1 Introduction

In the wake of conflicts and crises, the efficient and effective reconstruction of infrastructure is crucial. The supplier evaluation and allocation process play a critical role in this context, yet traditional procurement methods often fail to meet the dynamic and complex needs of post-conflict environments (Khaled et al. 2011; Pal et al. 2013). Various multi-criteria decision making (MCDM) methods have been used in different industries such as food, automotive, healthcare, and petrochemical; however, research on the construction industry in conflict areas is rare (Tushar et al. 2022). To address this gap, the research is structured around the following central research questions (CRQs):

CRQ1: What supplier selection criteria are considered important for the construction industry in conflict areas?

CRQ2: How can we best incorporate the identified supplier selection criteria into a supplier selection process in conflict areas?

CRQ3: How can we best conduct a risk analysis in conflict areas based on GIS?

CRQ4: How does the integration of multiple evaluation criteria and GIS-based risk analysis into the supplier evaluation and allocation process utilizing fuzzy multi-criteria decision-making affect the efficiency and outcomes of construction projects in conflict zones?

We hypothesize that incorporating various humanitarian, economic, and technical criteria into the evaluation process will make it more comprehensive. Utilizing GIS-based risk analysis will lead to a more realistic assessment of risks in conflict areas. Employing fuzzy multi-criteria decision-making (incorporating humanitarian, economic, technical, and risk criteria) will result in more effective and transparent supplier evaluations and allocations, thereby enhancing project success in these environments.

To address these CRQs, this study develops a structured hybrid fuzzy MCDM framework based on GIS-based risk assessment, F-AHP, and F-GP to tackle the supplier evaluation and allocation problem in the construction industry of northern Syria, a conflict area.

Motivated by the significant impact that procurement efficiency has on rehabilitation efforts, this research seeks to refine the procurement process in conflict-affected regions. Current methods often fall short due to the unpredictable nature of supply chains and the pressing needs of these areas (Sarkis and Talluri 2002; Christopher 2016). By developing a robust supplier evaluation and allocation model that accounts for a broad spectrum of criteria, this study aims to significantly improve transparency and effectiveness, impacting the success of reconstruction projects and the effective utilization of funds.

This study is significant as it addresses a notable gap in the literature concerning the application of fuzzy multi-objective decision-making models in post-conflict supplier selection (Ghodsypour and O’Brien 2001). The study aims to achieve the following main research objectives (MROs):

MRO1: Identify the supplier selection criteria relevant for the construction industry in conflict areas.

MRO2: Calculate weights of identified construction supplier selection criteria using the F-AHP method.

MRO3: Conduct GIS-based risk analysis in conflict areas.

MRO4: Evaluate and allocate construction suppliers using mathematical fuzzy goal programming.

MRO5: Discuss the implications of the proposed research.

This research contributes to the theoretical and practical understanding of procurement challenges and evaluation processes in conflict zones. It boosts transparency in the multi-criteria selection process, leading to a comprehensive framework, sustainability, and effective resource utilization, accelerating the return to normal life for affected populations, and contributing to the wider goal of effective reconstruction after crises. The findings are particularly relevant for non-governmental organizations, donors, local authorities, and other stakeholders engaged in reconstruction, offering them a novel approach that can lead to more informed and strategic decisions in construction supplier evaluation and allocation.

The paper is organized as follows: The literature review section outlines current methodologies and identifies their limitations. The methodology section introduces our innovative approach to supplier evaluation and allocation, while the application of this model is demonstrated through a detailed case study in section four. The final section discusses the implications of our findings and suggests potential avenues for future research.

2 Literature review

2.1 MCDM in supplier evaluation

MCDM represents a methodical process aimed at identifying the most favorable option among various feasible alternatives based on a set of defined criteria or attributes (Garg 2016a, b, 2017; Hallak et al. 2019, 2021; Hallak and Polat 2021). Selecting the correct supplier can significantly lower purchase costs and enhance an enterprise’s competitiveness, illustrating why many researchers believe that selecting a supplier is one of the most important activities of the purchasing department (Ghodsypour and O’Brien 2001).

