enhancement of MTIPs, and then provide a more scientific and effective strategic basis for project managers.

Based on the existing research results, this paper focuses on the organizational perspective of MTIPs, divides the organizational resilience system of engineering projects into four subsystems, namely, organization, management, material technology, and environment, and analyzes the affecting factors of organizational resilience of MTIPs in terms of resilience stability, redundancy, adaptability, and rapidity, which is a research breakthrough of this paper. Taking the Nanchang Metro Line 4 project as the survey object, a series of factors affecting the project’s organizational resilience are identified and screened out in combination with the literature analysis method, and the influencing factors are systematically summarized and integrated. In this paper, with reference to the methods of analyzing influencing factors in existing studies, we integrate the hybrid fuzzy DEMATEL-ISM-MICMAC method to build a system of impact factors for the organizational resilience of MTIPs. Recognize the critical factors of organizational resilience of MTIPs based on fuzzy DEMATEL. ISM-MICMAC is used to study the hierarchical structure and factor attribute characteristics among the influencing factors within the project organization system. This will provide ideas for improving the organizational resilience of MTIPs.

3. Materials and Methods

3.1. Organizational Resilience Framework for MTIPs

Currently, the academic community has not provided a clear definition of organizational resilience in the field of MTIPs.

Table 1. Comparative overview of resilience perspectives.

Resilience	Definition of Resilience	Organizational Resilience Applied to MTIPs
Engineering resilience	The ability to restore a system’s performance level from an interrupted state to an operational state, emphasizing absorptive capacity.	Emphasis on the absorptive capacity of the project organization in the face of crisis or disruption, without consideration of the capacity to optimize learning after the crisis event is completed.
Ecological resilience	The ability of ecosystems to return to their original state after experiencing external disturbances, emphasizing buffering capacity [8].	Emphasis on whether the project organization can be restored to its original state after a disruption, without consideration of the project organization’s capacity to absorb, cope and optimize.
Evolutionary resilience	The whole process of the dynamic response of the system, and more emphasis on the system’s ability to adapt, change and learn throughout the perturbation period [12].	Emphasis on the capacity of the project organization to absorb, cope and optimize in the face of crisis or disruption, as a dynamic evolutionary process.

provides a comparative summary of different views of resilience. Based on the organizational resilience theory and combined with the characteristics of MTIPs, this study defines the organizational resilience of MTIPs as the capacity of the project organizational system to remain vigilant, quickly assimilate and adjust to the impacts of emergencies on the project and the project organization in the face of adversity and emergencies, maintain the smooth operation of the system functions, and promptly return to the normal operating state after the emergencies are over and take corresponding measures to adapt to optimize the impacts brought about by the emergencies.

Resilience theory emphasizes the whole process of the organizational system to manage emergencies or risks before, during, and after the event, and through the process, the resilience performance of the organizational system is restored to a new state of safety. The four characteristic elements of resilience are stability, redundancy, adaptability, and rapidity. The organizational resilience of MTIPs emphasizes the systematic construction of the absorptive, restorative, and adaptive capacities of the project organizational system to risks and potential threats from an endogenous perspective, and the organizational resilience framework of MTIPs is constructed with the organizational system, the management system, the material and technological system, and the environmental system as the systematic elements, as shown in Figure 3. The organizational resilience capability of MTIPs considers the stability, redundancy, adaptability, and rapidity of resilience management.

3.2. Identification of Influencing Factors

MTIPs generally have the characteristics of a long engineering construction cycle, a large amount of investment, high engineering risks, and engineering management difficulties [46], which will all have a certain impact on the identification and confirmation of factors affecting their organizational resilience. This study uses the literature research approach to further find the elements affecting the organizational resilience of MTIPs, ensuring the correctness of the indicator system identification. Using the keywords “major transport infrastructure projects”, “organizational resilience”, “enterprise resilience”, and “project resilience” to search Chinese and global databases, including China Knowledge, Web of Science, Google Scholar, along with others, the data retrieval timeframe was set to 2010–2024. The preliminary search yielded 782 relevant English and 631 Chinese documents, and a total of 58 valid documents were obtained after screening the titles, abstracts, keywords, and research areas. There are fewer studies on the organizational resilience of MTIPs, and in this paper, we refer to the relevant studies on the organizational resilience of enterprises and the resilience of the engineering field [47,48,49,50,51,52,53,54] to select the resilience impact factors.

To ensure the reliability of the established indicator system of organizational resilience influencing factors, this study uses the case study methodology to rectify the initially detected influencing factors. This paper takes the Nanchang Metro Line 4 project as a case study and determines the project organizational resilience influencing factors by interviewing and investigating the project managers and construction personnel. The personnel structure of the project interviewees is shown in

Table 2. Information of project interviewees.

Classification	Distribution	Frequency	Percentage (%)
Project sector	Project construction department	5	22
	Project technical department	4	17
	Project contract department	3	13
	Project quality department	2	9
	Project design department	2	9
	Project administration	4	17
	On-site construction personnel	3	13
Education level	Bachelor’s degree	15	65
	Master’s degree	6	26
	Doctor’s degree	2	9
Working years	4–7 years	7	31
	8–10 years	12	52
	10–15 years	4	17

. For the project organizational resilience influencing factors initially selected by the literature analysis method, an interview questionnaire was created to assess the degree of influence of the initially selected factors. By summarizing the results of the on-site interviews and research, the influencing factors were organized and refined, and 20 influencing factors were finally identified. The influencing factors and their connotations are shown in

Table 3. Influencing factors and connotation of organizational resilience of MTIPs.

