Informatica

Buildings are responsible for an important portion of global energy consumption and carbon emissions. Energy consumption is related to the construction, operation, renovation, and demolition processes of buildings, as well as the production of building materials and various systems (Abass and Muthulingam, 2025). Rapid population growth and migration from rural areas to cities have made the construction sector one of the most energy-intensive and waste-producing sectors (Llantoy et al., 2020; Movaffaghi and Yitmen, 2023), making environmental impact management challenging in the construction industry (Tushar et al., 2022). Based on the Global Alliance for Buildings and Construction (GABC) 2024/25 global status report, in 2023, buildings accounted for 32% of global energy demand and 34% of CO₂ emissions. It also notes that despite a small reduction in carbon emissions and a slight increase in the use of renewable energy, these gains are insufficient to meet the Paris Agreement targets (GABC, 2025).

With the abandonment of the traditional “take-make-dispose” approach (Attri et al., 2025; Wahyu Adi and Wibowo, 2020), the circular economy (CE) perspective, focusing on material circularity, waste reduction, and efficient resource use has also emerged as a strategic framework for the transformation of the construction industry, like other industries (Tan et al., 2025; Gupta et al., 2025; Jain et al., 2025). CE intends to preserve the value of products, materials, and resources throughout their life cycle and to decrease wastes via useful techniques such as reuse, recycling, refurbishment, recovery, and cleaner production (Nayeri et al., 2025; Hasheminezhad et al., 2024; Salman and Hasar, 2023). In this way, CE aims to maximize the effective and efficient use of production elements, thereby extending their lifespan (Govindan et al., 2020). In the construction industry, this could be achieved by using recycled or reusable materials (Melikoglu, 2025; Timm et al., 2023), making modular designs (Abdelmageed and Zayed, 2020; Oteng et al., 2025), realizing energy-efficient production (Ferreira Junior et al., 2025; Ludger Bernsmann and Schleifenbaum, 2025), reducing carbon footprint (Almusaed et al., 2024; Iqbal et al., 2025), using reverse logistics (Mishra et al., 2022), gathering environmental certification systems, and considering life cycle management (Llantoy et al., 2020; Manu, 2024).

Sustainability has become a major concern in supply chain management due to increasing environmental pressures, resource scarcity, climate change, and rigourous environmental regulations (Govindan et al., 2020). It is evident that supply chains exert a direct influence on various key performance indicators, including energy consumption, waste generation, carbon emissions, transportation activities, and resource efficiency throughout the life cycle of products and materials (Tushar et al., 2022). There is an increasing expectation that firms will integrate sustainability considerations into purchasing, logistics, production, and supplier selection processes (Kannan et al., 2020). In the construction industry, where material-intensive production and transportation activities create substantial environmental impacts, sustainable supply chain management plays a critical role in reducing waste, improving resource efficiency, lowering carbon emissions, and supporting the circular economy practices (Tushar et al., 2022). For this reason, supplier selection decisions are no longer based solely on traditional factors such as cost, quality, and delivery performance, but also increasingly include environmental and circularity-related considerations (Kusi-Sarpong et al., 2023; Tramarico et al., 2025).

Aerated concrete is frequently preferred in the construction industry for its light weight, satisfactory thermal insulation, and material savings (Gyurkó et al., 2019). Collaboration with suppliers who possess those competencies significantly enables aerated concrete companies to support CE. Consequently, supplier selection in the construction supply chain for the aerated concrete industry is crucial for successful CE applications (Tushar et al., 2022). Additionally, selecting the right suppliers may reduce costs and increase supply chain efficiency, thereby enhancing the firms’ profitability and ability to meet customer expectations (Ecer et al., 2024). Whereas traditional supplier selection processes prioritize factors such as cost, quality, delivery, and financial flexibility (Paul and Pal, 2025), CE-focused supplier selection could consider environmental and social criteria, including waste management, energy efficiency, carbon emissions, environmental compatibility, recyclable raw materials, social responsibility, clean technology, life cycle impacts, etc. (Ulutaş et al., 2025; Tramarico et al., 2025). Such a broad assessment involves a multi-dimensional decision-making process with inherent uncertainty.