In recent years, several supplier selection and evaluation methods have been developed. A number of MCDM methods have been employed in this research area: the analytical hierarchy process (AHP) (Etlanda and Sutawidjaya 2022); the fuzzy technique for order of preference by similarity to ideal solution (FTOPSIS) (Cakar and Çavuş 2021); the best–worst method (BWM) (Amiri et al. 2020); the fuzzy decision-making trial and evaluation laboratory (FDEMATEL) (Giri et al. 2022); and the multi-criteria optimization and compromise solution (VIKOR) (Gupta and Kumar 2022). To maximize the advantages and minimize the weaknesses of specific MCDM methods, researchers often integrate two or more MCDM methods into a hybrid method (Govindan et al. 2020). The literature related to supplier selection is dominated by these hybrid MCDM approaches. Further, researchers have begun employing combinations of more than two MCDM methods, such as F-AHP, F-TOPSIS and Fuzzy inference systems (FIS) (Mina et al. 2021), and integrating F-AHP with PROMETHEE II for selecting suppliers based on circular economy principles (Tushar et al. 2022).

This research follows a novel approach by surveying the affected population to define criteria, using F-AHP to calculate weights, and implementing GIS-based risk assessments to evaluate risks associated with each supplier. Finally, all parameters are integrated into fuzzy goal programming to optimally evaluate and allocate construction suppliers for projects in the area.

2.2 Criteria used in the supplier evaluation

Although supplier selection criteria often differ from industry to industry, surveys of the literature show that criteria related to delivery, quality, and price are commonly considered across various sectors. The quality of materials and services has been a focal point in several studies (Dickson 1966; WEBER et al. 1991; Tam 2001; Chan and Kumar 2007; Gencer and Guerpinar 2007; Guo et al. 2009; Wang 2010; Balezentis and Balezentis 2011; Zeydan et al. 2011; Kilic 2013; Cristea and Cristea 2017; Tamosaitiene et al. 2017; Wang et al. 2017; Fallahpour et al. 2017), while aspects such as total price and delivery time have been emphasized by others (Dickson 1966; Weber et al. 1991; Tam 2001; Chan and Kumar 2007; Gencer and Guerpinar 2007; Lee 2009; Guo et al. 2009; Lam et al. 2010; Wang 2010; Balezentis and Balezentis 2011; Kilic 2013; Hruska et al. 2014; Cristea and Cristea 2017; Wang et al. 2017; Buyukozkan and Gocer 2017; Fallahpour et al. 2017). Certifications of products and materials (Ting and Cho 2008; Hudymacova et al. 2010; Cristea and Cristea 2017), supplier reputation (Lin and Chang 2008; Rezaei et al. 2014), and warranty periods post-delivery have also been considered key factors in supplier evaluations (Cristea and Cristea 2017; Wang 2010). Consistent performance and reliability are crucial, as emphasized in the literature (Wang et al. 2004; Chan and Kumar 2007; Gencer and Guerpinar 2007; Lee 2009; Cristea and Cristea 2017). Despite the evolving landscape and increasing emphasis on qualitative criteria, financial parameters, delivery, and quality consistently emerge as core criteria in nearly all research on supplier selection. This trend is affirmed by the aforementioned studies, which significantly influence decision-making in the supplier evaluation process.

In this study, related literature served as the primary source for identifying supplier selection criteria, supplemented by feedback from surveying 32 NGOs implementing construction projects in Syria. These surveys focused more on risk and humanitarian factors in supplier evaluation in conflict areas, whereas previous studies often concentrated on stable communities and regions. Most earlier research predominantly relied on criteria weights determined through AHP, MCDM, and/or TOPSIS under normal and stable conditions. However, there has been a notable lack of research focused on developing optimal strategies for evaluating and allocating construction suppliers in crisis areas, especially considering humanitarian and risk factors assessed through GIS. This study extends existing methodologies by identifying and weighting criteria using F-AHP, performing GIS-based risk assessments, and deriving a risk value for each supplier. These values are integrated into a mathematical model that employs fuzzy goal programming to optimally allocate construction suppliers to projects. This approach aims to provide a robust framework for supplier selection in challenging environments, ensuring that projects are equipped with the best possible resources under demanding conditions.

3 Methodology

A novel hybrid approach has been introduced to address the complexities of this multi-criteria problem. To begin, a GIS-based risk assessment methodology was employed. This entailed determining the risk values associated with each potential construction supplier within the conflict-affected region through the integration of spatial analysis. Subsequently, a multi-criteria technique was applied utilizing F-GP to effectively solve the model and identify the optimal solution. Within this framework, particular emphasis was placed on the assignment of construction suppliers to the required projects that best align with a spectrum of goals encompassing financial, technical, spatial, humanitarian, environmental, and risk considerations. The ensuing sections provide a meticulous delineation of the steps and procedures undertaken in this approach, as shown in Fig. 1:

  1. 1.

