Evaluation of the Barriers to Maintenance 4.0 for the Textile Industry via Pythagorean Fuzzy SWARA

Abstract

Maintenance 4.0 studies have become a focus for managers and employees when developing effective and efficient maintenance policies. In this study, the barriers to Maintenance 4.0 applications in the textile industry are investigated, and these barriers are weighted using the Stepwise Weight Assessment Ratio Analysis (SWARA) method based on Pythagorean fuzzy numbers. Solutions to address these barriers are presented. As a result of this study, Organizational and Managerial emerged as the most important main criterion. Operational was identified as the second most significant main criterion, followed by Technical Competence. Data-Related and Cybersecurity ranked fourth in terms of importance. On the other hand, Human Resources and Training and Financial were found to be the least important main criteria. These two criteria received lower importance scores compared to the others, with Financial being the criterion with the lowest overall significance. Sensitivity analyses were performed for six different scenarios by changing the importance weights of the decision-makers. The ranking of the criteria only slightly changed with the weights; this means that the results obtained in Case 1 are robust and reliable. Even in Case 6, where the expert weight ratios were completely reversed, the results did not change significantly. This highlights an important point regarding the reliability of the assessment.

1. Introduction

Maintenance 4.0, as an important part of the digital transformation in the textile industry, plays a critical role in increasing the efficiency and sustainability of production processes. Thanks to the integration of technologies such as the Internet of Things (IoT), big data analytics, and artificial intelligence, machine failures can be predicted in advance, and unplanned downtimes are minimized. This both reduces production costs and increases competitiveness by increasing quality. It also contributes to environmental sustainability, with positive effects on energy consumption and resource use. In a sector such as textiles, which requires intensive machine use, Maintenance 4.0 applications offer a strategic advantage in terms of production continuity and business performance.

The concepts of sustainability and digitalization are encountered in the manufacturing industry, as in every field. The optimum use of limited resources has become one of the most fundamental policies of nations, countries, and companies. While businesses develop production and management policies that consider sustainability in terms of environmental and social impacts and sustainability, they are trying to lay the foundations of Industry 4.0 with digitalization [1,2]. Industry 4.0 has entered our lives with the application of digital technologies such as the Internet of Things (IoT), cyber–physical systems (CPS), artificial intelligence (AI), virtual reality (VR), augmented reality (AR), and big data analysis (BDA) in the manufacturing sector [3,4,5].

Ensuring the continuity of production activities without interruptions or delays is of critical importance. The advent of Industry 4.0 technologies has necessitated the adoption of innovative approaches to maintenance operations. Maintenance 4.0 aims to enable failure prediction, optimize resource utilization, and establish continuously improving maintenance systems [6,7]. Survival in a highly competitive environment is closely linked to the uninterrupted operation of production processes. An effective maintenance policy enables cost-effective production and enhances customer satisfaction without compromising quality or quantity [8]. Maintenance 4.0 applications—where physical systems are synchronized with the virtual world through the integration of computers and technological devices—are a subset of Industry 4.0. These applications involve automated data collection, data analysis, and intelligent decision-making systems [1,9]. In today’s volatile and uncertain market environment, implementing a technically and administratively sound maintenance plan and management strategy is a highly complex process for businesses. Such a long-term strategic plan supports businesses in addressing the economic, environmental, and social dimensions underpinning the principle of sustainability [10,11,12].

The periodic maintenance of machines is a critical factor across all industries. It ensures the optimal functioning of machinery and equipment, minimizes the risk of unexpected downtime and emergency repairs, reduces workplace accident risks, extends the lifespan of assets, and ensures compliance with safety and environmental standards. These goals are achieved through regular inspections, preventive and corrective actions, part replacements, and adjustments [13,14]. A substantial contribution to operational excellence can be achieved through the strategic integration of various maintenance approaches, including corrective, preventive, predictive, autonomous maintenance, and focused improvement initiatives. Corrective maintenance refers to the maintenance activities undertaken when a machine, component, equipment, or system experiences a malfunction. Its primary objective is to restore the affected asset to its normal operational condition. As the machine or system ceases to operate, it necessitates immediate intervention, often leading to increased maintenance and downtime-related costs. The affected equipment remains non-operational until the underlying issue is fully resolved [15]. Preventive maintenance is a planned maintenance strategy carried out at certain intervals to keep machinery and equipment in an optimum working condition in order to prevent failures and to minimize the risk of breakdown time. Corrective actions and precautions are taken by focusing on potential problems before failure occurs. With this maintenance strategy, less idle time is spent, costs are lower, and equipment reliability is increased [16]. Predictive maintenance is a maintenance strategy based on continuous monitoring of the conditions and performance of equipment and machinery and determining when maintenance is necessary. It is the identification of signs of wear or abnormalities before failures occur using data collected with the help of devices such as oil, vibration analysis, sensors, etc. Predictive maintenance allows for maintenance planning, when necessary, thus avoiding unnecessary interventions. This means preventing unnecessary stops and excessive interventions, thus helping to minimize costs [17]. Autonomous maintenance is a maintenance strategy that involves machine operators in the process. In this maintenance strategy, operators are directly involved in the maintenance and repair of machines and are responsible for the operation of the equipment. Operators are responsible for the regular cleaning, lubrication, simple adjustments, and maintenance of machines and equipment. It is an approach that increases operational efficiency and reduces machine idle time [18]. Specific improvement is the creation of a specific improvement strategy by creating multifunctional teams to eliminate chronic machinery and equipment problems and increase performance. It focuses on developing effective solutions that aim to obtain to the root causes of problems. Specific improvement is the activities that create a culture of cooperation within the organization, eliminate chronic problems, and extend equipment life span [14].

To implement different maintenance policies within the company, some data needs to be collected. For Maintenance 4.0 applications, data collected from multiple data sources, such as sensors and the Internet of Things, are stored in cloud services and analyzed using artificial neural networks. As a result of these analyses, a Maintenance 4.0 policy is developed within the company [19].

Compared to traditional methods, the use of linguistic assessments in the fuzzy approach enhances its reliability by better capturing the nuances of expert judgment. Among various fuzzy techniques, the Pythagorean Fuzzy method yields more precise and sensitive results, which contributes to its growing preference in contemporary decision-making applications. Moreover, its inclusion among recent methodological advancements reinforces its suitability for addressing complex evaluation problems.

Importantly, this method incorporates the three fundamental components of uncertainty: membership, non-membership, and hesitation degrees. To improve the realism and robustness of the hesitation measurement, it calculates hesitation by squaring the membership and non-membership values. This enhances the reliability of the derived scores and strengthens the overall score function, allowing for more accurate interpretations of expert opinions.

Additionally, the method supports more dependable sensitivity analysis by offering a solid mathematical foundation for evaluating the stability of results under varying expert weights. Its straightforward implementation, even in problems involving numerous criteria, further contributes to its selection. The combination of analytical rigor and practical simplicity makes the Pythagorean Fuzzy SWARA method both a reliable and user-friendly tool for multi-criteria decision-making.