Dimension	Influencing Factor	Implications of Influencing Factors
Stability	S1 Risk warning and prediction	Identify and assess possible risks in advance to help the organization prepare for risks in advance.
	S2 Equipment condition and performance	The current operation of the equipment, including normal operation, failure, maintenance and other states.
	S3 Human resources management	Incorporate people who possess a high degree of specialized knowledge and abilities, such as training and experience.
	S4 Project environment	Combination of external and internal factors affecting project implementation.
	S5 Natural environment	Natural disasters, force majeure and other emergencies affecting the normal operation of the project.
	S6 Organizational culture	The unique cultural identity of the organization includes its values, beliefs, rituals, symbols, and methods of operation.
	S7 Organizational structure	Sequencing, spatial location, contacts and interrelationships between departments and positions within the organization.
Redundancy	S8 Emergency material security	Provision of the necessary material security for the response to and disposal of emergencies, including the procurement, storage, deployment and utilization of materials.
	S9 Inter-organizational synergies	Capacity of group members to cooperate and communicate well in order to bring together resources and enhance one another’s abilities.
	S10 Resource reserve situation	Resource accumulation and reserves in advance in response to possible future needs or risks.
	S11 Completeness of emergency plans	In the process of formulating and implementing the contingency plan, whether the various possible situations and risks have been fully considered, and whether corresponding countermeasures have been formulated.
	S12 Safety education efforts	The strength and effectiveness of the organization’s investment in safety education, including the number of safety trainings, their coverage, and the quality of their content, etc.
Rapidity	S13 Organizational leadership	Tasked with building trusting relationships, managing conflict, gathering and integrating knowledge, and garnering broad-based support.
	S14 Efficiency of information delivery	The speed and accuracy with which information is transmitted within or between organizations.
	S15 Decision making and responsiveness	The speed with which the organization makes decisions and takes action in the face of unexpected events or emergencies.
	S16 Organizational coordination	Actions taken by the organization to mobilize its internal and external resources in order to withstand crises, boost productivity, and accomplish its objectives.
Adaptability	S17 Accident cause investigation capacity	Ability of the organization to conduct in-depth investigations and analysis of the causes of accidents after they have occurred
	S18 Equipment replenishment and repair	Capacity to promptly restock and fix equipment in case of malfunction or damage.
	S19 Organizational learning	The process by which an organization continually adapts and optimizes its structure and behavior through the acquisition of new knowledge, skills and experience in a changing environment.
	S20 Organizational change	Adjustments and changes in the organization’s structure, culture, strategy, etc., in order to adapt to changes in the external environment or the development of internal needs.

. Based on the perspective of engineering projects, this paper divides the organizational system of major transportation infrastructure projects into organizational system, management system, environmental system, and material and technical system, and constructs the index system of project organizational resilience influencing factors from the four characteristic elements of resilience, namely, stability, redundancy, adaptability, and rapidity (Figure 4).

3.3. Methodology

The DEMATEL method was first proposed by the American scholars Fontela and Gabus [55]. The approach is a multi-criteria decision-making technique that can take into account both the direct and indirect relationships of effect between the two elements. The main purpose is to quantify the causal relationship between factors by determining the influence degree and the affected degree of each factor through the use of tools such as matrices and charts to determine the influencing relationship between the factors [56]. Fuzzy set theory, by simulating the process of the human brain processing fuzzy information, can effectively solve the problem of fuzzy correlation between elements. This work integrates triangular fuzzy numbers into the conventional DEMATEL approach by combining fuzzy set theory (FST) with DEMATEL. Expert judgment is quantified, and expert scoring becomes less subjective by fuzzifying the direct influence matrix and translating the expert semantics into the equivalent triangular fuzzy integers. The CFCS (converting fuzzy numbers into crisp scores) defuzzification approach, which was presented by Opricovic and Tzeng [57], then transforms the fuzzy numbers into exact values.

ISM was proposed by Warfield [58] as an emerging modeling tool that is widely used in systems engineering analysis. This tool is assisted by computer technology, based on directed graphs and matrix support, to transform and decompose a complex system into a manageable number of subsystems and analyze the relationship between the drivers. ISM is based on qualitative analysis by sorting out the logical relationships between factors, constructing multi-level recursive directed topology maps, and using directed maps and matrices to discover the main factors and their intrinsic connections. MICMAC is a quantitative method that uses the principle of matrix multiplication to reflect the interaction relationship between factors, and its core idea is to calculate the sum of rows and columns of the reachability matrix so as to classify the indicator factors into four categories of autonomous, dependent, associated, and independent elements, and to clarify the attribute characteristics of the factors at different levels.

Thus, this paper analyzes the impact factors of organizational resilience of major transportation infrastructure projects based on hybrid fuzzy DEMATEL-ISM-MICMAC and reduces the subjectivity of expert evaluation according to fuzzy DEMATEL to obtain the comprehensive impact matrix of organizational resilience of the project. Then the ISM is applied to obtain the reachability matrix, to divide the hierarchical relationship between organizational resilience factors, to analyze the direct and fundamental influences on the project’s organizational resilience, and to clarify the hierarchical structure and role relationship of organizational resilience influences. Finally, the MICMAC analysis model is established to generate the driving dependency diagram of project organizational resilience affecting factors, to deeply analyze the attribute characteristics of project organizational resilience affecting factors, and to offer both theoretical and practical guidance for the research on the organizational resilience of MTIPs.

3.4. Organization Resilience Influencing Factor Modeling

The specific steps are shown in Figure 5.

Step 1: According to the system of factors impacting the MTIPs’ organizational resilience, the interaction relationship between sets of influencing factors $X=\{x_i,i=1,2,\cdots n\}$ is determined through methods such as questionnaire surveys and expert interviews. Thus, obtaining the direct impact matrix $C={\left[c_{ij}\right]}_{n\times n}$, where $c_{ij}$ indicates how much the ith component directly influences the jth factor, when $i=j,c_{ij}=0$.

Step 2: The semantic scale for expert assessment was constructed. The degree of influence of each influence factor was categorized into five levels, as shown in

Table 4. Semantic transformation table.

Semantic Variable	Score	Corresponding Triangular Fuzzy Number
No effect	0	$\langle 0,0,0.25\rangle$
Very low effect	1	$\langle 0,0.25,0.5\rangle$
Low effect	2	$\langle 0.25,0.5,0.75\rangle$
High effect	3	$\langle 0.5,0.75,1\rangle$
Very high effect	4	$\langle 0.75,1,1\rangle$

. The initial direct influence matrix is transformed into the corresponding triangular fuzzy numbers based on the linguistic variables used by the expert group set by Wang and Chang [59]. The triangular fuzzy number can be denoted as $X=\left(l,m,r\right)$, where $l$ is the left-hand side value, that is, the conservative value; $m$ is the middle value, namely, the closest to the actual value; $r$ is the right-hand-side value, that is, the optimistic value, and $l\le m\le r$.