    Dual-Pronged Data Collection: This involves two methods, surveys and Focus Group Discussions (FGDs).

    • ➢Surveys: We collected responses from 32 NGOs to obtain a broad perspective on different practices and standards in the field. These surveys provided quantifiable and comparable data about the NGOs' methods for evaluating suppliers and criteria utilized.

      Fig. 1
      figure 1

      The proposed methodology to solve the supplier problem

      Full size image

      We employed a purposive sampling approach to select NGOs for our survey. Around 40 active local NGOs that conduct construction projects were invited to participate in the surveys, but 32 NGOs responded. This approach ensured that all participants were actively engaged in construction projects within Syria, allowing for a focused examination of construction supplier evaluation practices in this specific context. This method facilitated access to targeted insights into the challenges and practices peculiar to NGOs operating in conflict-affected regions.

      Survey Validation: To ensure the reliability and relevance of our questionnaire for surveying NGOs, we implemented a validation process with input from both academic researchers and donors experienced in construction projects in Syria. The validation process included:

      • Expert Selection: Two academics specializing in construction supplier evaluation, two experts from local NGOs, and two donors involved in construction project funding were selected for their extensive knowledge and field experience, as detailed in Table 1.

      Table 1 The participating experts
      Full size table

      • Review Process: Experts reviewed the draft questionnaire, assessing clarity, relevance to our objectives, and comprehensiveness in covering the NGO supplier evaluation process in conflict zones.

      • Criteria for Validation:

      (I) Clarity: Questions were designed to be clear and unambiguous.

      (II) Relevance: All questions directly related to construction supplier selection and evaluation practices.

      (III) Comprehensiveness: The questionnaire comprehensively covered all aspects critical to NGO decision-making in supplier evaluation and selection.

      • Feedback and Revisions: Experts provided structured feedback suggesting necessary revisions to improve question focus. These changes were incorporated to enhance the supplier evaluation process.

      • Final Validation: The revised questionnaire underwent a final review to confirm all modifications were effectively integrated. Approval from all experts confirmed the questionnaire was validated and ready for deployment. The whole survey is provided in Appendix 1.

    • ➢ FGDs: FGDs serve as a pivotal tool for delving deeper into issues initially highlighted through surveys. These discussions enable a nuanced exploration of the perspectives and practices of non-governmental organizations (NGOs), facilitating the identification of both common themes and divergent viewpoints. This process is crucial for developing a robust evaluation framework. For instance, when surveys reveal a lack of consensus on specific points, FGDs are instrumental in fostering collective agreement. Through FGDs, it was determined that the maximum acceptable distance for allocating a supplier to a construction project is 100 km. This limit is based on experiences showing that greater distances complicate project execution and increase logistical challenges. Similarly, it was established that suppliers within this region could be allocated to a maximum of three construction projects. This restriction reflects the administrative capacities of local suppliers who often lack the robust administrative structures necessary to manage multiple projects effectively. These conclusions are drawn from the practical insights and experiences of humanitarian actors and donors who have implemented construction projects in the region.

  2. 2.

    Prioritizing the criteria: Utilizing the F-AHP, a decision-making framework that allows for the incorporation of human judgment and uncertainty. It is used to determine the relative importance of a set of criteria that NGOs consider when evaluating suppliers. The "fuzzy" aspect allows it to handle imprecision, which is often the case in a crisis environment.

  3. 3.

    GIS Spatial Analysis: Geographic Information System (GIS) is a tool used to capture, store, manipulate, analyze, manage, and present spatial or geographic data. In this context, it helps in identifying and visualizing risks associated with different geographical locations where suppliers might operate, contributing to a comprehensive risk map. Two risks are defined: the first is the frontline risk, and the second is the hard access areas risk.

  4. 4.

    Building the Fuzzy Model: This involves creating a model based on fuzzy goal programming. This model can handle the complexity and ambiguity of real-world desired targets in each objective, making it suitable for evaluating suppliers where information may be incomplete or uncertain.

  5. 5.

    Model Validation: Before using the model for decision-making, it is critical to ensure that it accurately reflects the real-world scenario it's intended to represent. This involves testing the model against numerical examples to confirm its reliability.