Although businesses believe that Maintenance 4.0 applications are necessary and beneficial, it is obvious that the implementation rates of these policies are very low. Some barriers are encountered during the implementation of these policies. This study particularly addresses the barriers encountered in Maintenance 4.0 applications in the textile industry, which is the one of the largest industries in Turkey [20]. The main reason for choosing the textile industry is that it is a labor-intensive industry. The fact that workers are not highly qualified and that workers are fatigued due to long working hours increases the potential for workers to make mistakes, which inevitably results in damage to machines. In the second part of this study, a literature review was conducted on the subject, while in the third part, the identified barriers and the method used were introduced. In the fourth part, its application was performed in textile factories operating in Istanbul, and in the fifth part, the results obtained from the application were evaluated. In the final part, the difficulties encountered in this study and the potential for future studies were discussed.

2. Literature Review

In this section, many studies conducted in the literature within the scope of Maintenance 4.0 and the method used are summarized.

According to the search results made with the keywords “Maintenance 4.0” and “barriers” in the Web of Science database on 4 May 2025, four articles were found. Naik and Kirkire conducted a study to identify the barriers to the implementation of Maintenance 4.0 in manufacturing industries, categorize these identified barriers, rank the barriers according to their priorities, and suggest the digital technologies required to overcome these barriers [1]. James and his colleagues also conducted a study to identify and analyze the challenges faced in implementing the Maintenance 4.0 concept in industries. They collected the opinions of academics and industry experts through a survey study. As a result of their study, “technician skill challenges”, “cybersecurity challenges”, and “IoT application challenges” were determined to be the primary challenges [13]. El Kihel and his colleagues have worked to create a methodology for the implementation of Maintenance 4.0 with the data obtained as a result of digitalization and digitization. The main purpose of this methodology is to minimize downtime, optimize energy consumption, and increase the utilization rate and life of machines. As a result of this methodology, difficult workloads and risks are reduced by technologies, such as the digitalization of maintenance documents (procedures, risk prevention, machine history) and the rapid detection of machine failures [19]. In their study conducted in 2024, Samadhiya and his colleagues provided the integration of Industry 4.0 to ensure sustainability in total productive maintenance applications. It aimed to determine and prioritize the main drivers of total productive maintenance. As a result, it was determined that Industry 4.0 came first among the main drivers, followed by senior management support and the formal Industry 4.0 adoption program [21].

Since very few articles were reached because of the determined keywords, the word barrier was removed from the keywords, and a new search was conducted only for the keyword Maintenance 4.0. 77 studies were conducted on Maintenance 4.0 between 2016 and 2025. The distribution of publications by year is given in Figure 1. Of these publications, 40 are articles, 3 are early access, 1 is a book chapter, 32 are proceeding papers, 8 are review articles and 1 is editorial material.

Benhamza Hlihel and his colleagues developed a Maintenance 4.0 competency model that can be applied to all sectors in their study conducted in 2024. The Maintenance 4.0 competency model has three hierarchical levels: managers, engineers, and technicians. The determined competencies are organized according to four categories: technical, personal, social, and methodological. As a result of the study, basic competencies were determined for Maintenance 4.0 stakeholders [23].

In a study conducted on the railway industry, data on the operating conditions of a high-speed bogie were collected and analyzed using Industry 4.0 tools. The data collected allowed for the creation of indicators corresponding to abnormalities in a train’s behavior [24]. In their study, Di Nardo and his colleagues developed a fixed maintenance program using Industry 4.0 tools such as artificial intelligence, machine learning, big data analytics, the Internet of Things, and predictive analytics. This research contributes to the development of the Maintenance 4.0 concept by closing the gap between theory and practice [25]. Another study investigated the integration of Maintenance 4.0 in heating, ventilation, and air conditioning (HVAC) systems. With the help of tools such as artificial intelligence and the Internet of Things, it identified the obstacles and difficulties encountered in Maintenance 4.0 applications in HVAC systems [26]. In a study conducted in 2020, smart and sustainable maintenance was addressed from three different perspectives. The first of these is the historical perspective, which addresses the evolution of the maintenance approach. The second is the development perspective, which provides historical perspectives on maintenance data and data-driven maintenance technology. The last is the maintenance technology perspective, which addresses the dimensions of sustainable development and potential opportunities together [27]. In a study conducted on the chemical fertilizer industry, the Best Worst Method and Proximity Index Value methods were used in the selection of predictive maintenance suppliers to digitalize predictive maintenance. The weights of the supplier criteria determined by a literature review were calculated with BWM, and the final ranking of the suppliers was based on the overall proximity values obtained using the PIV method [28].

Although there are studies on Maintenance 4.0 applications, it is obvious that the subject is still new and not sufficiently studied. It is important to develop and implement models for Maintenance 4.0 applications, especially for different industries.

It is important to determine the barriers that may be encountered in Maintenance 4.0 applications in businesses and their importance levels. At this point, businesses faced with the decision-making process should analyze which criteria they will focus on and how they will propose a solution to these criteria. Human beings are in a decision-making process at every step of their lives. Sometimes they make efforts to determine the right choices by using instinctive and sometimes scientific methods. Multi-criteria decision-making (MCDM) problems and solution methods are also among the subjects that researchers often work on. Fuzzy sets are also used to eliminate the uncertainties experienced especially in the decision-making process.

There are many methods developed to determine the criteria by which a decision will be made in decision-making problems and the relative importance levels of these criteria. One of these methods is the SWARA method. As a result of the search made with the keywords “SWARA” and “Pythagorean” in the Web of Science database on 10 May 2025, 39 studies were found. All of these studies are articles published in different scientific journals. The distribution of publications obtained according to these keywords by year is shown in Figure 2.

In 2024, Pandey and Khurana used the hybrid SWARA–COmplex Proportional Assessment (COPRAS) method based on Pythagorean fuzzy numbers. In the study, solutions were developed to determine the risks encountered in Industry 4.0 applications and to reduce these risks. To determine the importance levels of the identified risks, the SWARA method based on Pythagorean fuzzy numbers was used, and after determining the weights of the risks, the solution suggestions developed using the COPRAS method based on Pythagorean fuzzy numbers were ranked [30]. Rani et al. addressed the solar panel selection problem and used the SWARA-VlseKriterijumska Optimizcija i Kaompromisno Resenje (VIKOR) method based on Pythagorean fuzzy numbers to solve the problem. While the SWARA method was used to determine the subjective weights of the evaluation criteria, the VIKOR method was preferred to rank the alternatives [31]. Another use of MCDM methods in different fields is in the evaluation of smart city waste management techniques. In a study conducted in India, the objective and subjective weights of the criteria to be used in the evaluation of smart city waste management techniques were determined using the MEthod based on the Removal Effects of Criteria (MEREC)–SWARA method based on Pythagorean fuzzy numbers. The Additive Ratio Assessment (ARAS) method was used in the ranking of alternatives [32]. Studies on the subject are summarized in

Table 1. Summary of Pythagorean fuzzy SWARA studies.

