Critical Success Factors for Enhancing Intelligent Loading and Unloading in Urban Supply Chains: A Comprehensive Approach Based on Fuzzy DEMATEL-AISM-MICMAC

Abstract

With the development of the smart logistics industry, the demand for intelligent loading and unloading (ILU) within urban supply chains (USCs) is increasing. However, few studies have examined the critical success factors (CSFs) for enhancing ILU in USCs. This study establishes a CSF model to support ILU improvement. Specifically, it integrates stakeholder theory, resource-based view theory, and innovation diffusion theory. Through research conducted in collaboration with 16 logistics industry specialists in Korea, 19 critical factors were identified. Fuzzy DEMATEL and the Adversarial Interpretive Structure Model (AISM) were then applied to analyze the identified factors. The results indicate that stakeholder collaboration, government support, and regulatory compliance are the most important factors affecting ILU improvement within USCs. Finally, cross-impact matrix multiplication applied to classification (MICMAC) analysis further verifies that these factors have a high driving power and low dependence, making them independent driving factors of the entire system. Furthermore, this study emphasizes the role of market research and automated system design. This work contributes to the knowledge on the intelligent logistics management of supply chains.

1. Introduction

Loading and discharging are critical in urban supply chains (USCs) [1]. They affect entire USCs’ operations depending on their reliability and efficacy [2]. However, with the increase in urban logistics demand, the efficiency and reliability of supply chain loading and unloading have been significantly affected. According to the latest report from the China Academy of Information and Communications Technology, by 2023, the adoption of smart logistics technologies had contributed to a 35% increase in urban supply chain efficiency, reinforcing their pivotal role in driving economic growth and strengthening supply chain resilience [3]. In this context, intelligence has evolved into a differentiation strategy to sustain supply chain competitiveness in the present enterprise environment [4]. Intelligence has emerged as a critical factor in the optimization of efficiency and the reduction in costs throughout USCs [5]. More specifically, technology enables intelligent loading and unloading (ILU) to provide fast, efficient, and accurate goods loading and unloading [6]. For instance, in Shanghai, automated warehouses equipped with AI-powered robotic arms and IoT-enabled tracking systems have optimized loading and unloading operations, reducing manual labor reliance and minimizing errors [7]. ILU enhances the hardware equipment utilized to convey products within USCs and renders the entire supply chain intelligent at the system level rather than at the individual production chain level [8]. Consequently, enhancing the level of ILU has emerged as a critical strategic domain for urban transportation organizations. Unfortunately, there is limited research focus on ILU within USCs.

In fact, the challenges faced by the automation of loading and unloading transcend organizational, managerial, and information systems, and even social issues [9]. As the logistics industry increasingly demands reduced time, lower costs, and improved overall efficiency, the adoption of intelligent logistics within USCs has become particularly urgent [10], and enhancing ILU within USCs has become an important focus for logistics companies [11]. However, due to the relative difficulty of popularizing ILU, most current ILU systems still require human intervention [8]. Additionally, the lack of personnel with specialized knowledge and the ability to fully realize new technologies’ potential remains a significant issue. Thus, these personnel must undergo training to better collaborate with automated loading and unloading machines [10]. Therefore, the main challenges faced by logistics companies potentially include a lack of awareness of new technologies, insufficient financial resources, insufficient expertise, and the need for customized intelligent equipment [10,12]. Thus, to succeed amidst today’s intense market competition, logistics companies must identify the critical success factors (CSFs) for enhancing ILU within USCs [11]. However, no scholarly discussions about the ILU phase of USCs exist. Therefore, there is an urgent need to systematically analyze the strategic framework of CSFs for ILU in USCs.

Currently, significant controversy surrounding research on intelligence in USCs remains. Firstly, the existing literature mainly focuses on optimizing single technologies such as artificial intelligence, the Internet of Things, and robotics to improve loading and unloading efficiency. However, research on how to achieve cross-technology integration is still lacking [13]. Secondly, the existing research focuses on analyzing the single factors affecting ILU, with few studies emphasizing the interrelationships between these factors [14]. Finally, few studies explore the strategic framework for implementing the automated loading and unloading system activities of USCs [8].

Therefore, the following investigative questions are proposed:

RQ1. What are the CSFs for increasing the automation of loading and unloading in USCs?

RQ2. What are the associations and effects of finalized critical factors?

To address the aforementioned research gap in this field of study, we propose integrating multidimensional factors such as technology, management, and environment into a systematic framework. We stress the “systematic” analytical research of critical factors to elucidate the strategic framework that promotes ILU within USCs. The key is to establish a systematic approach aligned with the development of intelligent strategies within USCs [15]. Therefore, this study comprehensively applies stakeholder theory (ST), resource-based view (RBV) theory, and innovation diffusion theory (IDT) to explore the critical factors driving the automation of USC loading and unloading.

The main contributions of this work are timely and diverse. Firstly, this work enriches the knowledge system of intelligent operations in USCs, provides a decision-making framework for logistics companies to enhance ILU within USCs, and promotes theoretical exploration in the ILU field. Secondly, this study fills the research gap in the literature regarding ILU implementation in urban areas and reveals the influence of various theoretical frameworks on the application of intelligent management in urban logistics. Thus, the research findings offer a strategic framework for logistics companies to implement ILU and help improve urban efficiency, reduce costs, and enhance the competitiveness of USCs.

Section 2 is a literature review and introduces information on improving automated loading. Section 3 presents the research methods employed and identifies critical factors. Section 4 outlines and discusses the findings of the data interpretation. Section 5 outlines the theoretical and managerial contributions of this paper. The final section discusses the study’s limitations and future developments.

2. Literature Review

2.1. Urban Smart Logistics

Smart logistics is an important driver of nations’ and enterprises’ competitiveness, playing a crucial role in economic growth. The logistics industry faces high costs and low efficiency; therefore, developing smarter methods to improve logistics efficiency and reduce logistics costs is an important issue in academic and industrial circles [13]. Smart logistics provides an effective way to seize opportunities created by new technologies and promotes new business models [16]. Urban smart logistics include core businesses, including smart transportation, smart warehousing, and ILU [6,13]. Recently, smart logistics have been used to plan, manage, or control USC development more intelligently [13,16]. Research has also focused more on areas such as smart transportation and smart warehousing.

Specifically, traditional transportation methods face issues including wasted loading capacity, loading task errors, inefficient operations, transportation safety, and cargo security [16]. With the development of IoT technology, intelligent transportation research is shifting to more integrated intelligent transportation systems [13]. This brings opportunities to solve the aforementioned problems. Using smart transportation within USCs can elucidate the logistics transportation process. This has improved transportation efficiency and traffic safety in urban logistics and provided economic benefits [6,13]. Through the synergistic effect of intelligent transportation and ILU, the time of the loading and unloading process and the transportation link can be effectively shortened.

Notably, with the continuous innovations in logistics operations and technology, smart city warehouse management is becoming increasingly complex and critical [17]. Smart warehousing optimizes warehouse space utilization, monitors storage environments, and improves warehouse management processes through IoT technology. The synergy between ILU and intelligent warehousing further enhances inventory management efficiency. It has revolutionized urban logistics warehousing management [8].

It is noteworthy that urban loading and unloading are necessary and repetitive activities that occur in transportation and warehousing operations. Automated loading and unloading can increase loading and unloading efficiency and reduce loading and unloading costs in urban logistics activities and warehousing costs [18,19]. ILU technology automates the loading and unloading process through the combination of information, computing, communication, and connectivity technologies. Its application has triggered significant changes in USCs [20], helping to address uncertain market conditions [21]. Research shows that it is crucial for the innovative development and survival growth of logistics companies in a dynamic business environment [22]. ILU research primarily focuses on improving the performance of intelligent equipment systems and applying machine learning to optimize loading and unloading efficiency [21,23]. Additionally, scholars also focus on researching ILU equipment and systems [24]. This includes technical interactions, control algorithm technology, and Automated Guided Vehicle (AGV) applications [25,26]. Furthermore, the cost-effectiveness of ILU technology has attracted significant attention from scholars [8]. Although ILU has become a research topic in the field of smart logistics [8,23], the existing research on smart logistics mainly focuses on smart transportation and smart warehousing, with relatively little research on improving ILU in the smart logistics field. Notably, in the smart logistics industry, decision support frameworks for improving ILU are lacking, along with comprehensive technologies and innovation management to promote ILU enhancements [27]. This demonstrates the necessity of exploring the CSFs for ILU in the USC delivery phase. Therefore, this field requires more attention.

Table 1. Research gaps.