  6. 6.

    Solving the model: The model is solved using GAMS, a high-level modeling system for mathematical programming problems. It is used to find the best solution or the most optimal supplier according to the criteria and data fed into the fuzzy model.

  7. 7.

    Sensitivity Analysis: After determining the optimal supplier, a sensitivity analysis is performed. This process involves changing one or more parameters in the model to see how these changes affect the outcome.

3.1 Building the model

3.1.1 Model parameters

I:

Set of tenders; i ϵ I

J:

Set of candidate construction suppliers; j ϵ J

N :

Set of targeted objectives; n ϵ N

\({d}_{ij}\) :

Distance between candidate construction supplier j and another (acquired from building road network dataset by GIS)

\(S{M}_{ij}\) :

Submission matrix for tender i by each construction supplier j (not all supplier is submitting their offers to all tenders)

\({FO}_{ij}\):

Financial offer submitted by supplier j for each tender i

\({TVC}_{j}\) :

Total value of previous contracts for each candidate supplier j.

\({RFL}_{j}\) :

Risk at each candidate location of Candidate supplier j in terms of proximity to frontlines.

\({RHA}_{j}\) :

Risk at each candidate location of Candidate supplier j in terms of hard access areas.

\({QM}_{ij}\) :

Quality of materials submitted by candidate supplier j for each tender.

\({SE}_{j}\) :

Staff experience for Candidate supplier j.

\({DT}_{ij}\) :

Delivery time submitted by candidate supplier j for each tender.

\({PR}_{j}\) :

Promptness response of quality and delivery issues for Candidate supplier j.

\({CE}_{j}\) :

Capacity of each Candidate supplier j.

\({MC}_{j}\) :

managemental capacity of each Candidate supplier j.

\({FC}_{j}\) :

Financial capacity of each Candidate supplier j.

\({RL}_{j}\) :

Recommendation letter submitted by third party for each Candidate supplier j.

\({CL}_{j}\) :

Child labor involvement for each Candidate supplier j.

\({HP}_{j}\) :

Commitment to humanitarian principles for each Candidate supplier j.

\({EV}_{j}\) :

Commitment to environmental regulation and compliance for each Candidate supplier j.

\({k}_{n}\) :

Weights of each fuzzy goal n.

\({AL}_{n}\) :

Aspiration level for each fuzzy goal n.

\({\Delta MA}_{n}\) :

Maximum allowable deviation for each fuzzy goal n

3.1.2 Model decision variables

\(O_n\):

Amount of an overachieved for each fuzzy goal n

\(U_n\) :

Amount of an underachieved for each fuzzy goal n

\({\mu }_{n}\):

Degree of membership for each fuzzy goal n

3.1.3 Objective functions

$$Maximize \sum_{i\in I}\sum_{j\in J}{t}_{ij} {QM}_{ij}$$
(1)
$$Manimize \sum_{i\in I}\sum_{j\in J}{t}_{ij} {Fo}_{ij}$$
(2)
$$Maximize \sum_{i\in I}\sum_{j\in J}{t}_{ij} {TV}_{j}$$
(3)
$$Maximize \sum_{i\in I}\sum_{j\in J}{t}_{ij} {SE}_{j}$$
(4)
$$Minimize \sum_{i\in I}\sum_{j\in J}{t}_{ij} {DT}_{ij}$$
(5)
$$Maximize \sum_{i\in I}\sum_{j\in J}{t}_{ij} {PR}_{j}$$
(6)
$$Maximize \sum_{i\in I}\sum_{j\in J}{t}_{ij} {CE}_{j}$$
(7)
$$Maximize \sum_{i\in I}\sum_{j\in J}{t}_{ij} {MC}_{j}$$
(8)
$$Maximize \sum_{i\in I}\sum_{j\in J}{t}_{ij} {FC}_{j}$$
(9)
$$Maximize \sum_{i\in I}\sum_{j\in J}{t}_{ij} {RL}_{j}$$
(10)
$$Minimize \sum_{i\in I}\sum_{j\in J}{t}_{ij} {CL}_{j}$$
(11)
$$Maximize \sum_{i\in I}\sum_{j\in J}{t}_{ij} {HP}_{j}$$
(12)
$$Minimize \sum_{i\in I}\sum_{j\in J}{t}_{ij} R{FL}_{j}$$
(13)
$$Minimize \sum_{i\in I}\sum_{j\in J}{t}_{ij} {RHA}_{j}$$
(14)
$$Maximize \sum_{i\in I}\sum_{j\in J}{t}_{ij} {EV}_{j}$$
(15)