Author (s)	Area	Method	Aim
Çağlar et al. [33]	Agriculture	SWARA based on Pythagorean Fuzzy Sets	To compare the results obtained from Pythagorean Fuzzy SWARA with the results from the formula developed by Wischmeier and Smith
Kumar and Mahanta [34]	Energy	Fuzzy Pythagorean MEREC-SWARA-VIKOR	To propose a matrix form-based Pythagorean fuzzy metric and apply it solar panel selection problem
Saeidi et al. [35]	Manufacturing	Extendend Pythagorean fuzzy SWARA-TOPSIS	To prioritize sustainable human resource management factors, and assess the alternatives
Alrasheedi et al. [36]	Manufacturing	Pythagorean fuzzy entropy-SWARA-WASPAS	To evaluate sustainable suppliers by different sustainable factors
Cui et al. [37]	Manufacturing	Pythagorean fuzzy SWARA-CoCoSo	To weight the degree of basic barriers and rank organizations in different sectors by these barriers
Kamali Saraji et al. [38]	Energy	Pythagorean fuzzy SWARA-TOPSIS	To determine criteria and rank EU countries
Yahya et al. [39]	Digital economy	Pythagorean fuzzy SWARA-DNMA (double normalization-based multiple aggregation)	To define potential barriers for cloud resource utilization
Ramya et al. [40]	Energy	Pythagorean fuzzy SWARA-WASPAS	To determine the best thermal energy storage technique
Khoshsepehr and Alinejad [41]	Industrial organizations	Pythagorean fuzzy SWARA-CoCoSo	To evaluate the organizations by antifragility factors
He at al. [42]	Tourism	An extended interval-valued Pythagorean fuzzy SWARA-MULTIMOORA	To investigate sustainable community-based tourism

.

3. Materials and Methods

In this section, the challenges encountered in the implementation of Maintenance 4.0 and the SWARA based on the Pythagorean fuzzy sets used to determine the relative importance levels of these challenges will be explained.

3.1. Barriers to Maintenance 4.0

Based on the literature review and expert opinions, it was concluded that there are different barriers to Maintenance 4.0 applications in different industry branches. The barriers to Maintenance 4.0 applications in the textile industry are summarized in

Table 2. Categorization of barriers.

Code	Barrier Name (Sub-Criteria)	Main Criteria
B1	Lack of top management commitment	Organizational and Managerial
B2	Need for organizational restructuring of maintenance-related functions	Organizational and Managerial
B3	Lack of collaboration between departments	Organizational and Managerial
B4	Lack of digital culture and resistance to change	Organizational and Managerial
B5	Absence of a clear strategic roadmap	Organizational and Managerial
B6	Lack of strategy for AI and Industry 4.0 adoption	Organizational and Managerial
B7	Lack of managerial competencies and limited resources	Organizational and Managerial
B8	Low maturity level of enabling technologies	Technical Competence
B9	Insufficient physical and digital IT infrastructure	Technical Competence
B10	Difficulty in system and software integration (interoperability issues)	Technical Competence
B11	Lack of sensors/Low accuracy/High power consumption	Technical Competence
B12	Integration challenges between IT and OT systems	Technical Competence
B13	Issues in data quality and quantity	Data-Related and Cybersecurity
B14	Inability to perform real-time monitoring	Data-Related and Cybersecurity
B15	Data security and privacy concerns	Data-Related and Cybersecurity
B16	Complexity in big data management and analytics	Data-Related and Cybersecurity
B17	Lack of digital literacy among staff	Data-Related and Cybersecurity
B18	Risks in information security and data sharing (internal and external)	Data-Related and Cybersecurity
B19	Lack of technical skills among maintenance staff	Human Resources and Training
B20	Lack of structured training programs for capacity building	Human Resources and Training
B21	Skill shortages in the labor market	Human Resources and Training
B22	Lack of cultural readiness for AI and digital transformation	Human Resources and Training
B23	High initial capital investment	Financial
B24	Uncertainty regarding economic returns	Financial
B25	Lack of governmental and institutional support	Financial
B26	Legacy equipment incompatible with digitalization	Operational
B27	Difficulties in process control and performance optimization	Operational
B28	Scheduling and planning optimization challenges	Operational
B29	Challenges in estimating remaining useful life	Operational
B30	Need for accurate predictive maintenance models	Operational
B31	Lack of a structured framework specific to Maintenance 4.0 implementation	Operational

.

3.1.1. Organizational and Managerial Main Criteria

• Lack of top management commitment: One of the barriers encountered in integrating Maintenance 4.0 applications into a company is the lack of support from top management. Maintenance 4.0 means integrating new technologies and tools such as artificial intelligence, the IoT, machine learning, and data mining into the business. For this integration to be implemented successfully, top management must provide the necessary strategic support and sufficient budget, leading the entire team in this transformation process. Otherwise, Maintenance 4.0 integration should not be expected to be successful [1,8,43,44].
• Need for organizational restructuring of maintenance-related functions: Maintenance 4.0 transformation should be perceived not only as a technological transformation, but also as an organizational transformation process within the business. If the traditional organizational structure is continued, it may not be possible to successfully implement Maintenance 4.0 applications. Therefore, the duties and responsibilities of maintenance teams should be redefined, a new organizational structure should be established, new job descriptions should be made (such as data analytics, digital monitoring), a connection should be established between maintenance teams and information technology teams, and horizontal and flexible organizational structures should be developed [1,8,13,43,45].
• Lack of collaboration between departments: In a successful Maintenance 4.0 application, good cooperation should be established between departments. What is required for Maintenance 4.0 is that not only maintenance departments, but also the other departments of the business use technologies such as IoT, artificial intelligence, big data, cloud computing. If the connection between departments is weak or absent, data sharing will be limited, necessary and sufficient data may not be obtained and analyzed from sensors, cybersecurity and software infrastructures may be inadequate, and the digital integration of processes cannot be achieved. This causes Maintenance 4.0 applications to slow down or fail [13,23,46].
• Lack of digital culture and resistance to change: For Maintenance 4.0 applications to be successful, employees within the company must adopt a digital culture and be open to continuous learning. When an innovation is integrated into the company, the employees’ adoption of this innovation and their lack of resistance will be a factor affecting the success of the new application [1,47].
• Absence of a clear strategic roadmap: For a successful Maintenance 4.0 implementation, a clear strategic roadmap must be created within the company. If such a roadmap cannot be created, confusion and uncertainty will occur, resources will be wasted, and integration problems will occur. With a proper roadmap, resources will be transferred correctly, senior management and employee participation will increase, and changes in management will be more effective [1,13,48].
• Lack of strategy for AI and Industry 4.0 adoption: Integration of artificial intelligence and Industry 4.0 technologies is very important in Maintenance 4.0 applications. For this, a strategy should be determined within the company around this issue. If this strategy is not determined correctly, technological investments will be unnecessary, artificial intelligence applications will not be integrated into maintenance operations, employees will remain in uncertainty, and long-term gains will not be achieved. However, with a correctly determined strategy, it will become clear how the processes will be digitalized, that data infrastructure and sensor systems are planned correctly, and that employees are motivated and involved in the process [1,49].
• Lack of managerial competencies and limited resources: Maintenance 4.0 applications are a digital transformation process, and managers are expected to be proficient in digital transformation and Industry 4.0 applications during this process. However, in many businesses, managers are not competent in managing issues such as data literacy, artificial intelligence strategies, and risk management. These deficiencies lead to the incorrect construction of the maintenance planning strategy, the incorrect selection of technologies, and employee resistance. In addition, the use of technologies, such as sensors, data collection and processing, artificial intelligence algorithms, and digital twins, to be used in the Maintenance 4.0 application process are quite costly and require qualified labor. It is difficult to overcome these barriers, especially in small- and medium-sized businesses (SMEs) [23,50].