Step 3: Defuzzification based on CFCS (converting the fuzzy data into crips scores) method is performed to obtain the nth order direct influence matrix $Z$. It includes the following links:

(1)

Normalize the triangular fuzzy numbers.

$$\begin{array}{l}\chi l_{ij}^k=\frac{l_{ij}^k-\mathit{min}{l}_{ij}^k}{{\Delta }_{\mathit{min}}^{\mathit{max}}}\\ \chi m_{ij}^k=\frac{m_{ij}^k-\mathit{min}{m}_{ij}^k}{{\Delta }_{\mathit{min}}^{\mathit{max}}}\\ \chi r_{ij}^k=\frac{r_{ij}^k-\mathit{min}{r}_{ij}^k}{{\Delta }_{\mathit{min}}^{\mathit{max}}}\end{array}$$

(1)

where $\chi l_{ij}^k$, $\chi m_{ij}^k$, and $\chi r_{ij}^k$ denote the normalized values of the triangular fuzzy numbers for the left-hand side value $l_{ij}^k$, the middle value $m_{ij}^k$, and the right-hand-side value $r_{ij}^k$, respectively. ${\Delta }_{\mathit{min}}^{\mathit{max}}=\mathit{max}{r}_{ij}^k-\mathit{min}{l}_{ij}^k$, denotes the value of the gap between the right-hand side value and the left-hand-side value.

(2)

Normalization of left-hand-side and right-hand-side values

$$\begin{array}{l}{u}_{ij}^k=\frac{\chi m_{ij}^k}{1+\chi m_{ij}^k-\chi l_{ij}^k}\\ v_{ij}^k=\frac{\chi r_{ij}^k}{1+\chi r_{ij}^k-\chi m_{ij}^k}\end{array}$$

(2)

where $u_{ij}^k$, $v_{ij}^k$ are the normalized values of the left-hand-side and right-hand-side values, respectively.

(3)

Calculation of crips values

$$z_{\mathit{ij}}^k={\mathit{minc}}_{\mathit{ij}}^k+{\Delta }_{\mathit{min}}^{\mathit{max}}\left[{\mathit{minu}}_{\mathit{ij}}^k\left(1-u_{\mathit{ij}}^k\right)+{\left(v_{\mathit{ij}}^k\right)}^2\right]/\left[1-u_{\mathit{ij}}^k+v_{\mathit{ij}}^k\right]$$

(3)

(4)

Calculate the average of the crips values to obtain the direct impact matrix $Z={\left|z_{ij}\right|}_{n\times n}$.

$$z_{\mathit{ij}}=\left(z_{\mathit{ij}}^1+z_{\mathit{ij}}^2\cdots +z_{\mathit{ij}}^k\right)/k$$

(4)

Step 4: Normalizing the direct impact matrix. In order to eliminate the influence caused by the difference in magnitude, the direct influence matrix $Z$ is normalized according to Equation (5), and finally the standardized direct influence matrix $G$ is obtained.

$$G=\frac{1}{\underset{1\le i\le n}{max}{\displaystyle {\sum }_{j=1}^n{z}_{ij}}}$$

(5)

where $G$ is the normalized direct influence matrix. $\underset{1\le i\le n}{max}{\displaystyle {\sum }_{j=1}^n{z}_{ij}}$ is the rows and maximum values in the direct influence matrix, normalized to have $0<z_{ij}<1$.

Step 5: Calculate the integrated impact matrix. Based on the normalization of the direct influence matrix $G$, in order to further express the influence relationship and the degree of impact between the organizational resilience influences, the combined influence matrix $T$ is derived through Equation (6).

$$T=G+G^2+\cdots +G^n=G\frac{I-G^{n-1}}{I-G}$$

(6)

where $T$ is the integrated impact matrix and $I$ represents the unit matrix of the same order as the normalized direct impact matrix $G$.

Step 6: Calculate the influence degree, affected degree, centrality degree, and cause degree. There are four main and relatively important analytical indicators involved in DEMATEL, namely: The influence degree $D_i$, which indicates the extent to which a particular factor exerts an influence on other factors. Affected degree $C_i$, which reflects the degree to which the element is influenced by other elements. Adding the influence degree to the affected degree to obtain the centrality degree $N_i$, which indicates the importance of the element in the system. The factor’s influence on the project’s organizational resilience increases with its degree of centrality. Subtracting the influence degree from the affected degree gives the cause degree $R_i$, which reflects the logical relationship between the factors. The correlation between the elements is higher with a greater degree of cause. The notion of “cause factor” refers to a factor that has a higher influence on other variables when the cause degree is positive; conversely, the expression “effect factor” refers to a factor that is more influenced by other factors.

According to the integrated impact matrix T, the influence degree, affected degree, centrality degree, and cause degree of the organizational resilience influencing elements of major transportation infrastructure projects are calculated, respectively, as shown in the following formulas.

$$D_i={\displaystyle {\sum }_{j=1}^n{t}_{ij}},(i=1,2,\cdots n)$$

(7)

$$C_i={\displaystyle {\sum }_{j=1}^n{t}_{ji}},(i=1,2,\cdots n)$$

(8)

$$N_i=D_i+C_i,(i=1,2,\cdots n)$$

(9)

$$R_i=D_i-C_i,(i=1,2,\cdots n)$$

(10)

where ${\displaystyle {\sum }_{j=1}^n{t}_{ij}}$ denotes the row-by-row sum and ${\displaystyle {\sum }_{j=1}^n{t}_{ji}}$ denotes the column-sum of the integrated impact matrix $T$.

Step 7: Create the adjacency matrix and reachability matrix. By utilizing ISM to further consider the effects of interactions and coupling between influencing factors, it is possible to show the pathways and hierarchical structure of influences between complex factors. The adjacency matrix $Q$ is built on the basis of the integrated influence matrix $T$. In this process, a threshold value $\lambda$ needs to be introduced in order to streamline the structure of the system, and the larger the $\lambda$, the more pronounced the effect of structural simplification. Simplifying the elements in the integrated influence matrix $T$ based on the value $\lambda$ yields the adjacency matrix $Q$.

$$\lambda =\overline{t_{ij}}+\sigma$$

(11)

$$Q_{ij}=\{\begin{array}{l}1,t_{ij}\ge \lambda \left(i,j=1,2,\cdots n\right)\\ 0,t_{ij}\le \lambda \left(i,j=1,2,\cdots n\right)\end{array}$$

(12)

where $\overline{t_{ij}}$ is the mean of all values in the matrix $T$. $\sigma$ is the standard deviation of the matrix $T$. $t_{ij}$ is the element in the overall influence matrix $T$.