Equations (1) to (2) delineate the objectives of the proposed model. Equation (1) is designed to maximize the overall quality of materials submitted by candidate suppliers. In contrast, Eq. (2) focuses on minimizing the financial bids tendered by the construction suppliers to conduct the construction works in the area. Equation (3) seeks to maximize the aggregate value of previous contracts held by the chosen suppliers, while Eq. (4) aims to enhance the cumulative experience of their staff. Equation (5) is intended to minimize the total delivery time for completing the construction projects as proposed by the suppliers.

Subsequent equations, namely Eqs. (6) through (7), are directed towards maximizing various operational capacities of the selected suppliers. These include promptness in response (Eq. 6), equipment capacity (Eq. 7), managerial capacity (Eq. 8), financial stability (Eq. 9), and the quantity of recommendation letters (Eq. 10). Equation (11) is oriented towards minimizing the incidence of child labor among the selected suppliers.

Equation (12) addresses the commitment of suppliers to humanitarian principles, aiming for its maximization. Equations (13) and (14) target risk minimization related to the proximity of suppliers to frontlines and hard-to-access areas. Lastly, Eq. (15) is dedicated to maximizing adherence to environmental regulations by the selected construction suppliers.

Subject to:

$$\begin{array}{cc}\sum\limits_{i\in J}t_{ij}=1&\forall i\in I\end{array}$$
(16)
$$t_{ij}\leq SM_{ij} \quad \forall\;i\in I,\;\forall\;j\in J$$
(17)
$$\sum\limits_{i\in I}{t}_{ij}\le 3 \quad \forall j\in J$$
(18)
$$\sum\limits_{i\in I}{t}_{ij}{d}_{ij}\le 100 \quad \forall j\in J$$
(19)
$${t}_{ij} \in \left[\text{0,1}\right]$$
(20)

The hard constraints of this model are encapsulated in Eqs. (16) to (20). Equation (16) mandates that each construction sub-project is to be allocated to only one supplier. Equation (17) specifies that a project can only be assigned to candidate suppliers who have submitted an offer for that particular project, as indicated by the submission matrix. Equation (18) limits the assignment of each construction supplier to a maximum of three projects. Equation (19) imposes a geographical constraint, ensuring that the projects assigned to a supplier are within a maximum distance of 100 km. Finally, Eq. (20) defines the binary variables associated with the selection of suppliers.

In existing literature, a variety of methods have been proposed to address multi-objective problems (Ulungu et al. 1994; Aiello et al. 2006; Ye and Zhou 2007; Singh and Singh 2011; Xu and Li 2012; Hathhorn et al. 2013; Emami and Nookabadi 2013; Xu et al. 2016; Li et al. 2017) In this particular study, the problem was addressed using fuzzy goal programming, F-AHP, GIS-based risk assessment, humanitarian and environmental context of the Syrian crisis.

3.2 Fuzzy goal programming

Goal programming is an approach used for solving multi-objective optimization problems that balances trade-offs in conflicting objectives. It allows for balancing all desired objectives (from Eq. 1) to Eq. 15 through direct trade-offs between all unwanted deviational variables by placing them in a normalized single-achievement function that includes all the objective deviations in just one equation (Jones and Tamiz 2010). In this study, a fuzzy goal programming model has been utilized because it provides a more realistic approximation to real case studies. This model was presented by Yaghoobi et al. (2008). The model consolidates all the objective functions into a single objective function, as formulated in Eq. 21, where efforts are made to minimize the deviations from the desired goal values in each goal, taking into account the weight of each objective function.