3.1.2. Technical Competence Main Criteria

• Low maturity level of enabling technologies: New technologies, such as IoT, sensors, big data analytics, artificial intelligence and machine learning, cyber–physical systems, digital twins, and edge and cloud computing, are still-developing technologies with low maturity levels. Low maturity levels negatively affect reliability and stability. Early adoption of these technologies, which have high investment costs, can cause both material and moral losses to companies [1,13,23].
• Insufficient physical and digital IT infrastructure: For Maintenance 4.0 applications to be successful, continuous data collection, analysis, prediction, and automation systems must be established. This requires a strong physical and digital information technology infrastructure [1,13,51].
• Difficulty in system and software integration (interoperability issues): For the implementation of Maintenance 4.0 to be successful, many different systems must exchange data in real-time and work in collaboration. Each of these systems may have different data formats, protocols, API structures, and communication standards. This causes integration problems [1,13,51].
• Lack of sensors/low accuracy/high power consumption: These three factors are among the elements that affect the success of Maintenance 4.0. In the event of insufficient sensors, predictive maintenance cannot be performed since sufficient data cannot be collected. In addition, low-quality sensors result in incorrect or incomplete data. This causes artificial intelligence and analysis systems to make incorrect decisions, and maintenance planning is incorrect. The use of wireless sensors results in very frequent battery replacement, which leads to additional costs and operational burdens [19,46].
• Integration challenges between IT and OT systems: Information technologies (IT) (software, data management, network infrastructure, data security, and digitalization) and operational technologies (OT) (physical machines, production devices, sensors, and industrial system management) are important to the success of Maintenance 4.0 policies. Weak or missing integration can lead to inefficient data collection, analysis, and maintenance processes [1,52].

3.1.3. Data-Related and Cybersecurity Main Criteria

• Issues in data quality and quantity: Maintenance applications will also be successful if accurate, complete, and reliable data is provided. With a successful maintenance application, production continuity and the efficient operation of the machines are ensured [53].
• Inability to perform real-time monitoring: Real-time monitoring is an important criterion for predictive maintenance 4.0 applications. Predictive analytical maintenance models are developed by analyzing the data coming from IoT sensors in real-time. In this way, maintenance costs are also minimized [2].
• Data security and privacy concerns: Data security and privacy are key factors for the continuity and success of Maintenance 4.0 implementations. Therefore, cybersecurity is an integral part of maintenance strategies [53].
• Complexity in big data management and analytics: To implement Maintenance 4.0 policies flawlessly, a large and diverse amount of data needs to be collected and analyzed. For this, a problem-free storage and access service must be provided. In addition, data analysis tools must be compatible. Inaccurate and unreliable data analysis leads to incorrect maintenance estimates [54].
• Lack of digital literacy among staff: Increasing the digital literacy of personnel within the business is important for the success of Maintenance 4.0 applications. Good digital literacy is required for the use of new technologies and the analysis and interpretation of the data obtained [55].
• Risks in information security and data sharing (internal and external): This is a criterion that is often overlooked, but which is very important when it comes to sharing data with both internal and external stakeholders. The fact that the system is so digital and that all devices are connected and communicate with each other underlines the need for intensive security measures [1].

3.1.4. Human Resources and Training Main Criteria

• Lack of technical skills among maintenance staff: Maintenance 4.0 is about the digitalization and automation of industrial machines, systems, and devices. Therefore, technical personnel need to have high digital skills as well as technical knowledge [1,23].
• Lack of structured training programs for capacity building: Maintenance 4.0 focuses not only on technological infrastructure, but also on the integration of human resources into this technology. The absence of structured training programs will cause personnel to not be able to use these digital systems effectively [23].
• Skill shortages in the labor market: Shortages in the labor market are a factor affecting the success of Maintenance 4.0 implementation. A qualified workforce is needed for successful implementation [56].
• Lack of cultural readiness for AI and digital transformation: Maintenance 4.0 applications do not only mean technological change and transformation, but also the formation of a new digital culture within the organization. In organizations without a digital transformation culture, employees cannot be enthusiastic about producing and adopting new ideas [1,13,23].

3.1.5. Financial Main Criteria

• High initial capital investment: The main criterion that prevents the adoption of Maintenance 4.0 is the high initial capital investment. Integrating advanced technologies, such as sensors, IoT devices, big data infrastructures, artificial intelligence-based software, and digital twins, required for Maintenance 4.0, are activities with high licensing, training, and operating costs [1,43].
• Uncertainty regarding economic returns: Uncertainty around the return rate is a significant barrier for Maintenance 4.0 applications. The advantages offered by Maintenance 4.0 emerge in the long term. In addition, the performance of the system depends entirely on the correct collection and analysis of data. Uncertainty in the realization of the economic benefits of Maintenance 4.0 poses a barrier for businesses [1,43].
• Lack of governmental and institutional support: Lack of government and institutional support is a systemic barrier to the spread of Maintenance 4.0 applications. Because the Maintenance 4.0 technological transformation process is quite costly, it requires multi-dimensional external support mechanisms such as financial incentives, infrastructure investments, educational support, and the establishment of standards [13,46].

3.1.6. Operational Main Criteria

• Legacy equipment incompatible with digitalization: One of the barriers to Maintenance 4.0 applications is old equipment and machines that are not compatible with digitalization. The technological tools used to collect data in Maintenance 4.0 applications are very difficult to integrate into old equipment and machines [1,13].
• Difficulties in process control and performance optimization: The success of Maintenance 4.0 applications is based on stable and optimized processes. If the process is unstable and not optimized, it is not possible for digital maintenance systems to work efficiently and correctly [1,13].
• Scheduling and planning optimization challenges: Difficulties in optimizing scheduling and planning directly negatively impact the goal of Maintenance 4.0 to transition from reactive to proactive and strategic maintenance. Therefore, not only technology, but also operational processes need to be digitalized and optimized [13,46].
• Challenges in estimating remaining useful life: The use of continuously updated algorithms with high-quality data to accurately estimate remaining useful life is essential for the success of Maintenance 4.0 applications [53].
• Need for accurate predictive maintenance models: This is an important barrier because cost reduction, business continuity, and efficiency increases, which are among the main goals of Maintenance 4.0, can be achieved with correct predictive maintenance models [53].
• Lack of a structured framework specific to Maintenance 4.0 implementation: To effectively implement Maintenance 4.0 applications, a structured framework is needed for a proper digital transformation process, the integration of technologies, and performance monitoring and continuous improvement processes [13].

3.2. Pythagorean Fuzzy SWARA Method

The Pythagorean Fuzzy SWARA method is designed to minimize decision-makers’ errors by utilizing linguistic variables. It employs membership (μ), non-membership (ν), and indeterminacy (π) degrees to represent uncertainty in expert evaluations. Occasionally, the sum of these assessments may exceed 1, which is a limitation addressed by the intuitionistic fuzzy method. To overcome this, the method calculates the indeterminacy by squaring the values. It remains a widely used and contemporary method in the literature.

The SWARA method, on the other hand, is based on ranking the criteria according to their importance. Unlike Analytic Hierarchy Process (AHP), it does not include pairwise comparisons or consistency ratio calculations.