In order to convert to the reachable matrix required by the ISM method, the adjacency matrix $Q$ is added to the unit matrix to obtain the multiplication matrix $B$. Based on the arithmetic properties of a boolean matrix, MATLAB R2022b software is used to carry out a number of boolean operations on the matrix $B$, which results in the reachable matrix $K$.

$$K={\left(Q+I\right)}^{n+1}={\left(Q+I\right)}^n\ne {\left(Q+I\right)}^{n-1}\ne Q+I$$

(13)

where $I$ represents the unit matrix of the same order as the adjacency matrix $Q$.

Step 8: Hierarchy of impact factors. The construction of a multilayer recursive structural model requires the computation of the reachable set $E\left(s_i\right)$, the prior set $F\left(s_i\right)$ and the common set based on the reachable matrix. According to Equation (14), the reachability matrix of organizational resilience influencing factors of MTIPs is hierarchically processed to construct a multilevel recursive order structure model.

$$F\left(s_i\right)=F\left(s_i\right)\cap E\left(s_i\right)$$

(14)

Step 9: MICMAC Analysis

Calculate the driving force values and dependency values between the factors through the reachability matrix and draw a graph of the MICMAC result analysis.

$$\{\begin{array}{l}{W}_i={\displaystyle \sum _{i=1}^n{k}_{ij},\left(i=1,2,\cdots ,n\right)}\\ Y_j={\displaystyle \sum _{j=1}^n{k}_{ij},\left(j=1,2,\cdots ,n\right)}\end{array}$$

(15)

where $W_i$ is the i^th row sum of the reachability matrix $K$ and $Y_j$ is the j^th column sum of the reachability matrix $K$.

4. Results

4.1. Data Collection

This study invites relevant experts to conduct interviews, surveys, and scores based on the established indicator system of factors affecting the organizational resilience of MTIPs.

A total of 10 researchers and experts in major transportation-related fields (with an average working time of more than 12 years) scored the relationship between the roles of the 20 influencing elements in Figure 4. These include five design engineers in the field of major transportation, all of whom have been engaged in full-time transportation engineering design work for more than six years, three experts for operation and management in the field of major transportation, all of whom are working in transportation group operating companies and have been working for more than six years and two university teachers in the field of major transportation infrastructure research, all with the title of professor.

Expert group members combine their professional knowledge and work experience to judge the role of each influencing factor in the project’s organizational resilience influencing factor indicator system. To ascertain the degree of interaction among the 20 resilience impact indicators, the scoring range is 0–4 [60]. In order to ensure the objectivity and accuracy of the scoring, the highest and lowest scores were excluded from the expert scoring results, and the arithmetic average of the remaining scores was taken as the final measured value of the indicator. The semantic transformation table (Table 4) was used to convert the experts’ language variables into triangular fuzzy integers. The initial direct influence matrix is defuzzified through Equations (1)–(4) using MATLAB R2022b software to obtain the direct influence matrix $Z$, as shown in

Table 5. Direct impact matrix.

Factor	S1	S2	S3	S4	S5	S6	S7	S8	S9	S10	S11	S12	S13	S14	S15	S16	S17	S18	S19	S20
S1	0	0.0595	0.4891	0.2273	0.4891	0.0595	0.2273	0.8333	0.4891	0.4891	0.8333	0.2273	0.2273	0.0595	0.8333	0.0595	0.2273	0.4891	0	0.8333
S2	0.2273	0	0.2273	0.2273	0.4891	0.2273	0.2273	0.8333	0.2273	0.2273	0	0.0595	0.2273	0	0.4891	0.2273	0.4891	0.0595	0.2273	0
S3	0.4891	0.2273	0	0.4891	0.4891	0.4891	0.2273	0.2273	0.4891	0.2273	0.4891	0.2273	0.2273	0.4891	0.0595	0.2273	0.2273	0.0595	0.4891	0.2273
S4	0.4891	0.0595	0.4891	0	0.0595	0.2273	0.4891	0.0595	0.2273	0.0595	0.8333	0.2273	0.8333	0.0595	0.0595	0.2273	0.0595	0.0595	0.4891	0.2273
S5	0.2273	0.0595	0.4891	0.4891	0	0.4891	0.4891	0.0595	0.4891	0.2273	0.2273	0.4891	0.4891	0	0.2273	0.2273	0.0595	0.0595	0.8333	0.2273
S6	0.4891	0.0595	0	0.4891	0.8333	0	0.4891	0.4891	0.4891	0.4891	0.2273	0.2273	0.4891	0	0.0595	0.2273	0.0595	0.2273	0.2273	0.0595
S7	0.2273	0.0595	0.4891	0.4891	0.0595	0.4891	0	0.0595	0.2273	0.2273	0.0595	0.2273	0.4891	0.2273	0.2273	0.2273	0	0	0.2273	0.2273
S8	0.4891	0.8333	0.2273	0.4891	0.2273	0.4891	0.0595	0	0.2273	0.8333	0.2273	0.0595	0.2273	0.2273	0.0595	0.4891	0.2273	0.2273	0.4891	0
S9	0.2273	0.4891	0.4891	0.4891	0.4891	0.2273	0.4891	0.2273	0	0.4891	0.0595	0.8333	0.2273	0	0.4891	0.4891	0.4891	0.2273	0.2273	0.8333
S10	0.8333	0.2273	0.4891	0.4891	0.0595	0.2273	0.2273	0.8333	0.2273	0	0.2273	0.0595	0.4891	0.2273	0.8333	0.2273	0.2273	0.2273	0.4891	0.8333
S11	0.2273	0.0595	0.2273	0.4891	0.2273	0.0595	0.0595	0	0.2273	0.0595	0	0.4891	0.4891	0.2273	0.2273	0.0595	0	0.2273	0.4891	0.4891
S12	0.4891	0.0595	0.4891	0.8333	0.2273	0.2273	0.2273	0.0595	0.4891	0.2273	0.2273	0	0.2273	0.2273	0.2273	0.0595	0	0.0595	0.8333	0.8333
S13	0.2273	0.2273	0.4891	0.4891	0.2273	0.2273	0.4891	0.2273	0.4891	0.4891	0.2273	0.2273	0	0.4891	0.2273	0.4891	0.2273	0.2273	0.4891	0.0595
S14	0.8333	0.0595	0.8333	0.2273	0.4891	0.4891	0.4891	0.2273	0.2273	0.2273	0.4891	0.2273	0.4891	0	0.8333	0.4891	0.2273	0.4891	0	0.2273
S15	0.2273	0.0595	0.2273	0.0595	0.2273	0.0595	0.4891	0.0595	0.4891	0.2273	0.0595	0.4891	0.4891	0.0595	0	0.0595	0.0595	0.2273	0.4891	0.8333
S16	0.2273	0.2273	0.8333	0.4891	0.4891	0.2273	0.4891	0.4891	0.2273	0.4891	0.2273	0.2273	0.4891	0.0595	0.2273	0	0	0.0595	0.8333	0.2273
S17	0.2273	0.2273	0.0595	0.2273	0.0595	0.2273	0	0.2273	0	0.4891	0.2273	0	0.2273	0.2273	0.8333	0.2273	0	0.2273	0.4891	0.0595
S18	0.8333	0.4891	0.2273	0.4891	0.0595	0.2273	0.2273	0.4891	0.2273	0.2273	0.2273	0.2273	0.0595	0.2273	0.4891	0.0595	0.4891	0	0.8333	0.0595
S19	0.2273	0	0.4891	0.4891	0.2273	0.4891	0.2273	0.4891	0.0595	0.2273	0.4891	0.2273	0.4891	0	0.2273	0.2273	0.2273	0.2273	0	0.2273
S20	0.4891	0.0595	0.8333	0.2273	0.2273	0.0595	0.4891	0.0595	0.4891	0.4891	0.8333	0.2273	0.4891	0.0595	0.2273	0.2273	0.0595	0.2273	0.4891	0