Serial Number	Research Gap	Research Topic	Author
1	Improve loading and unloading efficiency through the optimization of one technology.	Applications of the Internet of Things (IoT) in smart logistics: A comprehensive survey	[13]
		Energy-efficient rail guided vehicle routing for two-sided loading/unloading automated freight handling system	[23]
		From plant and logistics control to multi-enterprise collaboration	[21]
2	Exploring the reasons affecting intelligent loading and unloading through a single factor.	An exploratory investigation of the effects of supply chain complexity on delivery performance	[14]
3	Research on the Key Strategic Framework for the Lack of Automation Systems.	Automation in cargo loading/unloading processes: do unmanned loading technologies bring benefits when both purchase and operational cost are considered?	[8]
		Optimal marketing strategy: A decision-making with ANP and TOPSIS	[27]

shows the gap in the literature.

2.2. Development of Factor Theory Research

To systematically explore the CSFs for promoting ILU within USCs, three closely related theories are summarized. Specifically, stakeholders are “any group or individual who can affect or is affected by the achievement of an organization’s objectives” [28]. In the study of ILU, ST emphasizes the effects of the relationship between logistics companies and various stakeholders on decision-making and practice. This theoretical perspective considers pressures from various stakeholders, which may be coercive, normative, or mimetic, thereby affecting enterprises’ ILU decisions [15]. Therefore, ST helps identify the main external factors affecting automated loading and unloading. ST emphasizes that USC decisions and performance are greatly affected by the extensive involvement of multiple stakeholders. Therefore, numerous stakeholders create complexity in USC decision-making [11]. Therefore, to strengthen competitive advantages, many companies consider designing and implementing a good supply chain system to be important. In this context, establishing close and long-term trust relationships between buyers and suppliers is a crucial element for the successful construction of a supply chain system [29]. Fortunately, ST encourages managers to clarify how they want to conduct their business and what kind of relationships they want and must establish with stakeholders to achieve their goals [15]. At the same time, ST places higher demands on organizations regarding compliance with environmental regulations [11]. Therefore, USC development depends on the effective management of stakeholder relationships and the optimization of energy efficiency [11,15].

RBV theory emphasizes the importance of firm resources [30]. It posits that numerous firm resources within the same industry exhibit significant heterogeneity and comparatively fixed mobility [31]. Therefore, each enterprise’s resource mix is unique [30]. RBV theory helps to identify the internal resources that companies need to focus on developing when enhancing their automated loading and unloading capabilities. Additionally, operational training, innovative improvement, integrated support, and functional scalability between logistics companies and stakeholders are strategic resources for enterprises [15]. From an operational strategy perspective, ILU capabilities, as well as a company’s strategic resources, are both valuable, rare, and difficult to substitute [27]. ILU technology also helps drive customer and market knowledge creation, thereby enhancing marketing decision-making capabilities and improving the overall enterprise performance [32]. Therefore, through business alliances, the integration of existing resources with those of partners helps generate synergies and add value under integration conditions [30]. However, cost and technical implementation issues are receiving increasing attention [22]. Notably, companies with competitive advantages must simultaneously gain economic, technological innovation, and organizational benefits [17,22]. Therefore, ILU in USCs critically focuses on innovation and organizational learning based on RBV theory to enhance the competitive vitality of enterprises [15].

IDT emphasizes the process of technology spreading across numerous organizations [33,34]. It describes the decision-making process of enterprises adopting an innovation service or product, and how such technology increasingly creates decisions that influence enterprises to adopt new technologies [35]. Therefore, IDT helps to analyze the key factors influencing enterprises’ adoption of automated loading and unloading. The emergence of innovative products or technologies challenges manufacturing enterprises’ marketing [34]. Generally, software is a technology that can be evaluated based on its usability (i.e., suitability) and the benefits it provides an organization [36]. In the ILU field, diffusion refers to the process by which ILU technology spreads over time through specific channels among stakeholders or within the logistics industry. In this dissemination process, channel relationships and partners are particularly important for the promotion of ILU [34]. The supply chain management field especially requires continuous change, innovation, and the subsequent dissemination of these innovations. Therefore, supply chain managers integrate innovative structures and technologies to extend these innovations across the entire organization’s areas of responsibility [36,37]. Simultaneously, logistics companies are establishing framework rules and standards to enhance their legality with ILU equipment companies [28]. Therefore, in the ILU phase of USCs, IDT-based attention focuses on scalability features and meeting customized needs, making it crucial to enhance the market dissemination and application of ILU [34].

3. Research Methodology

3.1. Research Design

This research is divided into three phases. Figure 1 shows the analytical framework for constructing CSFs to improve ILU using fuzzy DEMATEL and AISM techniques.

The first phase includes identifying the CSFs to enhance ILU in USCs. A literature review was conducted and expert opinions were obtained to identify important facilitating factors for implementing sustainable initiatives. In terms of theoretical support, IDT describes how types of innovation or technology are accepted and spread within societies of various sizes, providing a theoretical foundation for the promotion of automated loading and unloading in USCs [15]. Secondly, ST emphasizes that an organization’s success depends on how it manages its relationships with key groups to promote technology adoption [11]. Finally, RBV theory posits that a company’s competitive advantage stems from the alloplasm properties of resources in the provision chain, providing support in identifying the technologies and resources needed for ILU [31]. Through literature searches and interviews with logistics industry specialists to validate the usefulness of specific critical factors, 25 success factors were identified.

The second phase is the specialist choice and data-gathering procedure. Professional logistics opinions were collected through interviews and questionnaires. First, 87 relevant companies and research institutions were identified in Seoul, Republic of Korea. Based on their web pages and logistics-related operations, 32 institutions were selected. These 32 institutions were then contacted, outlining the research goals and their importance. Within three weeks, 25 phone and email responses were received, of which nine specialists were unable to participate in the interviews due to commitments. In the end, 16 logistics industry specialists were selected to participate in the interview based on professional experience, industry influence, and familiarity with logistics innovation. They primarily have over 10 years of experience in logistics, supply chain management, or related fields, and possess professional knowledge in areas such as corporate management and technology development.

Table 2. Professionals’ demographic data.

Profile Information	Number of Respondents	Percentage (%)
Job position
Department Manager or above	8	50.00
Professor	6	37.50
Government Manager	2	12.50
Working experience in the company (years)
5–10	4	25.00
11–15	3	18.75
>16	9	56.25
Enterprise type
Logistics Company	2	12.50
Logistics Automation Equipment Manufacturing Company	3	18.75
Universities and Research Institutions	6	37.50
Government and Public Institutions	2	12.50
Other Technology Providers	3	18.75
Number of employees
<50	0	0.00
51–100	2	12.50
101–200	1	6.25
201–300	2	12.50
>301	11	68.75

Note. N = 16.

shows the professionals’ demographic data.

The data were collected during interviews through questionnaires and webinars from 2 July 2024 to 6 September 2024. An expert interview seminar was held on 3 September. During the meetings, the goals and significance of the research were explained in detail to the specialists. Subsequently, their opinions were gathered through anonymous questionnaires. Finally, through discussions, the specialists reached a consensus.

In the third phase, to form “systematic insights”, a fuzzy DEMATEL-AISM was established through expert scoring to help decision-makers understand the causal relationships between and hierarchy of various factors. The fuzzy DEMATEL method was used to analyze the mutual influence of factors in complex relationships, further categorizing the identified critical factors into causal groups. This method effectively reduces subjective bias and uncertainty by converting expert opinions into fuzzy numbers. At the same time, it clearly demonstrates the causal relationships and impact intensity between factors [38,39,40]. However, this method cannot identify factors’ hierarchical relationships. Therefore, the AISM method was employed to establish a hierarchy and causal relationships between factors. By introducing result priority and cause priority hierarchical extraction rules through the AISM, a set of antagonistic hierarchical topological maps is established. This method has proven advantageous in revealing the causal relationships in complex systems [41]. Compared to other methods, this approach can better explain the hierarchy of and logical relationships between factors affecting ILU by analyzing the degrees of influence, driving force, and dependency of factors.

Table 3. Recent research using fuzzy DEMATEL-AISM.

Serial Number	Author	Research Topic	Research Methods
1	[40]	Enablers to implement sustainable initiatives in agri-food supply chains	ISM-Fuzzy DEMATEL
2	[42]	Fuzzy DEMATEL method for developing supplier selection criteria	Fuzzy theory-Fuzzy DEMATEL
3	[39]	Using fuzzy DEMATEL to evaluate the green supply chain management practices	DEMATEL-Fuzzy set theory
4	[43]	A DEMATEL-based method for prioritizing strategic projects using the perspectives of the Balanced Scorecard	DEMATEL
5	[44]	Analysis of supply chain vulnerability factors in manufacturing enterprises: a fuzzy DEMATEL approach	Fuzzy DEMATEL
6	[45]	Identification of barriers and their mitigation strategies for Industry 5.0 implementation in emerging economies	DEMATEL
7	[46]	Industry 5.0 challenges for post-pandemic supply chain sustainability in an emerging economy	ISM-MICMAC
8	[41]	Influencing factors of novice pilot SA based on DEMATEL-AISM method: From pilots’ view	DEMATEL-AISM
9	[47]	An evaluation of factors influencing the vulnerability of emergency logistics supply chains	DEMATEL-AISM
10	[15]	Identifying CSFs for the agri-food cold chain’s sustainable development: When the strategy system comes into play	AISM-MICMAC
11	[33]	Key Success Factors for Export Structure Optimization in East Asian Countries Through Global Value Chain (GVC) Reorganization	DEMATEL-ISM

summarizes our recent research using fuzzy DEMATEL-AISM-MICMAC. Finally, critical factors with high potential for improving automated loading and unloading were identified. These critical factors guide leaders in developing effective strategies for implementing ILU management within USCs. Additionally, this research identifies and examines the CSFs affecting the practical execution of automated loading and unloading through correlation and structural analysis among these variables to meet the needs of academia and decision-makers.