The Pythagorean Fuzzy SWARA method incorporates the degree of hesitation, which is a limitation of the traditional SWARA approach. While IFS address part of this issue by including membership and non-membership degrees, they fall short when the sum of these values, along with hesitation, exceeds one. In such cases, PFS offer a more flexible and appropriate modeling framework. From the perspective of SWARA, when dealing with a large number of criteria, pairwise comparisons—as used in methods such as AHP—become extremely time-consuming and difficult to maintain consistently. Although the AHP provides a consistency check mechanism for up to 15 elements, the complexity increases significantly beyond this threshold. Therefore, SWARA is considered to be a more efficient method for weighting in multi-criteria decision-making problems compared to AHP, especially in situations involving a large set of criteria.

The concept of Pythagorean fuzzy sets (PFS), first proposed by Yager in 2013, is an advanced extension of both classical and intuitionistic fuzzy sets. PFS presents more available to conduct intuitionistic fuzzy structure. Membership degree ${\mathsf{\mu }}_{\mathrm{A}}\left(\mathrm{x}\right)$ and non-membership degree ${\mathsf{\vartheta }}_{\mathrm{A}}\left(\mathrm{x}\right)$ reflect each element x for a universe expression. Since their squared sum cannot sometimes surpass 1, this situation is significantly restricted. The formal definition of PFS is provided in Equation (1) [57,58].

$${\left({\mathsf{\mu }}_{\mathrm{A}}\left(\mathrm{x}\right)\right)}^2+{\left({\mathsf{\vartheta }}_{\mathrm{A}}\left(\mathrm{x}\right)\right)}^2\le 1$$

(1)

The structure of Pythagorean fuzzy sets (PFS) is illustrated in Equation (2).

$$\mathrm{P}=\left\{<\mathrm{x},\left({\mathrm{\mu }}_{\mathrm{P}}\left(\mathrm{x}\right),{\mathcal{V}}_{\mathrm{P}}\left(\mathrm{x}\right)\right)>:\mathrm{x}\in \mathrm{X}\right\}$$

(2)

The level of hesitation is displayed in Equation (3) [59]. The hesitation value represents the indeterminate portion that lies outside the membership and non-membership degrees. As the clarity of information increases, the level of hesitation—or indecisiveness—decreases.

$${\mathsf{\pi }}_{\mathrm{p}}\left(\mathrm{x}\right)=\sqrt{1-{\left({\mathsf{\mu }}_{\mathrm{A}}\left(\mathrm{x}\right)\right)}^2-{\left({\mathsf{\vartheta }}_{\mathrm{A}}\left(\mathrm{x}\right)\right)}^2},\mathrm{x}\in \mathrm{X}$$

(3)

The score and accuracy functions of α are defined in Equation (3). Given that S(α) lies within the interval [−1, 1], an enhanced score function for the PFN is subsequently proposed. The score function determines the direction of the evaluation by calculating the difference between the degrees of positive and negative assessments. The accuracy function quantifies the overall reliability of the evaluation by providing a numerical measure of certainty.

$$\mathrm{S}\left(\mathsf{\alpha }\right)={\left({\mathsf{\mu }}_{\mathsf{\alpha }}\right)}^2-{\left({\mathsf{\vartheta }}_{\mathsf{\alpha }}\right)}^2, \mathrm{K}\left(\mathsf{\alpha }\right)={\left({\mathsf{\mu }}_{\mathsf{\alpha }}\right)}^2+{\left({\mathsf{\vartheta }}_{\mathsf{\alpha }}\right)}^2 \mathrm{S}\left(\mathsf{\alpha }\right)\in \left[-\mathrm{1,1}\right],\mathrm{K}\left(\mathsf{\alpha }\right)\in \left[\mathrm{0,1}\right],$$

(4)

Normalized score and ambiguous function are displayed in Equation (5) [57]. To enhance interpretability, both the score and accuracy functions are normalized. Through normalization, the score function is rescaled to the interval [0, 1]. Similarly, by normalizing the accuracy function, the degree of uncertainty can be derived.

$${\mathrm{S}}^{\mathrm{*}}\left(\mathsf{\alpha }\right)={\displaystyle \frac{1}{2}}(\mathrm{S}\left(\mathsf{\alpha }\right)+1), {\mathrm{K}}^{\mathrm{*}}\left(\mathsf{\alpha }\right)=1-\mathrm{K}\left(\mathsf{\alpha }\right) \mathrm{S}\left(\mathsf{\alpha }\right),\mathrm{K}\left(\mathsf{\alpha }\right)\in \left[0,1\right],$$

(5)

There are n experts. Experts’ weights are calculated in Equation (6) [57]. These functions are also utilized to determine the weight of each expert in the decision-making process. The weights are calculated by taking into account the membership, non-membership, and hesitation values, thereby reflecting the consistency and decisiveness of expert judgments.

$${\mathrm{w}}_{\mathrm{m}}={\displaystyle \frac{\left({\mathsf{\mu }}_{\mathrm{m}}^2+{\mathsf{\pi }}_{\mathrm{m}}^2\ast \frac{\left({\mathsf{\mu }}_{\mathrm{m}}^2\right)}{\left({\mathsf{\mu }}_{\mathrm{m}}^2+{\mathsf{\vartheta }}_{\mathrm{m}}^2\right)}\right)}{\sum _{\mathrm{m}=1}^{\mathrm{n}}\left({\mathsf{\mu }}_{\mathrm{m}}^2+{\mathsf{\pi }}_{\mathrm{m}}^2\ast \frac{\left({\mathsf{\mu }}_{\mathrm{m}}^2\right)}{\left({\mathsf{\mu }}_{\mathrm{m}}^2+{\mathsf{\vartheta }}_{\mathrm{m}}^2\right)}\right)}}$$

(6)

The linguistic terms and their corresponding scales used in the evaluation of the experts are presented in

Table 3. Linguistic terms for experts.

Linguistic Terms	Brief	Pythagorean Fuzzy Numbers
Thrillingly Significant	THS	0.9	0.1
Typically Significant	TS	0.8	0.2
Noteworthy	N	0.6	0.4
Reasonable	R	0.5	0.5
Inconsequential	IC	0.45	0.55
Trivial	TR	0.3	0.75
Pitty	P	0.1	0.9

[31]. Table 3 presents the linguistic evaluation scale used to assess the importance weights of the experts.

The linguistic terms and scales used for evaluating the criteria are given in

Table 4. Linguistic terms for criteria.

Linguistic Terms	Brief	Pythagorean Fuzzy Numbers
Extremely Small	ES	0	1
Very Small	VS	0.1	0.9
Small	S	0.2	0.8
Slightly Small	SS	0.3	0.7
Below Intermediate	BI	0.4	0.6
Intermediate	I	0.5	0.5
Above Intermediate	AI	0.6	0.4
Slightly Big	SB	0.7	0.3
Big	B	0.8	0.2
Very Big	VB	0.9	0.1
Extremely Big	EB	1	0

[31]. Table 4 reflects the linguistic evaluation scale employed for evaluating the importance weights of the criteria.