Although there is some research on circular supplier selection in the available literature, many focus on automotive (Govindan et al., 2020), electronics (Kannan et al., 2020; Menon and Ravi, 2022), manufacturing (Liu et al., 2022; Tong et al., 2022), textile (Kusi-Sarpong et al., 2023; Ecer and Torkayesh, 2022), and chemical (Mina et al., 2021; Tramarico et al., 2025) industries. However, in the construction sector, supplier selection models that comprehensively incorporate CE principles are limited, and most evaluations are restricted to environmental drivers (Amarasinghe et al., 2024; Demirbağ et al., 2025). Furthermore, there are a few studies in which CE criteria have been systematically determined in conjunction with expert opinions to cope with uncertainty and actual industry practices. As far as the authors’ knowledge goes, there is no work on supplier selection in the aerated concrete industry. Besides, the existing literature shows that available circular supplier selection works largely focus on developed industrial sectors and global-scale examples. In developing countries, such as Türkiye, applied research on the construction sector and building materials is very scarce. For building materials like aerated concrete, which provides energy efficiency in buildings, has low carbon emissions, and is easy to transport due to its lightness, there is almost no circular supplier evaluation model in the literature. Circular supplier selection also involves managing uncertainty based on expert judgments and evaluations. In the construction industry, where many project-based, simultaneous, or sequential processes are carried out together, it is very difficult to measure supplier performance accurately and objectively. Therefore, decision-makers often rely on judgments based on their experiences or perceptions. In such an environment, supplier selection becomes a decision problem involving intensive uncertainty, and this problem needs to be addressed with reliable and powerful decision support systems. These research gaps stand out as critical for both practical applications and decision-makers for supplier selection in the construction supply chains to achieve sustainable development. This study, therefore, aims to fill those gaps by introducing a supplier selection methodology concerning CE principles. 16 drivers are determined by searching the literature and the opinions of experts. Five large companies operating in the aerated concrete sector in Türkiye are considered alternatives by experts using those drivers. To obtain trustworthy results under uncertainty, a comprehensive fuzzy hyperbolic (HyF) multi-criteria decision-making (MCDM) approach is developed. Criteria weights are decided using the Logarithmic Percentage Change-Driven Objective Weighting (LOPCOW) and Ranking Comparison (RANCOM) methods, balancing objective and subjective information, while alternatives are ranked using the hyperbolic extension of the Deviation-Based Pairwise Assessment Ratio Technique (DEPART) method with hyperbolic fuzzy information. Thus, a cutting-edge decision support system, named HyF-LOPCOW-RANCOM-DEPART, is introduced for the first time.

The research presents significant novelties and contributions in terms of contextual and methodological aspects. It makes significant contributions to the literature on CE-focused supplier selection, specifically in the construction sector. First, addressing supplier selection in the aerated concrete industry within the framework of CE principles fills a notable gap in the literature. The 16 evaluation criteria presented in this study encompass economic, environmental, social, and technological dimensions. These criteria were developed based on both a comprehensive literature review and expert opinions. The research offers an innovative approach by basing the evaluation set not only on theory but also on industry-specific information. The ranking of Türkiye’s leading aerated concrete manufacturers based on expert opinions has enabled the creation of a robust model based on real-world field data. This also increases the applicability of the findings to industry decision-making processes. Furthermore, expanding the CE criteria to include environmental indicators as well as waste management, life cycle assessment, carbon footprint, and technological capabilities enriches the perspective on supplier selection. With all these features, the study integrates theory and practice, producing an innovative approach and providing a significant reference for establishing a circular supply chain in the construction industry. Further, the novel methodological novelties and contributions of this research include:

The rationale for choosing these decision methods to develop a decision framework is given below:

It must be noted that hyperbolic fuzzy MCDM is considered a crucial research topic because the HyFS is an emerging orthopair variant that considers both preference (membership) and non-preference (non-membership) grades and has flexible constraints, which allow decision experts to effectively utilize this fuzzy variant for modeling uncertainty during the rating of decision entities. Since the HyFS was recently proposed, its exploration in decision frameworks and solving decision problems has just started, and so it is seen as a promising research topic for researchers to develop decision frameworks under HyFS for rational decision-making. In this paper, we are considering a problem of selecting a suitable supplier in the construction field by focusing on CE principles within the aerated concrete industry. Since the problem has a clear structure of MCDM with different alternative suppliers rated by decision experts on different criteria, we extend this problem into the HyFS-based MCDM construct and attempt to solve the decision problem through a structured methodical framework.