Therefore, in summary, we believe that fuzzy DEMATEL-AISM technology can help to address the following issues:

To enhance the development and application of ILU, logistics and loading and unloading equipment-related companies should understand which field to channel into their overall management.

The combination of fuzzy DEMATEL and AISM can reveal the causal relationships between factors.

The causal diagram generated by fuzzy DEMATEL and the hierarchical diagram generated by AISM help in understanding the importance of various factors.

It is noteworthy that method deviation can affect reactions while gathering data [15]. The selected expert is technically very skilled, and the specialists are highly capable in the fields of supply chain planning and operations management. Notably, each expert’s responses are confidential, thereby preventing biases in the research results.

3.2. Establishment of the Fuzzy DEMATEL-AISM Method

This study’s core method is the introduction of the fuzzy DEMATEL-AISM [40]. By constructing a key factor identification model based on the fuzzy DEMATEL-AISM method, an adversarial hierarchical topology diagram is drawn to visually display the influence relationships between various factors. This model’s basic process is as follows. We can recapitulate the fuzzy DEMATEL-AISM through the subsequent actions (Figure 2 and Figure 3):

3.2.1. Fuzzy DEMATEL

1. Determine influencing factor indicators and establish the initial influence matrix $W$. Determine the indicators of influencing factors $F=\left\{F_1,F_2,\cdots ,F_n\right\}$. Using specialist scoring, construct the initial influence matrix $W={\left|w_{ij}\right|}_{n\times n}$, where $n$ is the matrix dimension and $W_{ij}$ represents the degree of influence of factor $F_i$ on $F_j$ (For more details, please refer to Appendix A).

2. Calculate the fuzzy direct influence matrix and obtain the direct influence matrix $Z$.

According to

Table 4. The fuzzy linguistic scale.

Linguistic Terms	Influence Score	Influence Score
No influence (N)	0	(0, 0, 0.25)
Very low influence (VL)	1	(0, 0.25, 0.50)
Low influence (L)	2	(0.25, 0.50, 0.75)
High influence (H)	3	(0.50, 0.75, 1.00)
High influence (H)	4	(0.75, 1.00, 1.00)

, the expert scoring results are converted into triangular fuzzy numbers (TFNs). By using the converting fuzzy data into crisp scores (CFCS) method to defuzzify the fuzzy direct influence matrix, the direct influence matrix is obtained.

Calculate the crisp value of the TFN after annotation processing.

$$z_{ij}^k=min{w}_{ij}^k+{\displaystyle \frac{{\mathsf{\Delta }}_{min}^{max}\left[min{u}_{ij}^k\left(1-u_{ij}^k\right)+n_{ij}^k{n}_{ij}^k\right]}{1-u_{ij}^k+n_{ij}^k}}$$

(1)

Obtain the direct influence matrix $Z$.

$$z_{ij}^k={\displaystyle \frac{z_{ij}^1+z_{ij}^2+\cdots +z_{ij}^k}{k}}$$

(2)

$$Z={\left|z_{ij}^k\right|}_{n\times n}$$

(3)

3. Standardization direct influence matrix $G$.

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

(4)

$$G=\eta Z$$

(5)

In the formula, $G$ represents the direct influence matrix after normalization and fuzzification.

4. Calculate the comprehensive influence matrix $T$.

$$T={\mathrm{l}\mathrm{i}\mathrm{m}}_{n\to \mathrm{\infty }}(G+G^2+G^3+\cdots G^n)=G(I-G{)}^{-1}$$

(6)

In the formula, $I$ is the unit matrix and $(I-G{)}^{-1}$ is the inverse matrix of $(I-G)$.

5. Calculate the influence degree $D_i$ and the affected degree $C_i$.

$$D_i=\sum _{j=1}^n t_{ij},j=\mathrm{1,2},\cdots ,n$$

(7)

$$C_i=\sum _{i=1}^n t_{ij},i=\mathrm{1,2},\cdots ,n$$

(8)

In the formula, $t_{ij}$ represents the degree of influence of the $i$-th element on the $j$-th element in the comprehensive influence matrix $T$. The influence degree $D_i$ and the af fected degree $C_i$ represent the influence of factor $F_i$ on other factors or the influence received from other factors.

Influence degree $D_i$ is the row sum of each factor and represents the comprehensive influence value of that row’s factors on other factors.

The affected degree $C_i$ is the column sum of each factor and represents the comprehensive influence value of that column’s factors on other factors.

6. Calculate centrality $V_i$ and causality $N_i$.

$$V_i=D_i+C_i,i=\mathrm{1,2},\cdots ,n$$

(9)

$$N_i=D_i-C_i,i=\mathrm{1,2},\cdots ,n$$

(10)

Centrality $V_i$ indicates the importance of factor $F_i$ in the system; the larger this value, the more important the factor.

The causality degree $N_i$ indicates the direct causal relationship between factors. Factors with $N_i$ > 0 are causal factors, while those with $N_i$ < 0 are result factors.

7. Draw a causal relationship diagram of influencing factors, with the horizontal axis representing centrality and the vertical axis representing causality.

3.2.2. AISM

1.

Calculate the adjacency matrix $A$.

Based on the input from the expert panel, construct the structural self-interaction matrix, then, using the rules in

Table 5. Transformation rules.

	Self-Interaction Matrix	Adjacency Matrix
i, j	V	1
j, i		0
i, j	A	0
j, i		1
i, j	X	1
j, i		1
i, j	O	0
j, i		0

[15], convert the structural self-interaction matrix into an adjacency matrix that includes factor relationships.

2.

Establish the reachability matrix $R$.

$$H=A+I$$

(11)

$$R=H^{k-1}\ne H^k=H^{k+1},k=\mathrm{1,2},\cdots ,n$$

(12)

In the formula, $H$ is the product matrix and $I$ is the unit matrix. According to the properties of Boolean matrix operations, use the multiplication method to calculate the reachability matrix $R$.

3.

Establish a general skeletal matrix $S$.

$$S=R^{\prime }-1$$

(13)

For convenience in calculations, the reachability matrix $R$ is processed via edge and node reduction, thereby retaining the strong connectivity relationships in the reachability matrix to establish the reduced matrix $R^{\prime }$. This, in turn, leads to the most simplified representation of the reachability matrix, namely the skeleton matrix $S$. Notably, the skeleton matrix of each reachability matrix is unique.

4.

Draw the AISM topology hierarchy diagram.

Considering the result priority rules and cause priority extraction rules, hierarchical extraction is performed to obtain the final adversarial hierarchical division results. Subsequently, based on the hierarchical results and the skeleton matrix, the AISM topology hierarchical diagram was drawn. Finally, the cross-impact matrix multiplication applied to classification (MICMAC) technique was utilized to explain the pushing and dependency capabilities of CSFs, verifying the validity of the AISM.

3.3. The Development of Surveys

A set of 25 CSFs are identified for improving ILU per the literature recommendations. Next, a questionnaire survey was used to discuss the CSF list for improving ILU with 16 specialists. In this survey, we assign each factor according to the specialists’ experience and opinion: 1—Not important, 2—Not very important, 3—Average, 4—Important, and 5—Very important. Among the 25 influencing factors, the average score of 6 indicators was below 3, while the average score of 19 indicators was above 3. Finally, we selected 19 CSFs from the workshop for the next research phase (

Table 6. Critical success factors.