PFWA and PFWG are described in Equations (7) and (8) [57]. The Pythagorean Fuzzy Weighted Average (PFWA) and Pythagorean Fuzzy Weighted Geometric (PFWG) operators transform the opinions of multiple decision-makers into a single aggregated value based on their respective weights.

$$\mathrm{P}\mathrm{F}\mathrm{W}\mathrm{A}\left({\mathrm{P}}_1,{\mathrm{P}}_2,\dots ,{\mathrm{P}}_{\mathrm{n}}\right)=(\sqrt{(1-\prod _{\mathrm{i}=1}^{\mathrm{n}}{{(1-\left({\mathsf{\mu }}_{\mathrm{i}}\right)}^2)}^{{\mathrm{w}}_{\mathrm{i}}}}, \prod _{\mathrm{i}=1}^{\mathrm{n}}{\left({\mathsf{\vartheta }}_{\mathrm{i}}\right)}^{{\mathrm{w}}_{\mathrm{i}}})$$

(7)

$$\mathrm{P}\mathrm{F}\mathrm{W}\mathrm{G}\left({\mathrm{P}}_1,{\mathrm{P}}_2,\dots ,{\mathrm{P}}_{\mathrm{n}}\right)=(\prod _{\mathrm{i}=1}^{\mathrm{n}}{\left({\mathsf{\vartheta }}_{\mathrm{i}}\right)}^{{\mathrm{w}}_{\mathrm{i}}},\sqrt{(1-\prod _{\mathrm{i}=1}^{\mathrm{n}}{{(1-\left({\mathsf{\mu }}_{\mathrm{i}}\right)}^2)}^{{\mathrm{w}}_{\mathrm{i}}}})$$

(8)

The SWARA method is implemented to determine importance degree of the criteria [60].

The normalized score function (NSF) (${\mathrm{s}}_{\mathrm{j}}$) is calculated with Equation (9) [61]. Normalization of data is applied to facilitate comparability, ensuring that all values are confined within the [0, 1] interval. To determine the comparative significance of each score value, each normalized score function value was subtracted from the value that comes after it.

$${\mathrm{s}}_{\mathrm{j}}={\displaystyle \frac{1}{2}}(1+{\mathsf{\mu }}_{\mathrm{j}}-{\mathsf{\vartheta }}_{\mathrm{j}})$$

(9)

The relative coefficient (${\mathrm{k}}_{\mathrm{j}}$) is obtained with Equation (10). The relative coefficient is calculated to identify the differences in significance among the criteria.

$${\mathrm{k}}_{\mathrm{j}}=\left\{\begin{array}{cc}1,& \mathrm{m}=1\\ ({\mathrm{s}}_{\mathrm{j}}+1),& \mathrm{m}>1\end{array}\right.$$

(10)

The significance weight (${\mathrm{q}}_{\mathrm{j}}$) is obtained using Equation (11). This procedure enhances the consistency of the importance ranking derived from expert evaluations.

$${\mathrm{q}}_{\mathrm{j}}=\left\{\begin{array}{cc}1,& \mathrm{j}=1\\ {\displaystyle \frac{{\mathrm{q}}_{(\mathrm{j}-1)}}{{\mathrm{k}}_{\mathrm{j}}}},& \mathrm{j}>1\end{array}\right.$$

(11)

The final significant weight (${\mathrm{w}}_{\mathrm{j}}$) is determined by Equation (12). Normalization is also utilized to compute the final significance weights of the criteria.

$${\mathrm{w}}_{\mathrm{j}}={\displaystyle \frac{{\mathrm{q}}_{\mathrm{j}}}{\sum _{\mathrm{i}=1}^{\mathrm{m}}{\mathrm{q}}_{\mathrm{j}}}}$$

(12)

The linguistic terms and their corresponding scales used in the evaluation of the experts are presented in Table 3. The linguistic terms and scales used for evaluating the criteria are given in Table 4.

4. Determination of Barriers in the Textile Industry

Initially, experts with relevant experience in Maintenance 4.0 were selected to participate in the evaluation process. Preference was given to individuals with demonstrable expertise in the field. A summary profile of these experts is presented in Table 1.

Subsequently, a comprehensive literature review was carried out, supported by preliminary expert insights, to identify the criteria to be included in the study. The barriers associated with Maintenance 4.0 were determined based on this dual input from both the literature and expert opinions.

In the third phase, the Pythagorean Fuzzy SWARA method, which is capable of addressing the dimensions of truth, falsity, and indeterminacy independently, was employed. Following this, the linguistic variables used for the evaluations were clearly defined.

Decision-makers then assessed the criteria using these predefined linguistic terms. Drawing on their expertise and professional judgment, they evaluated the importance and impact of each criterion. This linguistic-based approach facilitates the standardization and comparability of inherently subjective assessments. To enhance the robustness of the analysis and reduce bias, additional validation techniques, such as sensitivity analysis, were employed to ensure the reliability of the assigned weights.

Next, individual evaluations were aggregated based on expert-specific weightings and then were subsequently transformed into a single representative value using normalization procedures.

Finally, the deneutrosophication process was conducted to translate the neutrosophic evaluations into crisp numerical values. The overall methodological framework of the study is depicted in Figure 3.

Subsequently, the crisp values were obtained. The score function was calculated and then normalized. To determine the comparative significance of each score value, the normalized score function values were subtracted sequentially, meaning each value from the one that follows it. To find the relative coefficient, 1 was added to each comparative significance score. Finally, the significant and final significant weight values were calculated in sequence.

Maintenance 4.0 is an increasingly critical topic in today’s industrial landscape. However, as a relatively new concept, it brings certain challenges and ambiguities. Therefore, this study aims to highlight these barriers and determine their significance levels. It is believed that identifying these significance values will contribute to more meaningful and effective management of the related issues.

The assessments were applied by DM_s in the Maintenance 4.0. DM_s is illustrated in

Table 5. Information of DM_s.

DMs	Graduation	Graduation Degree	Experience (Year)	Position
DM1	Mec. Eng.	Bachelor Science	8	OPEX Chief
DM2	Ind. Eng.	Bachelor Science	11	Maintenance Chief
DM3	Comp. Eng.	Bachelor Science	14	Data Analytic Expert
DM4	Mec. Eng.	Master Science	18	OPEX Manager
DM5	Ind. Eng.	Master Science	21	Maintenance Manager

. Considerable time was required both to identify suitable individuals for the evaluation and to ensure the completion of their assessments. Additionally, processing their inputs in the methodological framework was time-consuming. To enhance the reliability of the study, five DM_s were consulted during the application of fuzzy methods. The DM_s involved in this study are presented in Table 5.

The general roles of the experts within the company are presented in

Table 6. The general roles of the experts within the company.

DMs	Roles
DM1	Process improvement initiatives related to TPM and digitalization
DM2	TPM and digital transformation activities
DM3	Data collection, analysis, and modeling
DM4	The coordination of general process improvement
DM5	The systematic management of maintenance operations

.

The textile sector was deliberately chosen for this study due to its exceptionally high level of human capital and its relevance as a dynamic and complex industry. This focus allows for in-depth, context-specific insights and provides a robust environment to test the proposed framework. The textile sector is particularly labor-intensive and employs a large number of workers. In addition to the high number of employees, the generally low level of qualifications among workers contributes to frequent errors. These errors often lead to both product- and machine-related problems. For this reason, the textile sector was selected for this study. A company utilizing advanced technology was chosen to conduct the research.