Consecutively, we have converted each objective function into soft constraints by adding the deviations \({O}_{n}\)​ for those objective functions that we are striving to minimize, and \({U}_{n}\)​ ffor those objective functions that we are striving to maximize. The soft constraints are formulated in Eqs. 22, 24, 26, 28, 30, 32, 34, 36, 38, 40, 42, 44, 46, 48, and 50, while Eqs. 23, 25, 27, 29, 31, 33, 35, 37, 39, 41, 43, 45, 47, 49, and 51 are constraints that ensure the sum of the normalized negative deviations, normalized positive deviations, and the membership variable \({\mu }_{\text{n}}\)​ equals one. The fuzzy model is presented and validated by the Yaghoobi model (Yaghoobi et al. 2008).

$$Minimize \;Z=\sum_{n}{w}_{n}\left(\frac{{Ov}_{n}+{Un}_{n} }{{\Delta MA}_{n} }\right)$$
(21)
$$\sum\limits_{i\in I}\sum_{j\in J}{t}_{ij}\; {QM}_{ij}+{Un}_{1} \ge {AL}_{1}$$
(22)
$${\mu }_{1}+ \frac{{Un}_{1}}{{\Delta MA}_{1}}=1$$
(23)
$$\sum_{i\in I}\sum_{j\in J}{t}_{ij} \;{Fo}_{ij}-{On}_{2} \le {AL}_{2}$$
(24)
$${\mu }_{2}+ \frac{{On}_{2}}{{\Delta MA}_{2}}=1$$
(25)
$$\sum_{i\in I}\sum_{j\in J}{t}_{ij} {TV}_{j}+{Un}_{3}\ge {AL}_{3}$$
(26)
$${\mu }_{3}+ \frac{{Un}_{3}}{{\Delta MA}_{3}}=1$$
(27)
$$\sum_{i\in I}\sum_{j\in J}{t}_{ij} {SE}_{j}+{Un}_{4}\ge {AL}_{4}$$
(28)
$${\mu }_{4}+ \frac{{Un}_{4}}{{\Delta MA}_{4}}=1$$
(29)
$$\sum_{i\in I}\sum_{j\in J}{t}_{ij} {DT}_{ij}-{On}_{5}\le {AL}_{5}$$
(30)
$${\mu }_{5}+ \frac{{On}_{5}}{{\Delta MA}_{5}}=1$$
(31)
$$\sum_{i\in I}\sum_{j\in J}{t}_{ij} {PR}_{j}+{Un}_{6}\ge {AL}_{6}$$
(32)
$${\mu }_{6}+ \frac{{Un}_{6}}{{\Delta MA}_{6}}=1$$
(33)
$$\sum_{i\in I}\sum_{j\in J}{t}_{ij} {CE}_{j}+{Un}_{7}\ge {AL}_{7}$$
(34)
$${\mu }_{7}+ \frac{{Un}_{7}}{{\Delta MA}_{7}}=1$$
(35)
$$\sum_{i\in I}\sum_{j\in J}{t}_{ij} {MC}_{j}+{Un}_{8} \ge {AL}_{8}$$
(36)
$${\mu }_{8}+ \frac{{Un}_{8}}{{\Delta MA}_{8}}=1$$
(37)
$$\sum_{i\in I}\sum_{j\in J}{t}_{ij} {FC}_{j}+{Un}_{9} \ge {AL}_{9}$$
(38)
$${\mu }_{9}+ \frac{{Un}_{9}}{{\Delta MA}_{9}}=1$$
(39)
$$\sum_{i\in I}\sum_{j\in J}{t}_{ij} {RL}_{j}+{Un}_{10} \ge {AL}_{10}$$
(40)
$${\mu }_{10}+ \frac{{Un}_{10}}{{\Delta MA}_{10}}=1$$
(41)
$$\sum_{i\in I}\sum_{j\in J}{t}_{ij} {CL}_{j}-{On}_{11}\le {AL}_{11}$$
(42)
$${\mu }_{11}+ \frac{{On}_{11}}{{\Delta MA}_{11}}=1$$
(43)
$$\sum_{i\in I}\sum_{j\in J}{t}_{ij} \;{HP}_{j}+ {Un}_{12} \ge {AL}_{12}$$
(44)
$${\mu }_{11}+ \frac{{Un}_{11}}{{\Delta MA}_{11}}=1$$
(45)
$$\sum_{i\in I}\sum_{j\in J}{t}_{ij} R{FL}_{j} -{On}_{13}\le {AL}_{13}$$
(46)
$${\mu }_{13}+ \frac{{On}_{13}}{{\Delta MA}_{13}}=1$$
(47)
$$\sum_{i\in I}\sum_{j\in J}{t}_{ij} {RHA}_{j}-{On}_{14}\le {AL}_{14}$$
(48)
$${\mu }_{14}+ \frac{{On}_{14}}{{\Delta MA}_{14}}=1$$
(49)
$$\sum_{i\in I}\sum_{j\in J}{t}_{ij} {EV}_{j}+ {Un}_{15} \ge {AL}_{15}$$
(50)
$${\mu }_{15}+ \frac{{Un}_{15}}{{\Delta MA}_{15}}=1$$
(51)