CSFs	Definition	Author
Market research (F1)	Use market research to understand demand for ILU equipment, promote companies to improve products, and apply to logistics companies.	[5]
Operation training (F2)	Through workshops, seminars, and training, impart advantages and operational methods of ILU equipment to stakeholders and train employees to efficiently use and maintain automated systems.	[9,48]
Customization (F3)	Provide customized ILU solutions to meet the specific needs of different cities’ supply chains and logistics companies for various scales.	[25]
Equipment integration support (F4)	Design scalable devices and provide intelligent integration support to ensure compatibility with logistics processes and various types of vehicles, along with detailed installation and operation documentation.	[20]
Innovation improvement (F5)	Regularly update ILU technology, invest in the research and development of cutting-edge intelligent equipment, and continuously evaluate and improve processes to promote the application of equipment in logistics companies.	[43]
Equipment usability (F6)	Design intuitive and user-friendly ILU systems to lower the usage threshold and operational difficulty.	[49]
Environmental advantage (F7)	Showcase the advantages of the equipment in terms of environmental benefits, such as reducing pollution, optimizing energy efficiency, and lowering operating costs.	[46]
After-sales service (F8)	Provide robust after-sales support, including preventive maintenance, remote diagnostics, and on-site support, thereby ensuring equipment performance and stability and enhancing user confidence.	[50]
Stakeholder collaboration (F9)	Establish strategic partnerships with key players in the supply chain, design solutions to promote the adoption of smart loading and unloading equipment, and address the specific challenges of urban logistics.	[21,29]
Customer feedback mechanisms (F10)	Establish effective user feedback channels, continuously improve the technology and service levels of ILU equipment, and promptly respond to customer needs.	[17]
Performance control (F11)	Use real-time monitoring and data analysis to optimize the ILU process and track equipment performance to identify issues.	[40]
Government support (F12)	For logistics companies that adopt ILU equipment, provide financial subsidies, research funding, and infrastructure support to promote the development of ILU technology.	[40]
Reliability and safety (F13)	Introduce artificial intelligence in ILU equipment to optimize load sorting, space utilization, and route planning. Improve equipment reliability and safety and reduce downtime, workplace accidents, and equipment damage.	[42]
Scalability features (F14)	Ensure that ILU technology can scale its features according to business growth and the needs of ILU logistics companies.	[48]
Regulatory compliance (F15)	Ensure that ILU technologies comply with local and international regulations to avoid legal issues.	[26,29]
User-friendly design (F16)	Design an intuitive and simple ILU operation system to reduce the adoption barriers for users of the ILU logistics company.	[24]
Developing automation systems (F17)	Develop advanced automation systems, including robotic arms, autonomous mobile robots, and AGVs, to accelerate loading and unloading speeds, improve accuracy, and reduce manual labor and human errors in ILU logistics companies.	[43,45]
Multi-party collaboration platform (F18)	Deploy a multi-party collaboration cloud platform connecting logistics operators, drivers, and warehouse managers to achieve real-time coordination and the efficient utilization of urban zones.	[21]
Standardized packaging (F19)	Use standardized, modular packaging to optimize space in vehicles through automated loading and unloading.	[10,13]

). We then used text relationships similar to “causes” and asked the 16 specialists for external views on the interrelations of the 19 CSFs. All the specialists presented observations.

4. Results and Discussions

4.1. Fuzzy-DEMATEL Analysis

The initial influence matrix can be found in

Table 7. Initial influence matrix.

CSF	F1	F2	F3	F4	F5	F6	F7	F8	F9	F10	F11	F12	F13	F14	F15	F16	F17	F18	F19
F1	0	4	4	2	3	4	3	3	3	4	3	4	2	2	4	3	1	2	2
F2	2	0	1	1	2	2	1	1	2	1	4	1	3	2	3	4	4	1	3
F3	3	4	0	2	4	1	1	2	2	4	1	4	2	1	4	4	3	4	3
F4	0	2	2	0	3	1	0	2	4	1	2	3	2	4	1	2	2	4	2
F5	2	3	4	0	0	3	3	2	1	2	3	1	3	1	3	1	4	4	1
F6	2	4	1	3	3	0	3	4	2	2	2	1	1	2	2	4	1	4	2
F7	2	1	2	3	3	3	0	2	1	4	2	2	3	1	3	3	4	3	4
F8	2	4	3	3	4	3	2	0	4	1	2	3	3	3	2	3	2	1	3
F9	2	4	3	2	3	4	2	3	0	4	4	3	2	3	3	3	2	3	3
F10	0	1	1	3	4	1	2	3	1	0	1	4	0	2	3	1	1	4	2
F11	1	4	1	1	1	2	1	3	2	1	0	1	1	3	2	2	4	2	1
F12	3	3	4	2	4	3	2	4	4	3	3	0	2	4	4	4	4	3	2
F13	1	1	0	2	0	0	1	1	3	2	1	0	0	0	1	1	0	1	0
F14	1	2	1	3	1	3	3	4	2	4	2	4	2	0	1	2	2	4	4
F15	3	3	3	3	3	4	3	3	3	4	3	4	3	4	0	4	3	3	4
F16	3	1	1	3	3	2	2	2	3	1	3	1	1	2	2	0	1	4	1
F17	1	4	4	0	4	3	2	2	2	2	3	2	1	1	2	3	0	2	2
F18	1	2	2	2	3	2	1	2	2	3	2	3	0	2	2	2	2	0	1
F19	0	4	2	0	4	2	1	3	1	3	4	1	2	1	3	1	1	4	0

. Using Formulas (1)–(5), the standardized direct influence matrix for enhancing ILU within USCs was calculated, as shown in Appendix B (

Table A1. Standardization direct influence matrix $G$.

CSF	F1	F2	F3	F4	F5	F6	F7	F8	F9	F10	F11	F12	F13	F14	F15	F16	F17	F18	F19
F1	0.0000	0.9667	0.9667	0.5000	0.7333	0.9667	0.7333	0.7333	0.7333	0.9667	0.7333	0.9667	0.5000	0.5000	0.9667	0.7333	0.2667	0.5000	0.5000
F2	0.5000	0.0000	0.2667	0.2667	0.5000	0.5000	0.2667	0.2667	0.5000	0.2667	0.9667	0.2667	0.7333	0.5000	0.7333	0.9667	0.9667	0.2667	0.7333
F3	0.7333	0.9667	0.0000	0.5000	0.9667	0.2667	0.2667	0.5000	0.5000	0.9667	0.2667	0.9667	0.5000	0.2667	0.9667	0.9667	0.7333	0.9667	0.7333
F4	0.0333	0.5000	0.5000	0.0000	0.7333	0.2667	0.0333	0.5000	0.9667	0.2667	0.5000	0.7333	0.5000	0.9667	0.2667	0.5000	0.5000	0.9667	0.5000
F5	0.5000	0.7333	0.9667	0.0333	0.0000	0.7333	0.7333	0.5000	0.2667	0.5000	0.7333	0.2667	0.7333	0.2667	0.7333	0.2667	0.9667	0.9667	0.2667
F6	0.5000	0.9667	0.2667	0.7333	0.7333	0.0000	0.7333	0.9667	0.5000	0.5000	0.5000	0.2667	0.2667	0.5000	0.5000	0.9667	0.2667	0.9667	0.5000
F7	0.5000	0.2667	0.5000	0.7333	0.7333	0.7333	0.0000	0.5000	0.2667	0.9667	0.5000	0.5000	0.7333	0.2667	0.7333	0.7333	0.9667	0.7333	0.9667
F8	0.5000	0.9667	0.7333	0.7333	0.9667	0.7333	0.5000	0.0000	0.9667	0.2667	0.5000	0.7333	0.7333	0.7333	0.5000	0.7333	0.5000	0.2667	0.7333
F9	0.5000	0.9667	0.7333	0.5000	0.7333	0.9667	0.5000	0.7333	0.0000	0.9667	0.9667	0.7333	0.5000	0.7333	0.7333	0.7333	0.5000	0.7333	0.7333
F10	0.0333	0.2667	0.2667	0.7333	0.9667	0.2667	0.5000	0.7333	0.2667	0.0000	0.2667	0.9667	0.0333	0.5000	0.7333	0.2667	0.2667	0.9667	0.5000
F11	0.2667	0.9667	0.2667	0.2667	0.2667	0.5000	0.2667	0.7333	0.5000	0.2667	0.0000	0.2667	0.2667	0.7333	0.5000	0.5000	0.9667	0.5000	0.2667
F12	0.7333	0.7333	0.9667	0.5000	0.9667	0.7333	0.5000	0.9667	0.9667	0.7333	0.7333	0.0000	0.5000	0.9667	0.9667	0.9667	0.9667	0.7333	0.5000
F13	0.2667	0.2667	0.0333	0.5000	0.0333	0.0333	0.2667	0.2667	0.7333	0.5000	0.2667	0.0333	0.0000	0.0333	0.2667	0.2667	0.0333	0.2667	0.0333
F14	0.2667	0.5000	0.2667	0.7333	0.2667	0.7333	0.7333	0.9667	0.5000	0.9667	0.5000	0.9667	0.5000	0.0000	0.2667	0.5000	0.5000	0.9667	0.9667
F15	0.7333	0.7333	0.7333	0.7333	0.7333	0.9667	0.7333	0.7333	0.7333	0.9667	0.7333	0.9667	0.7333	0.9667	0.0000	0.9667	0.7333	0.7333	0.9667
F16	0.7333	0.2667	0.2667	0.7333	0.7333	0.5000	0.5000	0.5000	0.7333	0.2667	0.7333	0.2667	0.2667	0.5000	0.5000	0.0000	0.2667	0.9667	0.2667
F17	0.2667	0.9667	0.9667	0.0333	0.9667	0.7333	0.5000	0.5000	0.5000	0.5000	0.7333	0.5000	0.2667	0.2667	0.5000	0.7333	0.0000	0.5000	0.5000
F18	0.2667	0.5000	0.5000	0.5000	0.7333	0.5000	0.2667	0.5000	0.5000	0.7333	0.5000	0.7333	0.0333	0.5000	0.5000	0.5000	0.5000	0.0000	0.2667
F19	0.0333	0.9667	0.5000	0.0333	0.9667	0.5000	0.2667	0.7333	0.2667	0.7333	0.9667	0.2667	0.5000	0.2667	0.7333	0.2667	0.2667	0.9667	0.0000

).