The remaining part of the study is organized as follows. The second section contains a literature review. Within this scope, studies related to CE-focused supplier selection is included. Subsequently, studies employing different methods related to the subject are listed, and research gaps are conveyed. The third section introduces the research methodology. The fourth section presents the application section. The fifth and sixth sections respectively present the discussion, theoretical and practical implications, and conclusions.

The evaluation data obtained based on the expert profiles and business structures defined in this section were used as input in multi-criteria decision-making techniques to solve the circular economy-based supplier selection problem. In our study, we developed a methodological framework based on hyperbolic fuzzy sets to account for the uncertainty arising from the nature of the criteria, the subjective nature of expert opinions, and the uncertainty in the performance of the alternatives. In this context, we explain step by step in the following subsections the hyperbolic fuzzy set structure used in the study, how the criteria and expert weights were determined, and how the alternatives were ranked.

The proposed HyF-LOPCOW-RANCOM-DEPART model is applied to address the CE-focused supplier selection problem in the aerated concrete industry. In this context, researchers designed a multi-layered research process to comprehensively solve the problem, as shown in Fig. 1.

The preparation process for this study was carried out in 3 stages. In the first step, a pool of criteria containing 22 main criteria was created through face-to-face interviews with relevant literature and experts. The experts consulted on the construction industry’s circular economy-focused supplier selection was building construction, concrete manufacturing, and building market parts sales managers. These individuals were also involved in the data collection process in the third stage. Furthermore, the positions and experience of the industry experts were, respectively: building construction company owner (7 years), building materials store company owner (8 years), building materials store, and building construction company owner (25 years), building materials store marketing-sales manager (11 years), and construction materials supplier marketing-sales manager (11 years). Despite the fact that three of the participating experts are company owners or senior decision-makers, the study aimed to reduce potential individual bias by including experts from a variety of professional backgrounds, including purchasing, accounting, operations, and management. Furthermore, the entropy-based expert weighting approach was utilized to minimize the dominance of highly uniform evaluations by considering the consistency and variation of expert judgments. Given that supplier selection decisions in the construction materials sector are influenced by both operational and financial considerations, it was considered important to include company owners in order to reflect real-world decision-making dynamics.

In the second stage, experts from academia and industry checked the suitability of the criteria pool for the study. These experts were also among the practitioners mentioned earlier.

Specifically, the suitability of the 22-criterion pool was assessed through a structured voting exercise: each candidate’s criterion was put to the panel, and the experts voted on its relevance and appropriateness for circular-economy-based supplier selection in the aerated concrete industry. Criteria that did not attain the agreed majority support, chiefly those judged redundant with a retained criterion, not specific to the aerated concrete supply chain, or of limited practical measurability, were removed. As a result, six criteria were eliminated, and the remaining 16 criteria (Table 2) were carried forward to the data-collection survey in the next stage.

The criteria used in the study, along with their definitions and references, are shown in Table 2. The research data were collected in the third stage between October and November, 2025, from five experts (EXT1-EXT5). The participants consisted of company owners and managers in purchasing and marketing departments who were directly involved in the decision-making processes regarding supplier selection. Participants had between 7 and 25 years of experience, ensuring that supplier evaluations were based on both industry knowledge and field experience. Participants had at least a university degree, typically in business administration or accounting. Their age range was 30–45. Considering their industry experience and age, it is clear that they were directly involved in field work after their education. For evaluations in a multi-dimensional and in-depth field like the circular economy, this panel is considered a suitable data source.

Decision-makers evaluated five major firms that play an important role in the aerated concrete industry using the suggested framework in this research. In line with the principles of confidentiality and impartiality, the names of the companies included in the study are not explicitly stated; companies are referred to by codes AL1–AL5. Firms of varying scales and organizational structures operating in Türkiye’s construction industry are involved in the production and distribution of aerated concrete. They include those with nationwide recognition as well as those active in specific regional markets. The circular supplier selection process is not based on a single profile but rather reflects the industry’s overall structure.

The assessments obtained from the experts and businesses defined in this section were directly used in the application of the circular economy-based supplier selection model developed in the study. These data, based on expert opinions, were analysed within the framework of hyperbolic fuzzy sets to determine the criterion weights and alternative rankings. The steps followed during the application process are presented below.

The steps for determining the values of decision entities are given below:

Step 1: Collect data from five experts (EXT) on 16 criteria. Each expert rates the criteria to form a 5×16 weight calculation matrix (Table 3).