3.3 Model validation

In the scholarly exposition of the case study, the validation of the proposed allocation model is meticulously articulated through a two-pronged data analysis approach. Initial validation is undertaken via synthetically constructed datasets, derived post-consultation with domain experts. This preliminary phase encompasses two distinct numerical examples: the first involving 10 construction projects and 20 potential suppliers, and the second encompassing 20 construction projects paired with 30 candidate suppliers. These scenarios are rigorously tested against varied weight assignments, prioritizing the dual objectives of optimizing material quality and minimizing project costs. The detailed datasets and the resultant computational outcomes are systematically documented in Appendix 2.

To elucidate the model's operational efficacy, Fig. 2 is presented, delineating a comparative analysis of the synthetic datasets. This visual representation accentuates the model's capacity to negotiate between the competing objectives of material quality and cost-efficiency. Notably, the graphic illustrations within Fig. 2 delineate a positive correlation between the assigned weight to material quality and the model's propensity to enhance this particular objective. Conversely, an augmented emphasis on cost reduction is reciprocated by the model's inclination to curtail financial expenditures. Collectively, these outcomes substantiate the model's robustness and its capability to deliver balanced solutions within the complex operational landscape of Northern Syria's reconstruction endeavors.

Fig. 2
figure 2

Model validation results (a, b, c, d)

Full size image

Subsequent to the synthetic trials, the model's validity is further corroborated through empirical data amassed from field surveys and focus group discussions within the specific contexts of Al-Bab and Ar-Ra'ee, Syria. The integration of real-world data provides a pragmatic dimension to the model's applicability, with a sensitivity analysis cementing its relevance. The congruence between the model's outcomes and the practical requirements observed in the field serves as a testament to its validity and effectiveness.

4 Results and discussions

4.1 Case study

In the midst of ongoing crises in Northern Syria, the tendering process for construction projects has evolved into a multifaceted challenge. Various suppliers, eager to contribute to the rebuilding efforts, have submitted proposals to undertake one or multiple projects dispersed across the northern region of Syria in two sub-districts (AlBab and Ar-Ra'ee). An NGO, acting as the steward of these efforts, is tasked with the rigorous evaluation of these proposals against a comprehensive set of criteria. These criteria span humanitarian considerations, risk assessment considerations, environmental impact, ensuring sustainability amidst reconstruction, and technical and financial competencies, alongside the capacity to effectively deliver on project commitments.

The evaluation matrix is composed of fifteen distinct criteria, carefully designed to holistically assess each supplier's offer based on data collected from stakeholders in the area. This systematic approach aims to align the selection process with overarching objectives by proposing a novel hybrid fuzzy model. As a result, this study strives to promote equitable development and adherence to environmental and humanitarian standards, achieving a transparent framework for all stakeholders in the crisis region.

The NGO will allocate each project to the supplier that demonstrates the highest congruence with the defined criteria described in Table 2, thus ensuring the optimal alignment of project needs with supplier capabilities. It is stipulated that a single supplier may be awarded a maximum of three projects, with the stipulation that the geographical distance of these projects does not exceed 100 km for one supplier. This constraint is imposed to ensure logistical feasibility and effective project oversight, as lessons learned from previous projects as a result of FGDs.

Table 2 Criteria utilized to evaluate and select construction supplier in conflict areas
Full size table

Through this case study, the article elucidates the operational complexities and the intricate decision-making processes involved in post-conflict reconstruction. The narrative underscores the necessity of a transparent, balanced, and multi-objective approach that interweaves diverse evaluation criteria to foster comprehensive development and stability in crisis-afflicted regions.

4.1.1 Environmental regulation and compliance

Four factors were considered during the visits to each construction supplier, as shown in Fig. 3:

  1. 1.

    Water Conservation: Using water-efficient construction techniques, as water resources might be scarce or contaminated in crisis regions.

  2. 2.

    Low-Impact Materials: Choosing construction materials that have minimal environmental impact, such as locally sourced materials, to reduce transportation emissions and support the local economy.

  3. 3.