Subsequently, the comprehensive influence matrix was obtained using Formula (6), as shown in Appendix B (

Table A2. Comprehensive influence matrix $T$.

CSF	F1	F2	F3	F4	F5	F6	F7	F8	F9	F10	F11	F12	F13	F14	F15	F16	F17	F18	F19
F1	0.0885	0.2052	0.1739	0.1349	0.1942	0.1850	0.1449	0.1777	0.1662	0.1902	0.1758	0.1814	0.1289	0.1442	0.1913	0.1821	0.1386	0.1812	0.1475
F2	0.0941	0.1017	0.0961	0.0872	0.1342	0.1198	0.0865	0.1092	0.1163	0.1068	0.1550	0.0987	0.1148	0.1097	0.1364	0.1556	0.1459	0.1217	0.1265
F3	0.1285	0.1929	0.1047	0.1240	0.1985	0.1310	0.1069	0.1509	0.1420	0.1794	0.1376	0.1719	0.1202	0.1196	0.1813	0.1841	0.1576	0.1984	0.1508
F4	0.0635	0.1337	0.1117	0.0693	0.1493	0.1042	0.0702	0.1243	0.1463	0.1078	0.1240	0.1311	0.0993	0.1408	0.1064	0.1250	0.1172	0.1669	0.1120
F5	0.1003	0.1581	0.1482	0.0794	0.1115	0.1411	0.1218	0.1306	0.1074	0.1312	0.1449	0.1084	0.1204	0.0999	0.1459	0.1212	0.1551	0.1737	0.1048
F6	0.1026	0.1759	0.1071	0.1293	0.1649	0.0985	0.1249	0.1653	0.1284	0.1332	0.1369	0.1126	0.0952	0.1220	0.1343	0.1687	0.1143	0.1808	0.1242
F7	0.1044	0.1380	0.1274	0.1315	0.1723	0.1495	0.0808	0.1417	0.1172	0.1692	0.1405	0.1319	0.1267	0.1096	0.1545	0.1575	0.1611	0.1736	0.1564
F8	0.1116	0.1914	0.1480	0.1367	0.1922	0.1576	0.1194	0.1154	0.1691	0.1320	0.1496	0.1519	0.1352	0.1461	0.1472	0.1670	0.1404	0.1510	0.1497
F9	0.1177	0.2026	0.1557	0.1314	0.1903	0.1818	0.1274	0.1750	0.1150	0.1861	0.1881	0.1631	0.1256	0.1560	0.1724	0.1776	0.1501	0.1926	0.1588
F10	0.0614	0.1138	0.0960	0.1154	0.1629	0.1015	0.0981	0.1358	0.0984	0.0859	0.1046	0.1440	0.0679	0.1095	0.1336	0.1068	0.1011	0.1638	0.1100
F11	0.0735	0.1550	0.0892	0.0816	0.1108	0.1123	0.0799	0.1300	0.1089	0.0980	0.0832	0.0932	0.0791	0.1184	0.1130	0.1191	0.1395	0.1253	0.0918
F12	0.1431	0.2035	0.1847	0.1418	0.2207	0.1817	0.1387	0.2027	0.1907	0.1861	0.1869	0.1294	0.1363	0.1819	0.2003	0.2067	0.1917	0.2078	0.1567
F13	0.0447	0.0624	0.0377	0.0662	0.0491	0.0417	0.0478	0.0587	0.0865	0.0734	0.0589	0.0417	0.0300	0.0394	0.0580	0.0595	0.0399	0.0646	0.0391
F14	0.0883	0.1513	0.1107	0.1325	0.1419	0.1489	0.1272	0.1716	0.1322	0.1680	0.1396	0.1613	0.1113	0.0927	0.1239	0.1425	0.1314	0.1863	0.1572
F15	0.1424	0.2039	0.1696	0.1584	0.2080	0.1963	0.1538	0.1907	0.1775	0.2022	0.1884	0.1921	0.1517	0.1831	0.1394	0.2072	0.1774	0.2110	0.1868
F16	0.1077	0.1176	0.0965	0.1179	0.1482	0.1192	0.1001	0.1230	0.1304	0.1068	0.1371	0.1011	0.0837	0.1107	0.1205	0.0911	0.1010	0.1654	0.0962
F17	0.0862	0.1727	0.1478	0.0781	0.1741	0.1409	0.1068	0.1306	0.1212	0.1290	0.1463	0.1211	0.0910	0.1003	0.1316	0.1500	0.0931	0.1456	0.1191
F18	0.0766	0.1291	0.1092	0.1002	0.1469	0.1156	0.0832	0.1206	0.1123	0.1327	0.1191	0.1283	0.0665	0.1082	0.1189	0.1222	0.1145	0.1002	0.0944
F19	0.0626	0.1613	0.1086	0.0708	0.1619	0.1163	0.0840	0.1357	0.0977	0.1335	0.1509	0.0978	0.0985	0.0932	0.1353	0.1085	0.1024	0.1627	0.0768

). Finally, the causal set for improving ILU factors was obtained using Formulas (7)–(10). Centrality indicates the importance of factors in the system; the larger this value, the more important the factor. So, the factors affecting ILU improvement are ranked based on centrality. Causality indicates the classification of causes and effects. Factors with causality > 0 are causal factors, while those with causality < 0 are result factors. Thus, based on causality, the factors are divided into causal groups (see

Table 8. Causal set.

CSF	Influence Degree	Affected Degree	Centrality	Causality	Ranking
F1	3.1317	1.7977	4.9294	1.3339	10
F2	2.2161	2.9701	5.1862	−0.7540	7
F3	2.8802	2.3228	5.2031	0.5574	6
F4	2.2032	2.0867	4.2898	0.1165	18
F5	2.4039	3.0316	5.4355	−0.6278	5
F6	2.5192	2.5429	5.0620	−0.0237	9
F7	2.6436	2.0024	4.6460	0.6411	16
F8	2.8115	2.6894	5.5008	0.1221	4
F9	3.0674	2.4637	5.5311	0.6037	3
F10	2.1103	2.6515	4.7618	−0.5412	14
F11	2.0018	2.6674	4.6692	−0.6656	15
F12	3.3914	2.4610	5.8524	0.9305	2
F13	0.9993	1.9824	2.9816	−0.9831	19
F14	2.6189	2.2852	4.9041	0.3338	12
F15	3.4400	2.6442	6.0843	0.7958	1
F16	2.1741	2.7523	4.9264	−0.5782	11
F17	2.3854	2.4723	4.8578	−0.0869	13
F18	2.0985	3.0724	5.1709	−0.9738	8
F19	2.1585	2.3589	4.5174	−0.2004	17

).

Finally, construct a coordinate system with centrality on the horizontal axis and causality on the vertical axis, deriving the causal relationship diagram shown in Figure 4. The causal relationship diagram provides valuable insights for analyzing the CSFs in enhancing ILU within a USC.

In the context of the logistics industry, the importance ranking of factors affecting ILU improvement was obtained using centrality results. The ranking is as follows:

F15 > F12 > F9 > F8 > F5 > F3 > F2 > F18 > F6> F1 > F16 > F14 > F17 > F10 > F11 > F7 > F19 > F4 > F13.