.

4.2. Analysis of Fuzzy DEMATEL Results

According to Equations (5) and (6), to obtain the comprehensive impact matrix. After the calculation of Equations (7)–(10), the summary of the four important indicators of the organizational resilience of MTIPs influencing factors data influence degree $D_i$, affected degree $C_i$, centrality degree $N_i$ and cause degree $R_i$. The centrality ranking and the attribute information of the influencing factors are presented in

Table 6. The centrality and cause degree, influence and affected degree of each factor.

Factor	${\mathit{D}}_{\mathit{i}}$	${\mathit{C}}_{\mathit{i}}$	${\mathit{N}}_{\mathit{i}}$	${\mathit{R}}_{\mathit{i}}$	Order of Centrality	Factor Properties
S1	4.188	4.375	8.563	−0.187	1	Effect factor
S2	2.736	2.024	4.759	0.712	19	Cause factor
S3	3.616	4.788	8.404	−1.173	2	Effect factor
S4	3.039	4.658	7.697	−1.619	8	Effect factor
S5	3.426	3.286	6.712	0.14	10	Cause factor
S6	3.387	3.078	6.465	0.309	15	Cause factor
S7	2.582	3.642	6.224	−1.06	17	Effect factor
S8	3.641	3.406	7.047	0.235	9	Cause factor
S9	4.337	3.662	7.998	0.675	6	Cause factor
S10	4.4	3.735	8.135	0.665	4	Cause factor
S11	2.607	3.919	6.525	−1.312	13	Effect factor
S12	3.592	3.057	6.648	0.535	12	Cause factor
S13	3.725	4.451	8.176	−0.726	3	Effect factor
S14	4.454	1.783	6.237	2.671	16	Cause factor
S15	2.951	3.737	6.688	−0.786	11	Effect factor
S16	3.83	2.676	6.506	1.154	14	Cause factor
S17	2.509	1.922	4.431	0.587	20	Cause factor
S18	3.558	2.142	5.7	1.415	18	Cause factor
S19	3.064	4.978	8.042	−1.914	5	Effect factor
S20	3.695	4.017	7.712	−0.323	7	Effect factor

.

According to Table 6, the cause–result diagram of factors influencing the organizational resilience of the project is plotted (Figure 6), in which the distribution and values of each influencing factor can be clearly seen to determine the influence of each factor on the project’s organizational resilience.

4.2.1. Analysis of Centrality and Causality

Among the organizational resilience-affecting factors for MTIPs, the more important a factor is, the higher its centrality degree signifies. The significance of a factor’s influence on the organizational resilience of the project and its relationship to other factors are reflected in its centrality. From the analysis of Table 4 and Figure 6, it can be seen that risk prediction and warning (S₁), human resource management (S₃), organizational leadership (S₁₃), resource reserve situation (S₁₀), organizational learning (S₁₉), and inter-organizational synergies (S₉) are the top 6 influencing factors in terms of centrality. This implies that they are important variables impacting the organizational resilience of the project and that their consideration should be enhanced during project implementation.

In terms of stability, S₁ and S₃ have higher centrality, indicating that effective risk prediction and early warning mechanisms can help project teams identify potential risks in advance and take appropriate measures to cope with them. Furthermore, excellent human resource management strategies will also have a positive impact on project outcome outputs. In the area of redundancy, S₉ and S₁₀ are more important influences on the organizational resilience of a project, suggesting that the ability of the project team to integrate resources and work together is very important when the project is faced with unexpected events or crises. With regard to the degree of rapidity, S₁₃ has the greatest degree of centrality, showing that organizational leadership is the critical factor influencing the organizational resilience of the project at this stage. On the adaptation dimensions, S₁₉ is considered to be the most critical factor. Organizational learning enhances the project organization’s ability to perceive and adapt to the environment and helps the project organization accumulate experience and wisdom.

Causality is an indicator that classifies the attributes of a factor; a causality greater than zero is a cause factor, and less than zero is an effect factor. Contributing factors are generally indirect factors that influence organizational resilience in MTIPs, while resulting factors are direct factors. Eleven cause factors and nine factors make up the organizational resilience influencing factors of MTIPs, as shown in Figure 6. Among the cause factors, the top four factors in terms of influence are inter-organizational synergy (S₉), resource reserve situation (S₁₀), safety education efforts (S₁₂), and emergency material security (S₈). The above four causal factors are indirect motivators that affect organizational resilience in major transportation infrastructure projects. Among the result factors, the top four factors with the absolute value of the degree of cause are organizational learning (S₁₉), project environment (S₄), organizational structure (S₇), and completeness of contingency plans (S₁₁), which suggests that the above factors are more influenced by other factors and are also the direct motivators affecting the organizational resilience of MTIPs.