The aim of this study was to examine Maintenance 4.0 barriers within this sector and to demonstrate the feasibility of integrating Maintenance 4.0 strategies with Industry 4.0 technologies. Bursa, home to one of Turkey’s largest organized industrial zones, holds significant prominence in the textile industry and serves as a leading hub in this field. The sector was selected with the intention of setting a precedent for other industries and encouraging digital transformation and maintenance initiatives across different segments of the textile industry.

Since consultancy and training activities had previously been conducted with the DM_s, their respective areas of expertise were already known. To enhance the efficiency of the study, the opinions of five DM_s were collected. Five DM_s currently working in the textile sector in Bursa participated in the study. Professionals with experience in both maintenance practices and technological implementation were purposefully chosen. The interviews were conducted face-to-face, and pairwise comparisons were performed using linguistic variables. Linguistic terms were utilized to enhance the clarity and precision of the assessments.

The methodology was explained with examples to ensure a clear understanding of the process. The scales and criteria used for evaluation were also introduced in detail. Additionally, a sensitivity analysis was introduced to verify the reliability of the weight assignments given to the DM_s.

The primary objective of this study was to prioritize criteria through the use of the Pythagorean Fuzzy SWARA method.

This approach, recognized in the contemporary literature for its effectiveness, had not yet been applied to analyze Maintenance 4.0 barriers. It was selected for its robustness and ability to yield reliable, nuanced insights. The evaluations were carried out based on the linguistic scale provided in Table 4. The implementation calculations of this study are demonstrated in Figure 4.

The hesitation degree for Case 1 is calculated using Equation (3), as demonstrated in

Table 7. Linguistic terms for DM_s.

DMs	μ	ν	π
DM1	0.6	0.4	0.69282
DM2	0.8	0.2	0.565685
DM3	0.8	0.2	0.565685
DM4	0.9	0.1	0.424264
DM5	0.9	0.1	0.424264

. First, the decision-makers were evaluated in Equation (5), as shown in Table 3. The evaluation table of the five decision-makers is presented in Table 7. Once the hesitation values are determined, the importance weights of the DM_s are computed using Equation (6).

Subsequently, the linguistic evaluations of the decision-makers are presented in

Table 8. Evaluation results for criteria by DM_s.

Criteria	DM1	DM2	DM3	DM4	DM5
B1	VS	I	S	I	VS
B2	VB	VB	VB	VB	VB
B3	AI	B	SB	I	I
B4	AI	AI	I	I	SB
B5	VS	ES	VS	ES	VS
B6	VS	ES	S	ES	VS
B7	ES	ES	VS	ES	VS
B8	B	VB	B	B	VB
B9	AI	I	BI	SB	I
B10	SB	AI	SB	AI	SB
B11	ES	ES	ES	ES	ES
B12	EB	EB	VB	EB	VB
B13	VB	SB	SB	I	VB
B14	S	SS	SS	S	S
B15	S	VS	ES	VS	S
B16	VS	S	S	VS	VS
B17	S	BI	SS	BI	S
B18	AI	AI	I	BI	AI
B19	AI	BI	BI	BI	AI
B20	S	BI	S	BI	S
B21	AI	I	I	SB	I
B22	ES	ES	ES	VS	VS
B23	SS	SS	S	BI	S
B24	BI	SS	I	SS	S
B25	AI	I	I	BI	S
B26	AI	B	AI	SB	B
B27	VS	S	VS	S	VS
B28	BI	I	BI	AI	I
B29	I	I	I	SB	I
B30	S	VS	ES	VS	VS
B31	VB	SB	AI	VB	B

, based on the linguistic scale provided in Table 4.

The calculation of the criteria weights using Equations (3)–(12) is shown in

Table 9. Barriers’ weights.

Criteria	Normalized Score Function	Comparative Significance of Normalized Score Value (${\mathit{s}}_{\mathit{j}}$)	Relative Coefficient (${\mathit{k}}_{\mathit{j}}$)	Significance Weight (${\mathit{q}}_{\mathit{j}}$)	Final Significant Weight (${\mathit{w}}_{\mathit{j}}$)
B12	1	-	1	1	0.0541
B2	0.9	0.1	1.1	0.909091	0.0492
B31	0.831315	0.068685	1.068685	0.850663	0.0461
B8	0.831176	0.000139	1.000139	0.850545	0.0461
B13	0.715614	0.115561	1.115561	0.762437	0.0413
B26	0.703345	0.01227	1.01227	0.753195	0.0408
B10	0.647618	0.055727	1.055727	0.713438	0.0386
B3	0.644896	0.002722	1.002722	0.711501	0.0385
B21	0.596817	0.048079	1.048079	0.678861	0.0368
B29	0.582681	0.014136	1.014136	0.669399	0.0362
B9	0.582143	0.000537	1.000537	0.669039	0.0362
B4	0.560443	0.0217	1.0217	0.654829	0.0355
B18	0.523264	0.037179	1.037179	0.631356	0.0342
B28	0.506094	0.01717	1.01717	0.620699	0.0336
B19	0.463598	0.042496	1.042496	0.595397	0.0322
B25	0.460446	0.003152	1.003152	0.593526	0.0321
B1	0.35398	0.106466	1.106466	0.536416	0.0290
B24	0.352593	0.001387	1.001387	0.535673	0.0290
B17	0.326961	0.025632	1.025632	0.522286	0.0283
B20	0.308639	0.018323	1.018323	0.512888	0.0278
B23	0.299045	0.009594	1.009594	0.508015	0.0275
B14	0.243975	0.05507	1.05507	0.481498	0.0261
B27	0.148159	0.095816	1.095816	0.439397	0.0238
B16	0.143866	0.004293	1.004293	0.437519	0.0237
B15	0.118485	0.025381	1.025381	0.426689	0.0231
B30	0.098138	0.020347	1.020347	0.418181	0.0226
B6	0.083178	0.01496	1.01496	0.412017	0.0223
B5	0.058957	0.024221	1.024221	0.402273	0.0218
B22	0.045892	0.013065	1.013065	0.397085	0.0215
B7	0.043767	0.002125	1.002125	0.396243	0.0215
B11	0	0.043767	1.043767	0.379628	0.0206

.

Based on the linguistic evaluation results provided in Table 8 for the criteria, the PFWA value is calculated using Equation (7). Subsequently, the score function is computed using Equation (9). This score function value is then normalized using Equation (5) to obtain the normalized score function. Following this, the comparative significance of the normalized score value ($s_j$) is determined. The relative coefficient ($k_j$) is calculated using Equation (10), while the significance weight ($q_j$) is obtained using Equation (11). Finally, the final significant weight ($w_j$) is computed using Equation (12).

The weights of the main and sub-criteria are presented in

Table 10. The weights of the main and sub-criteria.

Main Criteria	Weights	Sub-Criteria	Global Weights
Organizational and Managerial	0.2178	B1	0.0290
		B2	0.0492
		B3	0.0385
		B4	0.0355
		B5	0.0218
		B6	0.0223
		B7	0.0215
Technical Competence	0.1956	B8	0.0461
		B9	0.0362
		B10	0.0386
		B11	0.0206
		B12	0.0541
Data-Related and Cybersecurity	0.1766	B13	0.0413
		B14	0.0261
		B15	0.0231
		B16	0.0237
		B17	0.0283
		B18	0.0342
Human Resources and Training	0.1183	B19	0.0322
		B20	0.0278
		B21	0.0368
		B22	0.0215
Financial	0.0886	B23	0.0275
		B24	0.0290
		B25	0.0321
Operational	0.2031	B26	0.0408
		B27	0.0238
		B28	0.0336
		B29	0.0362
		B30	0.0226
		B31	0.0461

.