Based on the degree of causation, factors are divided into a cause group and an effect group. The causal factors are crucial as driving forces within the entire system and their performance will affect the achievement of overall goals [39]. As independent driving factors, they have a direct effect on the system and are of high priority [40]. Therefore, the factors in the cause group should receive more attention. The centrality ranking of the factors classified as causal group factors is as follows: F15 > F12 > F9 > F8 > F3 > F1 > F14 > F7 > F4. The order of causality is as follows: F1 > F12 > F15 > F7 > F9 > F3 > F14 > F8 > F4. Among them, F15 has the highest centrality and a relatively high causality, indicating that regulatory compliance is an extremely important factor [51]. This indicates that the promotion and application of ILU technologies must comply with local and international regulations to ensure their legality [48]. The second rank in terms of centrality and causality is F12. This indicates that the government has an extraordinary effect in promoting ILU development. F9 ranks third in centrality and fifth in reason, indicating that strategic cooperation and resource sharing among key participants in the supply chain are crucial for promoting ILU [11]. Moreover, with service competition intensification, the manufacturing industry faces greater challenges; thus, F8 is important in addressing product usage issues [21,52]. The diversity and complexity of loading and unloading products highlight the importance of F3 [14]. F1 is an important way to understand customer needs for ILU products [5]. In different loading and unloading tasks, F14 combined with various functional modules can effectively reduce operational complexity [48]. F7 helps companies to establish a good image, while also playing an important role in marketing equipment [53]. F4 helps to manage the operation of multiple equipment models in complex operational environments [24]. Ultimately, the cause group factors are considered the origin of the effect group factors. Emphasizing the cause group factors facilitates understanding their effect on the effect group factors.

Outcome factors in the system are easily influenced by other factors, and other factors in the system affect the entire system through these outcome factors [39,54]. This indicates that the outcome group can be improved, while the cause group can meet the effectiveness standards [39]. Therefore, the factors in the outcome group are considered feedback indicators of the system’s operational state. The centrality ranking of the factors classified as outcome group factors is as follows: F5 > F2 > F18 > F6 > F16 > F17 > F10 > F11 > F19 > F13. The order of causality is as follows: F6 > F17 > F19 > F10 > F16 > F5 > F11 > F2 > F18 > F13. The centrality and causality of F13 are both minimal, indicating that they are the most affected. All factors will affect reliability and safety, thereby affecting the application of ILU.

As shown in Figure 3, the first and second quadrants include nine causal factors. Among them, the factors in the first quadrant are core influencing factors with high centrality and causality. Additionally, they are closely related to other factors. The factors in the first quadrant include F1, F12, F15, F9, F3, and F8. The second quadrant consists of self-driven factors with lower centrality but close relationships, including F7, F14, and F4. Among all the cause groups, F15 has the highest centrality, indicating that regulatory compliance is more important to the entire system. F1 ranks the highest in terms of causality, meaning that F1 has a greater effect on the entire system. Furthermore, Table 8 shows that the influence of F1 ranks third among all factors causing causal relationships. Therefore, F1 has a significant effect on other factors.

The third and fourth quadrants include 10 outcome factors. The third quadrant consists of independent factors with lower importance and fewer effects, including F17, F19, F10, F16, F11, and F13. The fourth quadrant contains quite prominent influencing factors, but the connections between these factors are relatively weak and include F6, F5, F2, and F18. Among them, F5 ranks high in centrality (fifth) with an influence slightly less than 0, indicating that it is less affected by other factors but has a significant effect on the system.

4.2. AISM Analysis

Based on input from the specialist panel, the structural self-interaction matrix is constructed and notations to record specialist judgment are applied to represent the interaction orientation among two CSFs (for instance, i and j). The notations used have the following implications:

V: CSF i causes CSF j;

A: CSF j causes CSF i;

X: CSF i and CSF j are mutually causal;

O: CSF i and CSF j are not related.

Table 9. Structural self-interaction matrix.

CSF	F19	F18	F17	F16	F15	F14	F13	F12	F11	F10	F9	F8	F7	F6	F5	F4	F3	F2
F1	O	V	V	V	O	O	O	O	O	O	A	O	O	O	O	V	V	O
F2	A	A	A	A	O	V	O	O	V	X	O	X	V	V	V	A	A
F3	X	X	X	X	O	O	O	O	V	V	O	O	O	O	O	X
F4	X	X	X	X	O	O	O	O	V	V	O	O	O	O	O
F5	O	O	O	O	O	X	V	O	O	O	O	A	X	X
F6	O	O	O	O	O	X	V	O	X	A	O	A	X
F7	A	O	O	O	O	X	V	O	X	A	O	A
F8	A	A	A	A	O	V	O	O	V	X	O
F9	V	V	V	V	A	O	O	A	O	O
F10	A	A	A	A	O	V	O	O	V
F11	O	O	O	O	O	X	V	O
F12	O	O	O	O	X	O	O
F13	O	O	O	O	O	A
F14	O	O	O	O	O
F15	O	O	O	O
F16	X	X	X
F17	X	X
F18	X

shows the structural self-interaction matrix developed to enhance ILU.

According to the rules in Table 5, convert the structural self-interaction matrix into an adjacency matrix that contains factor relationships. See Appendix B,

Table A3. Adjacency matrix.

CSF	F1	F2	F3	F4	F5	F6	F7	F8	F9	F10	F11	F12	F13	F14	F15	F16	F17	F18	F19
F1	1	0	1	1	0	0	0	0	0	0	0	0	0	0	0	1	1	1	0
F2	0	1	0	0	1	0	1	1	0	0	1	0	0	1	0	0	0	0	0
F3	0	1	1	1	0	0	0	0	0	1	1	0	0	0	0	1	1	1	1
F4	0	1	1	1	0	0	0	0	0	1	1	0	0	0	0	1	1	1	1
F5	0	0	0	0	1	0	1	0	0	0	0	0	1	1	0	0	0	0	0
F6	0	0	0	0	0	1	0	0	0	0	0	0	1	1	0	0	0	0	0
F7	0	0	0	0	1	0	1	0	0	0	1	0	1	1	0	0	0	0	0
F8	0	1	0	0	1	1	1	1	0	1	1	0	0	1	0	0	0	0	0
F9	1	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	0	1
F10	0	0	0	0	0	1	1	1	0	1	1	0	0	1	0	0	0	0	0
F11	0	0	0	0	0	0	1	0	0	0	1	0	1	1	0	0	0	0	0
F12	0	0	0	0	0	0	0	0	1	0	0	1	0	0	1	0	0	0	0
F13	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0
F14	0	0	0	0	1	1	1	0	0	0	1	0	1	1	0	0	0	0	0
F15	0	0	0	0	0	0	0	0	1	0	0	1	0	0	1	0	0	0	0
F16	0	1	1	1	0	0	0	1	0	1	0	0	0	0	0	1	1	1	1
F17	0	1	1	1	0	0	0	1	0	1	0	0	0	0	0	1	1	1	1
F18	0	1	1	1	0	0	0	1	0	1	0	0	0	0	0	1	1	1	1
F19	0	1	1	1	0	0	1	1	0	1	0	0	0	0	0	1	1	1	1

.

Use Formulas (11)–(12) to calculate the final reachable matrix, as shown in

Table 10. Reachability matrix.

CSF	F1	F2	F3	F4	F5	F6	F7	F8	F9	F10	F11	F12	F13	F14	F15	F16	F17	F18	F19	DRP
F1	1	1	1	1	1	1	1	1	0	1	1	0	1	1	0	1	1	1	1	16
F2	0	1	0	0	1	1	1	1	0	1	1	0	1	1	0	0	0	0	0	9
F3	0	1	1	1	1	1	1	1	0	1	1	0	1	1	0	1	1	1	1	15
F4	0	1	1	1	1	1	1	1	0	1	1	0	1	1	0	1	1	1	1	15
F5	0	0	0	0	1	1	1	0	0	0	1	0	1	1	0	0	0	0	0	6
F6	0	0	0	0	1	1	1	0	0	0	1	0	1	1	0	0	0	0	0	6
F7	0	0	0	0	1	1	1	0	0	0	1	0	1	1	0	0	0	0	0	6
F8	0	1	0	0	1	1	1	1	0	1	1	0	1	1	0	0	0	0	0	9
F9	1	1	1	1	1	1	1	1	1	1	1	0	1	1	0	1	1	1	1	17
F10	0	1	0	0	1	1	1	1	0	1	1	0	1	1	0	0	0	0	0	9
F11	0	0	0	0	1	1	1	0	0	0	1	0	1	1	0	0	0	0	0	6
F12	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	19
F13	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	1
F14	0	0	0	0	1	1	1	0	0	0	1	0	1	1	0	0	0	0	0	6
F15	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	19
F16	0	1	1	1	1	1	1	1	0	1	1	0	1	1	0	1	1	1	1	15
F17	0	1	1	1	1	1	1	1	0	1	1	0	1	1	0	1	1	1	1	15
F18	0	1	1	1	1	1	1	1	0	1	1	0	1	1	0	1	1	1	1	15
F19	0	1	1	1	1	1	1	1	0	1	1	0	1	1	0	1	1	1	1	15
DEP	4	13	10	10	18	18	18	13	3	13	18	2	19	18	2	10	10	10	10

Abbreviations: DEP, dependence power; DRP, driving power.