4.2.2. Analysis of Influence Degree and Affected Degree

Influence degree and affected degree are measures of the strength of the factor’s correlation with other factors. In Figure 7, the top five factors ranked in terms of influence degree are risk warning and prediction (S₁), inter-organizational synergies (S₉), resource reserve situation (S₁₀), efficiency of information delivery (S₁₄), and organizational coordination (S₁₆), demonstrating that these five factors have a greater influence on the other factors. Among these factors, risk warning and prediction (S₁) is closer to being influenced, and the factor is ranked first in centrality. This suggests that risk prediction and warning capability are critical to the enhancement of project organizational resilience.

For the affected degree factors, human resource management (S₃), project environment (S₄), organizational structure (S₇), completeness of emergency plans (S₁₁), organizational leadership (S₁₃), decision making and responsiveness (S₁₅), and organizational learning (S₁₉) were less influential than the affected degree factors, reflecting the fact that these types of factors are more susceptible to the other factors in general. There is a relatively high centrality of human resource management (S₃), organizational leadership (S₁₃), and organizational learning (S₁₉) among the above factors, suggesting that these types of factors are also more perturbing to the organizational resilience of the project when they are being influenced.

4.3. Analysis of ISM Results

The adjacency matrix is determined by Equation (12). The value of $\lambda$ is usually selected based on the experience of experts, yet it is less objective. In this paper, referring to existing studies [61], the algorithm of adding the mean and standard deviation of all factors in the integrated influence matrix $T$ is adopted to reduce the subjective influence. In this paper, $\lambda =0.235$ is taken to obtain the adjacency matrix, and the reachability matrix (

Table 7. Reachability matrix.

Factor	S1	S2	S3	S4	S5	S6	S7	S8	S9	S10	S11	S12	S13	S14	S15	S16	S17	S18	S19	S20
S1	1	1	0	1	0	1	0	0	0	1	1	1	0	1	1	1	1	1	0	0
S2	1	1	0	1	0	1	0	0	0	1	1	0	0	1	1	1	1	1	0	0
S3	0	0	1	1	1	0	1	0	0	1	1	0	1	1	1	1	1	1	0	0
S4	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
S5	0	0	0	1	1	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0
S6	0	0	0	0	0	1	0	0	1	0	1	0	0	1	0	1	0	0	0	0
S7	0	0	0	1	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0
S8	0	0	0	0	0	0	0	1	0	0	1	0	0	1	0	1	0	0	0	0
S9	0	0	0	0	0	0	0	0	1	0	1	0	0	1	0	1	0	0	0	0
S10	0	0	0	1	0	0	0	0	0	1	1	0	0	1	1	1	1	1	0	0
S11	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	0
S12	1	0	0	0	0	0	0	0	1	0	1	1	0	1	0	1	0	0	0	0
S13	0	0	1	1	1	0	1	0	0	1	1	0	1	1	1	1	1	1	0	0
S14	0	0	0	0	0	0	0	0	0	0	1	0	0	1	0	1	0	0	0	0
S15	0	0	0	0	0	0	0	0	0	0	1	0	0	1	1	1	0	0	0	0
S16	0	0	0	0	0	0	0	0	0	0	1	0	0	1	0	1	0	0	0	0
S17	0	0	0	1	0	0	0	0	0	1	1	0	0	1	1	1	1	1	0	0
S18	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0
S19	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0
S20	0	0	0	0	0	0	0	0	1	0	1	0	0	1	0	1	0	0	0	1

) is calculated by Equation (13).

According to the principle of hierarchical division, the reachable matrix $K$ is hierarchized to obtain the reachable set, the prior set, and its intersection. When $F\left(s_i\right)=F\left(s_i\right)\cap E\left(s_i\right)$, the filtered influencing factor $S_i$ is the 1st level factor, and then the above factor label is deleted from the reachable matrix $K$, thus obtaining the factors of each level. The set of influencing factors is shown in

Table 8. The set of influencing factors.

Factor	Reachable Set	Prior Set	Intersection Set
S1	1,2,4,6,10,11,12,14,15,16,17,18	1,2,12	1,2,12
S2	1,2,4,6,10,11,14,15,16,17,18	1,2	1,2
S3	3,4,5,7,10,11,13,14,15,16,17,18	3,13	3,13
S4	4	1,2,3,4,5,7,10,13,17,18	4
S5	4,5,7	3,5,13	5
S6	6,9,11,14,16	1,2,6	6
S7	4,7	3,5,7,13	7
S8	8,11,14,16	8	8
S9	9,11,14,16	6,9,12,20	9
S10	4,10,11,14,15,16,17,18	1,2,3,10,13,17	17,10
S11	11	1,2,3,6,8,9,10,11,12,13,14,15,16,17,20	11
S12	1,9,11,12,14,16	1,12	1,12
S13	3,4,5,7,10,11,13,14,15,16,17,18	3,13	3,13
S14	11,14,16	1,2,3,6,8,9,10,12,13,14,15,16,17,20	16,14
S15	11,14,15,16	1,2,3,10,13,15,17	15
S16	11,14,16	1,2,3,6,8,9,10,12,13,14,15,16,17,20	16,14
S17	4,10,11,14,15,16,17,18	1,2,3,10,13,17	17,10
S18	4,18	1,2,3,10,13,17,18	18
S19	19	19	19
S20	9,11,14,16,20	20	20

.

Based on the results of the element level division in Table 7, the following final stratification results were obtained: L₁ = {S₄,S₁₁,S₁₉}; L₂ = {S₇,S₁₄,S₁₆,S₁₈}; L₃ = {S₅,S₈,S₉,S₁₅}; L₄ = {S₆,S₁₀,S1₂,S₁₇,S₂₀}; and L₅ = {S₁,S₂,S₃,S₁₃}. The ISM hierarchical structure of organizational resilience of MTIPs is illustrated, and based on the findings of the hierarchical analysis, a multilevel recursive structural model is constructed to ascertain the recursive structural linkages among the system’s elements. As shown in Figure 8, the index system of organizational resilience of MTIPs is a five-level multilevel recursive rank system.