Step 2: Considering the dimension of the data, the feature reduction algorithm is applied from Section 3.2 for determining the Laplacian score. From the algorithm, it is clear that a criterion with a low score gets a high preference, meaning that the criterion has greater representation ability compared to others.

Table 4 gives the similarity matrix and degree vector values that are calculated by using the Laplacian algorithm. The values 0 to 4 denote EXT1 to EXT5. A Laplacian matrix is determined by using these two results, and that is shown in Table 5. Finally, the score vector of each criterion is determined, and based on the ascending order of scores, the criteria are arranged. From the list, the best K criteria are chosen.

The score value for each criterion is determined as CRT1: 1.183, CRT2: 1.183, CRT3: 1.139, CRT4: 0.99, CRT5: 1.019, CRT6: 1.018, CRT7: 1.018, CRT8: 0.923, CRT9: 0.94, CRT10: 1.034, CRT11: 1.047, CRT12: 1.297, CRT13: 1.133, CRT14: 1.183, CRT15: 1.012, and CRT16: 0.885 based on the values in Table 5. By arranging these values in ascending order, the top 10 criteria chosen are CRT4, CRT5, CRT6, CRT7, CRT8, CRT9, CRT10, CRT11, CRT15, and CRT16.

These ten criteria constitute the reduced set used for the subsequent criteria weighting and supplier ranking; the remaining six are retained in the validated framework but excluded from the computational evaluation in order to reduce dimensionality. Hence, the framework validates 16 criteria, of which 10 are used in the final evaluation. It is noted that the Laplacian score is applied to the 16 expert-validated criteria (not the original 22-criterion pool): the 22 to 16 reduction is the prior Stage-2 expert-voting step, whereas the Laplacian score performs this data-driven 16 to 10 reduction while preserving the local structure of the rating data. Figure 2 reports the Laplacian score of each of the 16 criteria and the top-10 (lowest-score) criteria that are retained.

Step 3: Collect data about alternatives from the experts. Five alternatives are rated based on 10 criteria (feature reduced) by five experts. So, five matrices of $5\times 10$ are obtained (Table 6).

Step 4: Use the data from Step 3 (Table 6) and the procedure in Section 3.3 for determining the weights of experts. A vector of $1\times 5$ is obtained (Table 7).

Use the procedure in the expert weight calculation section to determine the entropy (Table 7) associated with each expert rating, and by normalization, we get weights as: 0.201, 0.198, 0.188, 0.212, and 0.202, respectively.

Step 5: Consider the rating of 10 criteria by five experts to form a weight matrix of $5\times 10$ order. By applying the criteria weight procedures from Section 3.3, the objective and subjective weights of the criteria are determined. Each weight yields a vector of $1\times 10$. They are combined to form a single weight vector of $1\times 10$.

Table 8 provides ratings from experts on the criteria that are considered for the study after feature reduction. Laplacian score supported the feature-reduction and facilitated dimensionality reduction, which, from the managerial context, corresponds to reduced complexity and computational overhead. Eqs. (10)–(12) corresponds to the determination of criteria weights by LOPCOW. Data in Table 8 is transformed to its corresponding HyFSs, and scores are determined. Eq. (11) determines the percentage value, which maps to a logarithmic scale for nullifying the spikes in value due to extreme value rating and yields a weight vector upon normalization, which is a vector of $1\times 10$ order (Table 9).

From Eqs. (11)–(12), the results in Tables 9 and 10 are depicted. From Table 9, data is given to the LOPCOW for calculating the objective criteria weights (refer to Table 10). Further, data in Table 8 is provided to Eqs. (13)–(14) for determining the subjective criteria weights using the RANCOM method (Table 11).

Table 12 presents the subjective weight values from each expert, which are further used for determining the subjective weights of criteria. Eq. (14) determines the weights, and it is shown in Table 13.

The results in Table 10 and Table 12 are used for determining the combined weights of criteria by considering the step size as 0.50. The weights are calculated as: CRT4: 0.126, CRT5: 0.117, CRT6: 0.072, CRT7: 0.072, CRT8: 0.059, CRT9: 0.075, CRT10: 0.079, CRT11: 0.140, CRT15: 0.152, and CRT16: 0.108, respectively.

Step 6: Apply the procedure in Section 3.4 for determining the individualistic and combined ranks of alternatives by considering data from Step 3. A rank vector of $1\times 5$ is obtained.