. This matrix outlines the transitivity between factors for improving ILU and evaluates all the other factors’ transitivity [40]. Subsequently, the general skeleton matrix is calculated using Formula (13) (see Appendix B,

Table A4. General skeleton matrix.

CSF	F1	F2	F3	F4	F5	F6	F7	F8	F9	F10	F11	F12	F13	F14	F15	F16	F17	F18	F19
F1	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
F2	0	0	0	0	1	0	0	1	0	0	0	0	0	0	0	0	0	0	0
F3	0	1	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
F4	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0
F5	0	0	0	0	0	1	0	0	0	0	0	0	1	0	0	0	0	0	0
F6	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0
F7	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0
F8	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	0
F9	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
F10	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
F11	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0
F12	0	0	0	0	0	0	0	0	1	0	0	0	0	0	1	0	0	0	0
F13	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
F14	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0
F15	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0
F16	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0
F17	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0
F18	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1
F19	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0

). For the reachable set, cause set, and common set, see Appendix B,

Table A5. Reachable set, Cause set, and Common set.

CSF	Reachable Set	Cause Set	Common set
F1	1, 2, 3, 4, 5, 6, 7, 8, 10, 11, 13, 14, 16, 17, 18, 19	1, 9, 12, 15	1
F2	2, 5, 6, 7, 8, 10, 11, 13, 14	1, 2, 3, 4, 8, 9, 10, 12, 15, 16, 17, 18, 19	2, 8, 10
F3	2, 3, 4, 5, 6, 7, 8, 10, 11, 13, 14, 16, 17, 18, 19	1, 3, 4, 9, 12, 15, 16, 17, 18, 19	3, 4, 16, 17, 18, 19
F4	2, 3, 4, 5, 6, 7, 8, 10, 11, 13, 14, 16, 17, 18, 19	1, 3, 4, 9, 12, 15, 16, 17, 18, 19	3, 4, 16, 17, 18, 19
F5	5, 6, 7, 11, 13, 14	1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 14, 15, 16, 17, 18, 19	5, 6, 7, 11, 14
F6	5, 6, 7, 11, 13, 14	1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 14, 15, 16, 17, 18, 19	5, 6, 7, 11, 14
F7	5, 6, 7, 11, 13, 14	1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 14, 15, 16, 17, 18, 19	5, 6, 7, 11, 14
F8	2, 5, 6, 7, 8, 10, 11, 13, 14	1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 14, 15, 16, 17, 18, 19	2, 8, 10
F9	1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 13, 14, 16, 17, 18, 19	9, 12, 15	9
F10	2, 5, 6, 7, 8, 10, 11, 13, 14	1, 2, 3, 4, 8, 9, 10, 12, 15, 16, 17, 18, 19	2, 8, 10
F11	5, 6, 7, 11, 13, 14	1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 14, 15, 16, 17, 18, 19	5, 6, 7, 11, 14
F12	1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19	12, 15	12, 15
F13	13	13	13
F14	5, 6, 7, 11, 13, 14	1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 14, 15, 16, 17, 18, 19	5, 6, 7, 11, 14
F15	1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19	12, 15	12, 15
F16	2, 3, 4, 5, 6, 7, 8, 10, 11, 13, 14, 16, 17, 18, 19	1, 3, 4, 9, 12, 15, 16, 17, 18, 19	3, 4, 16, 17, 18, 19
F17	2, 3, 4, 5, 6, 7, 8, 10, 11, 13, 14, 16, 17, 18, 19	1, 3, 4, 9, 12, 15, 16, 17, 18, 19	3, 4, 16, 17, 18, 19
F18	2, 3, 4, 5, 6, 7, 8, 10, 11, 13, 14, 16, 17, 18, 19	1, 3, 4, 9, 12, 15, 16, 17, 18, 19	3, 4, 16, 17, 18, 19
F19	2, 3, 4, 5, 6, 7, 8, 10, 11, 13, 14, 16, 17, 18, 19	1, 3, 4, 9, 12, 15, 16, 17, 18, 19	3, 4, 16, 17, 18, 19

. By analyzing the overlap and distribution patterns of these sets, key nodes and their roles in the system can be identified, providing a basis for further optimization and decision-making.

Finally, based on the final extraction results (

Table 11. Final extraction results.

Level	Results of Type UP	Results of Type DOWN
1	13	12, 15
2	5, 6, 7, 11, 14	9
3	2, 8, 10	1
4	3, 4, 16, 17, 18, 19	3, 4, 16, 17, 18, 19
5	1	2, 8, 10
6	9	5, 6, 7, 11, 14
7	12, 15	13

), the topological hierarchy diagrams for UP type (Figure 5) and DOWN type (Figure 6) are drawn. Orientation route segments represent the reachability connection among ILU elements, while double arrows indicate mutual reachability. The topological diagram reveals that the seven levels display the progressive causal relationships and action paths between various factors [41]. There are three loops among the influencing factors, indicating that the influencing factors within these three loops are mutually causal and closely interconnected.

Based on the results of the hierarchical diagram, the system is divided into seven levels. Then, based on the final reachability matrix and the final hierarchy of critical factors, a hierarchical structure model for improving the critical ILU factors has been constructed (Figure 4 and Figure 5). Among them, F12 and F15 are located at the causative layer L7, serving as the core driving forces for system implementation. This indicates that government support and regulatory compliance help promote F9. F1 helps the government to understand the demand for improving ILU, thereby formulating supportive policies, and highlights the importance of enhancing ILU within USCs. It also means that government support and regulatory compliance have a major effect on the entire system, and that the effect permeates the entire system and is not easily affected by other factors. Therefore, these four factors are the key elements affecting ILU improvement within USCs.

L2–L6, as important connecting factors, form the transitional factor layer, which is influenced by the causal layer factors and also affects the factors of other layers. As shown in Figure 4 and Figure 5, the system’s transitional factor layer consists of 16 factors. Among them, the factors in L2, L3, and L4 are in a strong correlation loop, indicating that there is a strong interconnection and mutual causality between the factors at this level. Therefore, F3, F4, F16, F17, F18, and F19 are crucial for the improvement of ILU technology and resource-sharing decisions. Additionally, through F2, F8, and F10, user experience and quality feedback can be effectively enhanced. F11 has driven the F5 of F6 and F14 and enhanced F7 by optimizing energy efficiency.

The top-level factor is the outcome layer, which has a direct effect. Therefore, F13 has the most direct effect on improving ILU within USCs. To promote ILU development, one can begin with reliability and safety. Since this factor is easily influenced by other factors, it is important to manage its antecedent factors.

Through the reachability matrix, MICMAC analysis can be conducted to verify the element classification in the AISM method based on dependence and driving power. MICMAC can help to identify key driving factors and dependent factors, clarifying the position and role of each factor within the system. The driving–dependence matrix is used to analyze the CSFs affecting ILU improvement, and the CSFs are divided into four regions: Autonomous (I), Dependent (II), Connected (III), and Independent (IV), as shown in Figure 7.

The first quadrant is the autonomous quadrant with a low driving force and low dependence. The factors within this quadrant are located near the origin in the diagram. Their weak association with other factors has limited effects on the entire system. However, MICMAC analysis indicates that the factors in the autonomous quadrant may be sources of system instability and variability. Notably, no critical factors were identified in this area in this study. This indicates that all factors within the system have a certain degree of interdependence and have varying effects on the entire system.

The second quadrant is the dependency quadrant with a low driving force and high dependency, which is the desired outcome. Research has found that factors F2, F5, F6, F7, F8, F10, F11, F13, and F14 occupy the top level of the structural model and show strong dependency on other factors. Therefore, these factors must rely on other factors to reduce limitations in the ILU improvement process and to play a greater role.

The third quadrant is the connected quadrant with high drive and dependence. Factors within this quadrant, F3, F4, F16, F17, F18, and F19, have a bidirectional effect on the system. They strongly affect other factors and are also easily affected by them. Factors in this quadrant exhibit significant instability. Any intervention involving these factors will affect other factors and feedback on them. Therefore, the factors within this quadrant are both system drivers and dependent on other factors [11].

The fourth quadrant is the independent quadrant with a high driving force and low dependence. Factors including F1, F9, F12, and F15 belong to the independent quadrant. These factors form the base of the structural model and are considered the main driving forces for ILU enhancement.

The MICMAC analysis of critical factors and hierarchy found that the driving and dependent forces come from F12 and F15, and have the lowest dependency and highest driving force (Table 10). F9 also has low dependency and a high driving force. Therefore, F9, F12, and F15 are positioned at the top of the hierarchy and are considered independent driving factors.

To verify the stability of the model, we conducted a sensitivity analysis by increasing the influence of F3 by 10% and 20% and decreasing it by 10% and 20% based on the original influence matrix. The changes in impact degree are shown in

Table 12. Change in impact degree (%).