The ISM model is able to visualize the mutual interactions and hierarchical structure among the influencing elements. Figure 8 demonstrates the complexity of the linkages between the organizational resilience influences of MTIPs, which can occur in adjacent tiers or across tiers. In the multilevel recursive order structure model, each affecting factor is divided into three levels and five ranks, and each affecting factor has a clear hierarchical structural change, and the closer to the bottom layer indicates that the affecting factor is more fundamental. First, the fifth layer includes risk prediction and warning (S₁), equipment condition and performance (S₂), human resource management (S₃), and organizational leadership (S₁₃), which are the underlying influences of organizational resilience, the most basic and important influences in the whole system and play a key role in the organizational resilience of MTIPs. It also suggests that risk prediction and warning, equipment condition and performance, human resource management, and organizational leadership do not directly affect organizational resilience but rather influence it by influencing other factors.

Second, the nine intermediate factors: Decision making and responsiveness (S₁₅), inter-organizational synergies (S₉), emergency material security (S₈), natural environment (S₅), organizational culture (S₆), resource reserve situation(S₁₀), safety education efforts (S₁₂), accident cause investigation capacity (S₁₇), and organizational change (S₂₀) are at the third and fourth levels of the model. In particular, organizational culture and organizational change belong to the organizational system. Decision making and responsiveness, inter-organizational synergy, safety education efforts, and accident cause investigation capacity belong to the management system. Emergency material security and resource reserves belong to the material and technical systems. The natural environment belongs to the environmental system, which is different from the enterprise organizational resilience of previous studies. Engineering projects are often limited by geographic location, and external natural environment risks need to be taken in the preliminary geological survey and design stage to take appropriate measures to resolve them. Although these factors do not directly affect the specific aspects of organizational resilience of MTIPs, they indirectly affect the organizational resilience of MTIPs by affecting the organization’s decision making, synergy, emergency response security, culture, resource reserves, safety education, accident prevention, and change capability. Therefore, in project management, these factors should be fully considered, and corresponding measures should be taken to enhance organizational resilience.

Third, the factors organizational learning (S₁₉), completeness of emergency plans (S₁₁), and project environment (S₄) at the first level of the model and organizational structure (S₇), efficiency of information delivery (S₁₄), organizational coordination (S₁₆), and equipment replenishment and repair (S₁₈) at the second level are direct influences, which have a direct driving effect on the project’s organizational resilience and are affected by indirect influences and fundamental influencing factors. The project environment and organizational structure belong to the stability category. Completeness of emergency plans belongs to the redundancy level. efficiency of information delivery and organizational coordination are at the rapidity level. Organizational learning and equipment replenishment and repair refer to the adaptability dimension. In summary, these factors directly contribute to the resilience of project organizations by affecting their stability, adaptability, and coping capacity. Consequently, these factors should be fully considered and optimized in project management in order to improve the resilience of project organizations.

4.4. Analysis of MICMAC Results

Based on the reachability matrix $K$, the driving force and dependency of each influencing factor were calculated according to Equation (15), and the results were obtained as shown in

Table 9. Driving forces and dependencies of factors.

Factor	Driving Force	Dependency
S1	12	3
S2	11	2
S3	12	2
S4	1	10
S5	3	3
S6	5	3
S7	2	4
S8	4	1
S9	4	4
S10	8	6
S11	1	15
S12	6	2
S13	12	2
S14	3	14
S15	4	7
S16	3	14
S17	8	6
S18	2	7
S19	1	1
S20	5	1

.

Based on the data in Table 9, the factors were divided into four quadrants in the form of two-dimensional axes as autonomous, dependent, associated, and independent factors, respectively, as shown in Figure 9; where the sum of the rows of the reachability matrix is the driving force and the sum of the columns is the degree of dependency. The set of autonomous factors has weak drivers and dependencies; the set of dependency factors has strong dependencies and weak drivers; the set of associated factors has strong drivers and dependencies; and the set of independent factors has strong drivers and weak dependencies.

The set of dependency factors in quadrant II with high dependency and low driving force includes completeness of emergency plans (S₁₁), efficiency of information delivery (S₁₄), organizational coordination (S₁₆), and project environment (S₄). Due to the weaker driving force, they are more susceptible to other factors, and these factors are located at the direct level of the ISM model, which can be driven by enhancing other factor indicators.

Autonomous factors in quadrant III with a lower driving force and a lower degree of dependency include factors such as organizational culture (S₆), safety education efforts (S₁₂), and organizational change (S₂₀), which are not easily affected by other factors, relatively independent, and easy to control, and are the focus and launching point of organizational resilience enhancement of MTIPs at this stage. At the same time, these factors are located at the middle level of the ISM model, playing the role of a bridge between the top and the bottom.

Quadrant IV contains six independent factors that are more strongly driven and less vulnerable to the influence of other factors, and are generally located at the bottom of the model. In this regard, risk warning and prediction (S₁), equipment condition and performance (S₂), human resource management (S₃), and organizational leadership (S₁₃), as the fundamental influencing factors of the ISM model, have far-reaching impacts on the organizational resilience of MTIPs. Meanwhile, risk warning and prediction (S₁), human resource management (S₃), resource reserve status (S₁₀), and organizational leadership (S₁₃) are also factors with high DEMATEL centrality.

In conclusion, the accuracy and validity of the modeling analysis in this study were demonstrated by the comparison between the results of the MICMAC analysis and the DEMATEL-ISM model.

5. Discussion

MTIPs are temporary, multi-stakeholder projects to achieve project objectives, which are characterized by complexity, longevity, and large scale, and there is a strong need to improve the organizational resilience of the project to guard against unknown situational risks. Organizational resilience is widely regarded as a capability, behavior, and outcome that is continuously absorbed when the system is faced with a perturbation, thereby enabling itself to perform resistance behaviors in order to restore the system to its original state or to achieve a more optimal state.

First, in the identification of factors affecting organizational resilience in MTIPs, some scholars in existing studies have proposed human resource management [28,29], organizational leadership [31], and decision making and responsiveness [32] as key factors affecting organizational resilience. There is a deficiency of study on the influencing variables of organizational resilience in the field of engineering projects, and the majority of resilience influencing factors are centered on the enterprise level. In fact, there are also similarities and differences between engineering project organizations and enterprise organizations. Therefore, this paper first adopts the factors related to corporate organizational resilience, and through the case study of the Nanchang Metro Line 4 project, combined with the characteristics of MTIPs, it is found that the project environment, the state and performance of equipment, the risk warning and prediction, the emergency material security, and the situation of the resource reserve are also factors of organizational resilience in MTIPs. The results of the influencing factor model analysis in this paper not only verify the accuracy of previous studies, but also offer helpful sources for future research on organizational resilience in the context of engineering projects.