The final significant weight ($w_j$) values presented in Table 8 have been used to derive the sub-criteria values shown in Table 9. Based on these values, the importance weights of the corresponding main criteria have also been determined. The findings of this study indicate that Organizational and Managerial is the most critical main criteria. Operational is next in significance, with Technical Competence closely following. Data-Related and Cybersecurity holds the fourth position in the ranking. Conversely, Human Resources and Training and Financial were deemed the least significant among all criteria. These two factors received comparatively lower weightings, with Financial emerging as the least influential overall. The most important criterion highlights the strategic and leadership-related shortcomings of management. It indicates that the organization needs to utilize changes in management practices to effectively prepare for transformation. Operational and technical aspects also stand out with high weightings, emphasizing the importance of process improvement and in establishing a robust technical infrastructure. The “Data-Related and Cybersecurity” dimension points to situations where artificial intelligence and big data are not adequately supported. Therefore, prioritizing the integration of AI and data utilization is essential. Human resources and financial issues are relatively less significant compared to other factors. However, deficiencies in these areas can be addressed through targeted training initiatives and the development of a long-term financial structure.

Among the sub-criteria, the most important one is “Integration challenges between IT and OT systems”, which belongs to the Technical Competence criteria. The second most significant sub-criterion is “Need for organizational restructuring of maintenance-related functions”, under the Organizational and Managerial criteria. The third most important is “Lack of a structured framework specific to Maintenance 4.0 implementation”, which falls under the Operational criteria. The fourth is “Low maturity level of enabling technologies”, also associated with Technical Competence.

To measure the sensitivity of the results, five additional scenarios were applied. These scenarios are presented in

Table 11. Sensitivity scenarios.

Scenarios	DMs	Linguistic Terms
Case 1	DM1	N
	DM2	TS
	DM3	TS
	DM4	THS
	DM5	THS
Case 2	DM1	TS
	DM2	TS
	DM3	TS
	DM4	TS
	DM5	TS
Case 3	DM1	TS
	DM2	TS
	DM3	TS
	DM4	THS
	DM5	THS
Case 4	DM1	N
	DM2	THS
	DM3	THS
	DM4	TS
	DM5	TS
Case 5	DM1	TS
	DM2	THS
	DM3	THS
	DM4	TS
	DM5	TS
Case 6	DM1	THS
	DM2	N
	DM3	N
	DM4	TS
	DM5	TS

. According to these results, the outcomes of the sensitivity analysis are illustrated in Figure 2. As seen from the results, the expert weightings had little effect on the final outcomes. Even in Case 6, where the expert weight ratios were completely reversed, the results did not change significantly. This highlights an important point regarding the reliability of the evaluation.

According to these results, the outcomes of the sensitivity analysis are illustrated in Figure 5. A sensitivity analysis was conducted to examine how variations in the weights of the decision-makers (DM_s) influence the criteria weights. As observed, only minor changes occurred in the ranking of the criteria. This outcome indicates that the results obtained in Case 1 are robust and reliable. A sensitivity analysis was conducted by altering the importance weights of the decision-makers and applying the Pythagorean Fuzzy SWARA method accordingly. For this purpose, six distinct cases were defined and evaluated.

As seen from the results, the expert weightings had little effect on the final outcomes. Even in Case 6, where the expert weight ratios were completely reversed, the results did not change significantly. This highlights an important point regarding the reliability of the evaluation.

5. Conclusions

With the rapid entrance of Industry 4.0 applications into our lives with cutting-edge technologies, the Maintenance 4.0 paradigm has also become vital for businesses. Maintenance 4.0 does not only aim to detect and prevent failures, but also helps businesses to survive in the face of intense competition by increasing productivity with the effective use of resources. Predictive, smart, proactive, and autonomous maintenance strategies are developed thanks to tools such as artificial intelligence, the Internet of Things, and data analytics used within the scope of Maintenance 4.0 applications. In this way, both equipment availability and reliability are increased, and maintenance costs and production breakdown due to failures are minimized. Therefore, the Maintenance 4.0 paradigm is important for the sustainability of manufacturing enterprises. Although the importance of Maintenance 4.0 is known, it is not very easy to implement them in real life. In this study, the importance of Maintenance 4.0 is emphasized and the barriers encountered are determined. They are defined in line with a comprehensive literature review and expert opinions. The barriers that Maintenance 4.0 applications will encounter in the textile industry, where there is intensive labor use, and the weights of these barriers are determined. The SWARA method, based on Pythagorean fuzzy sets, was used to determine the weights of the barriers.

The results of this study reveal that Organizational and Managerial stands out as the most crucial main criterion. It is followed by Operational, while Technical Competence ranks closely behind in third place. Data-Related and Cybersecurity takes the fourth position in the hierarchy. On the other hand, Human Resources and Training and Financial are identified as the least significant among the main criteria, with Financial having the lowest overall weight.

As for the sub-criteria, the top-ranked is “Integration challenges between IT and OT systems”, which falls under Technical Competence. The next in importance is “Need for organizational restructuring of maintenance-related functions”, categorized under Organizational and Managerial. Ranked third is “Lack of a structured framework specific to Maintenance 4.0 implementation”, belonging to the Operational criteria. The fourth most critical sub-criterion is “Low maturity level of enabling technologies”, also linked to Technical Competence.

This study demonstrates that addressing Maintenance 4.0 requires not only overcoming technological limitations, but also implementing cultural changes within the organization. From a managerial perspective, Maintenance 4.0 should be integrated into the organization’s strategic objectives, which will help promote internal acceptance of the initiative. A more flexible internal structure should be established to support the operations of the maintenance department. For technical barriers, system interfaces should be developed and pilot projects should be conducted to ensure the effective use of data. To overcome data-related barriers, a system that ensures data quality and security must be established. In terms of human resources, training programs and suggestion systems should be implemented to increase employee engagement. As for financial barriers, investments should be planned with a long-term perspective to ensure sustainability and continuity.

The textile sector is limited in terms of both its operational and technological capacities. In this regard, it imposes restrictions on the generalizability of findings compared to other industries. Moreover, differences in workforce structures and organizational cultures are likely to emerge, further limiting the applicability of sector-specific insights. Therefore, incorporating examples from diverse sectors would enable a more comprehensive understanding of the challenges associated with Maintenance 4.0, ultimately leading to more generalizable and robust conclusions.

The most important limitation of this study is the temporal and spatial difficulties experienced in collecting data to determine the barriers and their relative importance levels. In addition, the collection of data only from textile enterprises operating in Turkey may be another limitation. The fact that this study cannot be applied to wider target groups, different countries, and different industry branches may be a barrier. In the future, the same problem can be addressed by determining different barriers for different industries. In addition, different fuzzy sets can be used in solving the problem, and the results can be analyzed with different weighting methods based on the same or different fuzzy sets. Hesitant, Picture, Spherical Neutrosophic, and Fermatean methods can be used in future studies.