Eqs. (15)–(20) are applied for determining the ranks of alternative suppliers. Note that the ranks from each expert’s data are obtained, and the combined ranks are also determined using the Copeland strategy. We present the results from the DEPART method for one expert, and the combined ranks of alternative suppliers are presented from Eq. (20) and the Copeland strategy.

Table 13 depicts the results of the DEPART method by considering the rating from expert EXT1. Likewise, results are determined for other experts’ rating data, and the ranks at 0.50 are depicted in Table 14 for all five experts. Eq. (15)–(19) is used on the rating data provided in Table 7. Each parameter of DEPART is determined and shown in Table 14. Table 15 shows the rank values of alternative suppliers by considering data from each expert and strategy value at 0.50.

From Table 14, it is clear that supplier ‘AL2’ is the least preferred by five experts, and the conclusion on the top alternative supplier varies across experts. To arrive at a consensus, we apply the Copeland strategy, and the results are shown in Table 15.

From Table 15, the combined ranks of alternative suppliers are clear. Alternative ‘AL3’ is most preferred, while ‘AL2’ is least preferred. From the Copeland strategy, we are able to arrive at a consensus on the top priority and the least priority alternative. In the earlier situation, it was difficult as each expert’s ranking provided a new rank order that led to non-consensus in the final priority of alternatives. Besides, this approach facilitates the understanding of ranking by each expert, which otherwise ceases to exist. This improvement in rank determination is lacking in extant models.

This research focuses on supplier selection in the construction supply chain for the aerated concrete industry, considering CE drivers. To do this, we introduced the HyF-LOPCOW-RANCOM-DEPART framework, which allows for a more systematic and transparent approach to supplier selection, addressing uncertainty and subjective evaluations. The study contributes to the supply chain process by providing a practical tool for incorporating CE applications into the decision-making process in the construction materials industry. Another significant contribution of the study is that it clearly demonstrates that CE-based supplier selection is not solely based on environmental indicators. The research results emphasize that factors such as technological capability, management structure, and corporate reputation are at least as important as environmental criteria. The proposed HyF-LOPCOW-RANCOM-DEPART decision support approach for CE-focused supplier selection produces more systematic and balanced results by considering the characteristics of experts. Thanks to this model, which makes a significant methodological contribution to the literature, intuitive decisions are replaced by consistent evaluations. In this respect, the proposed model can be used not only for the aerated concrete sector but also for other sectors with similar characteristics.

Whereas the research offers valuable contributions, it also has some limitations. First, the study is based on the assessments of a limited number of experts. Although the expert panel consisted of individuals with extensive field experience, increasing the number of experts could strengthen the representativeness of the results. Another limitation is that the research was conducted within a specific industry and a limited number of firms. This limits generalization to other construction materials or different sub-industries. Although the present study focuses on the aerated concrete sector in Türkiye, the proposed decision-making framework is flexible and can be adapted to different geographic and industrial contexts. The methodology is not limited to a specific country or market structure, and the evaluation criteria may be revised according to regional regulations, environmental policies, supply chain characteristics, and market conditions. Future studies may apply the proposed framework to different construction material sectors, such as insulation materials or prefabricated building components, and compare circular economy priorities across developed and developing countries. Furthermore, applications in different developed and developing countries and regional contexts could reveal the impact of regulatory frameworks and market conditions on supplier selection outcomes. Third, this research provides a static assessment and does not consider changes in supplier performance over time. Future research could analyse the development of suppliers’ CE performance over the years using dynamic evaluation models. Fourth, the proposed framework assumes that the data is fully available, which might not always be practically viable. In such cases, the present model cannot handle the non-availability of data. Partial information about decision parameters, such as criteria and alternatives, cannot be embedded in the present framework. These limitations will be handled in the future. Furthermore, plans are made to extend the proposed framework for other decision applications in the field of logistic service providers, renewable energy-based transit mode selection, technology intervention assessment in supply chains, resilience, and humanitarian supply chain evaluation. Novel fuzzy variants, such as pentagonal fuzzy set, bipolar fuzzy set, quantum fuzzy set, etc., may also be explored for data interpretation, and novel decision models can be developed under these fuzzy constructs for rational decision-making with better modelling of uncertainty. Finally, plans are made to integrate machine learning methods with a decision support system for carrying out data-driven decision activities with the help of a larger volume of data.