Second, this study analyzes the critical factors influencing the organizational resilience of MTIPs based on fuzzy DEMATEL and adopts the fuzzy set theory to convert the linguistic evaluations of experts into mathematical expressions, which further eliminates human uncertainty. It was found that risk prediction and warning, human resource management, inter-organizational synergies, resource reserve situation, organizational leadership, and organizational learning are the critical factors influencing organizational resilience in MTIPs. The above factors belong to the four characteristic elements of resilience, which indicates that in the construction process of major transport infrastructure projects, dynamic attention should be paid to the characteristic elements of organizational resilience in the pre-disaster or post-disaster phases to ensure that they can maintain a stable, efficient, and sustainable operating state. The most important of these factors are risk prediction and warning, human resource management. In terms of the category of organizational resilience characteristic elements, both factors lie at the level of stability. Stability, as a vital aspect of organizational resilience, determines whether an organization can maintain smooth operation in the face of challenges and changes, and stability is closely related to continuous risk prediction management [62]. The impact of risk prediction and warning on organizational stability is self-evident. In a complex and changing external environment, organizations are faced with risk challenges from various sources. Effective risk prediction and warning can help organizations identify potential risks in advance so that they can respond quickly to avoid or reduce the losses caused by them. Human resource management is also essential to organizational stability. The quality, skills, and attitudes of employees have a direct impact on the operational efficiency and stability of an organization. The research makes it abundantly clear that in the process of enhancing organizational resilience, great importance should be attached to the stability of the project, and the stability of the organization should be enhanced through continuous optimization of the risk prediction and warning mechanism and strengthening of human resource management and other measures so as to enhance organizational resilience.

Third, this paper generates a three-level five-rank hierarchical diagram by constructing the ISM model of organizational resilience for MTIPs, which divides the factors influencing organizational resilience for MTIPs into direct, transitional, and fundamental influencing factors. In L₅, there exists a strong correlation between organizational leadership and human resource management, and the state performance of equipment affects the project’s ability to warn and predict risks. Leaders need to use modern scientific methods to train, organize, and deploy human resources in order to achieve organizational goals. Furthermore, projects with equipment in good condition and stable performance tend to have higher risk warning and prediction capabilities and are able to detect and respond to potential risks in a timely manner. One of the findings of this study is that organizational leadership is found to have a deep influence on the project’s organizational resilience. The findings of Teo et al. [63] support this view. In addition, organizational coordination is at the upper level of the hierarchical model and has a direct influence on the organizational resilience of the project. This further validates Bourne’s [64] view that the collaboration of members in an organization has a greater impact on project success. However, there are also works that are not quite consistent with the results of this study. For example, this study found that organizational learning has a direct influence on project organizational resilience, whereas Liu et al. [28] argued that organizational learning indirectly influences corporate organizational resilience. This is primarily because the relationship between the influencing aspects of project organizational resilience has not been thoroughly explored in the studies that have already been carried out. Organizational learning belongs to the category of adaptability and is the factor with the most centrality in this category. It is obvious to discover that during a crisis or disturbance, project organizations must quickly adjust to the new circumstances, gain pertinent expertise, and engage in continual learning to stay abreast of external developments [65]. Finally, this study calculates the driving force and dependence of each influencing factor based on the MICMAC method, divides the factors into autonomous, dependent, associated, and independent factors, and compares the results with the DEMATEL-ISM analysis to verify the validity of the model analysis.

6. Conclusions

In order to further enrich and develop the theoretical research on organizational resilience of MTIPs and to enhance the project’s anti-risk level and crisis prevention ability, this paper adopts the hybrid fuzzy DEMATEL-ISM-MICMAC method, which provides a comprehensive analytical framework for the organizational resilience influencing factors of MTIPs and visualizes the complex systematic relationship among the influencing factors of organizational resilience. The conclusions of this paper are as follows:

(1)

Literature analysis method and case study method are used to establish the index system of organizational resilience influencing factors of MTIPs by combining the characteristic elements of resilience (stability, redundancy, rapidity, and adaptability) and project organizational subsystems (organizational system, management system, environmental system, and material and technological system).

(2)

Organizational resilience-influencing factors were analyzed using the fuzzy DEMATEL method. The results of the analysis show that risk prediction and warning, human resource management, inter-organizational synergy, resource reserve situation, organizational leadership, and organizational learning have a higher degree of centrality and have the greatest contribution and importance to the project organizational resilience system. Therefore, project organizations should focus on the above important influencing factors from the dynamic evolution stage of organizational resilience when enhancing resilience.

(3)

ISM analysis shows that risk prediction and warning, equipment condition and performance, human resource management, and organizational leadership are the fundamental factors affecting the project’s organizational resilience. The above factors belong to the organizational system, management system, and material and technical system, respectively, which indicates that in the process of construction of MTIPs, it is necessary to construct the project system in an all-round, comprehensive, and multi-level way to ensure the smooth implementation of the project.

(4)

The MICMAC analysis found that risk prediction and warning, equipment condition and performance, human resource management, resource reserve situation, and organizational leadership have strong drivers and low dependency and are classified as independent factors, which can be considered foundational factors to drive the development of other factors. The findings of this paper can inform the optimization of organizational resilience in MTIPs.

However, with the application of new technologies and tools in the field of MTIPs, the study of organizational resilience in MTIPs in this paper has some limitations. Above all, in terms of the selection of indicators of organizational resilience of MTIPs, the selection of indicators of organizational resilience influencing factors in this study is based on the relevant literature to analyze the Nanchang Metro Line 4 project research. Due to the geographic environment, construction difficulties, etc., the selection of indicators will change, and the subsequent study can be based on the characteristics of the different projects to choose a more relevant indicator system. Furthermore, project organizational resilience is a dynamic process that requires regular review and updating of the indicator system based on project progress and changes in the external environment. In addition, the accuracy of indicator weights has a greater impact on the analysis of project organizational resilience. This paper adopts the expert scoring method to obtain the direct impact matrix. Despite the use of the fuzzy set theory for defuzzification, there is still the subjective impact of expert preference and professional level. In the subsequent research, a resilience indicator model can be constructed based on big data or artificial intelligence to improve the accuracy and validity of the indicator weights. Finally, in terms of the choice of research methods, interdisciplinary research methods can be applied to explore in depth the internal and external factors affecting the resilience of project organizations.