CSF	Decreasing Influence 20%	Decreasing Influence 10%	Increase Influence 10%	Increase Influence by 20%
F1	−3.6	−1.96	2.17	4.67
F2	−2.78	−1.53	1.62	3.56
F3	−2.14	−1.1	1.07	3.02
F4	−2.78	−1.53	1.59	3.48
F5	−2.78	−1.53	1.59	3.48
F6	−1.87	−1.05	1	2.93
F7	−2.36	−1.51	1.31	3.2
F8	−2.74	−1.51	1.62	3.75
F9	−2.14	−1.1	1.07	3.02
F10	−3.02	−1.66	1.68	3.71
F11	−2.57	−1.3	1.23	3.35
F12	−3.18	−1.92	1.93	4.05
F13	−1.07	−0.78	0.78	2.28
F14	−2	−1.11	1.07	2.87
F15	−2	−1.61	1.78	1
F16	−3.21	−1.76	1.91	4.16
F17	−2.99	−1.63	1.69	3.69
F18	−2.99	−1.63	1.69	3.69
F19	−2.99	−1.63	1.69	3.69

. The analysis results indicate that when the impact degree on F3 increases or decreases, the overall impact degree of the system also increases or decreases accordingly (as shown in Figure 8). This indicates the important role of F3 in the system. Moreover, the varying degrees of response of different factors to the influence of F3 indicate that F3 has a high driving force and high dependency. This analysis indicates that the model of this study remains stable under different parameter settings, verifying its reliability.

F12 plays a crucial role in promoting ILU development in the logistics industry, a view consistent with prior research [6]. This study suggests that government support is a crucial prerequisite for enhancing the development of ILU [22]. Simultaneously, government support positively affects the standardization of corporate behavior and investment activities [55]. Government funding support affects ILU operational methods and further determines the deployment of systems and machines [6]. The government promotes technological innovation through market incentives (such as providing financial rewards, research funding, and infrastructure investment) [56]. Moreover, government support is beneficial in regulating and promoting logistics equipment companies’ demand for F17 and F3 to carry out F5. This includes research and development in areas such as F6, F13, and the device F14 [49,55], thus increasing ILU levels within USCs.

F15 is among the CSFs for enhancing ILU implementation. When enterprises introduce ILU technologies, regulations are an important tool to address uncertainty in the management process [53]. Consistent with other research findings, digitalization and intelligent regulations are important external driving forces for promoting the adoption of ILU technologies by enterprises [56]. Establishing architecture regulations and programs helps companies to increase their legitimacy with external stakeholders [28]. Moreover, regulations related to employee safety can drive the normalization and standardization of operational training (F2), becoming a core element in reducing management costs during the ILU process [26]. As the attention paid by society to environmental protection increases, compliance with environmental regulations will positively affect a company’s reputation and performance, while also providing the company with an environmental advantage (F7) [56].

F9 is also among the CSFs for enhancing ILU. The strategies and challenges of stakeholders mutually influence and constrain one another [53]. Stakeholder collaboration encourages key participants in the supply chain to establish strategic cooperative relationships, with F18 as the basis for cooperation [49]. F12 guides corporate behavior through F15; production and operations are made transparent. It also ensures compliance in data sharing among enterprises in supply chain collaboration through effective communication and trust production, and operations are made transparent [15]. Additionally, F1 and the addition of F10 help in accurately understanding customer needs. F11 helps to identify and resolve equipment issues during use, which aids in optimizing ILU equipment and systems. Additionally, using F19 during the loading and unloading process helps to optimize space and improve loading and unloading efficiency [54]. Through F9, the development and application of ILU do not only improve the overall efficiency of the supply chain, but also enhance the long-term cooperation willingness between enterprises, laying the foundation for the widespread application of ILU.

5. Conclusions

5.1. Theoretical Contributions

The contributions of this paper are multifaceted and timely. In terms of theoretical contributions, this study first enriches the knowledge system in the field of USCs’ loading and unloading management by elucidating the CSFs that promote ILU during USCs. This differs from previous research that has focused on issues related to the costs and benefits of the intelligent unloading process. This study explores the factors promoting the intelligence of USCs’ ILU from a systemic perspective, addressing the gaps in the existing literature. For example, in the context of intelligence, the research reveals the theoretical significance of corporate stakeholder collaboration, government support, and regulatory compliance as core driving factors, providing new support for the theoretical framework of intelligent loading and unloading. Furthermore, by combining fuzzy DEMATEL, AISM, and MICMAC methods, a systematic study was conducted from multiple dimensions on the CSFs for promoting intelligent loading and unloading in urban supply chains. This provides a new theoretical perspective for enhancing ILU in USCs.

Secondly, this study introduces intelligent technology into loading and unloading management, promoting the development of intelligent logistics management theory, for example, improvements in automated loading and unloading equipment and human collaboration models. Finally, this study explores the facilitation system of ILU from the perspective of intelligent logistics management, with a focus on analyzing the reasons for its application in the delivery phase, supporting the coordinated development of ILU efficiency and technology.

5.2. Management Contributions

Our findings provide valuable suggestions for managers in management and policy fields. Above all, logistics companies must recognize that ILU plays an important role in improving smart logistics efficiency. Additionally, to promote ILU within USCs, logistics companies must establish multi-party collaboration platforms to facilitate effective information sharing, thereby accelerating loading and unloading speed and reducing human error. Additionally, logistics company managers should attach value to employee training when adopting new technologies and equipment. By regularly conducting operational training, they can standardize employee operations and reduce the losses caused by operational errors. During the loading and unloading process, logistics companies should use standardized packaging to optimize loading and unloading space. Additionally, to avoid legal issues during loading and unloading, logistics companies should strictly adhere to local laws and regulations.

Secondly, logistics equipment manufacturing companies must understand user needs through market research and other methods when innovating and developing new equipment. Based on supply chain characteristics and requirements in different cities, they should provide flexible, customized solutions and equipment integration support. For example, by referring to the automated loading and unloading modes of advanced logistics centers, optimizing equipment design, and improving operational efficiency. By innovating, improving, and developing automated systems that meet the business needs of logistics companies, they can improve loading and unloading efficiency. Furthermore, considering that logistics business information may involve privacy issues, equipment companies should focus on compliance with relevant regulations. From the perspective of equipment usage, logistics equipment manufacturing companies should establish a comprehensive customer feedback mechanism and learn from the user experience optimization strategies of successful companies to continuously optimize product performance to enhance their users’ experience. Furthermore, logistics equipment manufacturing companies should design more usable products and provide strong after-sales service to customers, thereby enhancing user satisfaction by reducing equipment use difficulties.

Finally, government support plays a crucial role in enhancing the goals of ILU within USCs. First, the government should improve policies and regulations, clearly defining the development goals and technical standards for ILU and providing logistics companies with clear policy guidance. Second, the government should increase financial investment in infrastructure construction to provide a solid foundation for the efficient operation of ILU. Furthermore, it should promote the opening of the ILU equipment market and encourage logistics equipment manufacturing companies to develop more advanced ILU technologies and equipment. Finally, governments must promote cooperation among stakeholders in the logistics industry, thereby creating favorable market environments to drive the application of ILU within USCs.

6. Limitations and Future Research

This study has certain limitations. First, the fuzzy DEMATEL and AISM methods rely on expert judgment, so they cannot completely eliminate subjective limitations. Secondly, the research is primarily based on the urban supply chain delivery environment in South Korea. Although the research conclusions have a certain degree of universality, their applicability in other regions still needs further verification. Additionally, the sample size is relatively small, which may limit the generalizability of the research findings when applied on a larger scale. Finally, although the IDT, RBV theory, and ST applied in this study provide valuable perspectives, they may not adequately consider dynamic external environments and complex stakeholder cooperation.

With the rapid development of urban logistics innovation and technology, factors that currently have a significant effect may become secondary in the future. Additionally, some factors may become more important as ILU operations mature. Therefore, future research could consider collecting data from different time periods to analyze the importance of various factors, the reasons for changes in causal relationships, and trends. This will help enterprises and decision-makers to adjust their strategies and decisions in a timely manner based on industry development trends, thereby better promoting the development of intelligent loading and unloading. Additionally, that is why future investigators should consider more influencing factors and expand the application of relevant theories, for example, using artificial intelligence to optimize loading and unloading routes, implementing equipment monitoring, interconnection, and data sharing through the Internet of Things, and using blockchain technology to enhance data security and logistics transparency. Concurrently, the further exploration of improvements and extensions of the fuzzy DEMATEL and AISM methods could combine emerging technologies to develop more precise analytical models. Additionally, further research could test the applicability of the findings in specific supply chain environments, such as emerging market cities or fresh food cold chain logistics, to assess the feasibility of their promotion and further enrich the research in the intelligent logistics field.