    Waste Reduction: Implementing strategies to minimize construction waste and ensure proper disposal, as waste management systems in crisis areas might be compromised.

  4. 4.

    Energy Efficiency: Incorporating energy-efficient designs in construction to reduce the long-term environmental footprint, considering the limited energy resources in such areas or utilizing photovoltaic energy to produce electricity.

Fig. 3
figure 3

Factor considered in the process of Environmental Regulation and Compliance

Full size image

4.2 Findings

In this section, we present the results from applying the proposed hybrid methodology. We extracted the risk value for each candidate supplier by creating risk maps according to each criterion and its location on the related risk map, as shown in Fig. 4.

Fig. 4
figure 4

Risk values in the target area

Full size image

After calculating the risk value for each candidate supplier, we integrated this value into the mathematical fuzzy goal programming model as previously described. We solved the model using the software package, incorporating all fifteen goals to determine the optimal solution for the entire problem.

After calculating the risk value for each candidate supplier, we integrated this value into the mathematical fuzzy goal programming model as previously described. We solved the model using the software package, incorporating all fifteen goals to determine the optimal solution for the entire problem.

The inclusion of more criteria in the model enhances its transparency for the affected populations and suppliers within the humanitarian context and yields more varied values. This results in a marked variance between the values of candidate suppliers, which facilitates the selection process for decision-makers.

The results obtained from solving the proposed hybrid fuzzy model revealed the optimal solution achieved for each goal. We have presented these in Table 3 to compare the actual values against the planned values (identified by three dedicated experts considering humanitarian aspects and similar previous projects in northern Syria). Additionally, we depict the results in Fig. 5, showcasing the actual achieved values for each objective as a percentage and obtained value.

Table 3 Planned goals vs actual as percentage
Full size table
Fig. 5
figure 5

Planned goals vs actual

Full size image

For example, our cost target was approximately $300,000, but the actual value achieved was $321,526, reflecting an increase of 7%, which is considered relatively satisfactory as we strive to minimize this goal. While our aim was to reach a material quality scale of 66, we achieved a higher scale number of 94, which is approximately 44% higher. Our target for allocating suppliers with previous contracts was around $1,200,000, but after running the model, we achieved a very close value of $1,276,414. Throughout the model, we aimed to achieve 66 for staff experience, 150 for total delivery time to complete the projects, and scales of 66 for promptness response, equipment capacity, managerial capacity, financial capacity, recommendation letters, 40 for child labor involvement, 60 for commitment to humanitarian principles, 30 for risks related to frontlines, 30 for risks related to hard-to-access areas, and 60 for commitment to environmental regulations. In the results, we obtained 67, 260, 66, 67, 66, 94, 78, 44, 71, 26, 35, and 61, respectively.

The potential reasons for obtaining values that deviate significantly from the target values can be described as follows:

  • The decision-makers in the target area consider previous projects as a baseline and attempt to predict optimistic targets for this project based on that baseline.

  • Occasionally, the model may have already achieved the optimal values related to a specific criterion, indicating that there is no further possibility to improve the solution.

  • The model consistently strives to balance the achievement of goals according to weights determined by the F-AHP and does not focus on any single criterion in isolation.

4.3 Sensitivity analysis

4.3.1 Scenario 1: financial offers changing

In this Scenario, adjustments were made to the financial offers in response to the unstable market conditions in the area. Starting with a planned value of $300,000, the model endeavored to minimize this financial goal. A resultant value of $305,449.7 was obtained at a -5% change, as illustrated in Fig. 6. When transitioning from a 0% to a 5% change, there was an observed increase in the financial offers. However, this increase still remained less than the 5% threshold relative to the 0% change scenario. This indicates that the model consistently prioritizes minimizing financial costs while simultaneously considering the fulfillment of other objectives.

Fig. 6
figure 6

Changes of financial offer

Full size image

4.3.2 Scenario 2: some projects and suppliers are outside the calculation because of its location withing the frontline region

In this scenario, stakeholders acknowledge the possibility that some areas on the frontlines may become uncontrollable. This implies that construction projects and suppliers in these areas would be excluded from consideration, necessitating a resolution of the problem without their involvement. In this case, the projects and suppliers in the frontline area can be characterized as follows:

  • The construction suppliers: S4, which was not selected under normal circumstances, and S5, S6, and S7, which were selected in the normal situation.

  • Project P3 is also excluded from consideration.

The results obtained after solving the model under these conditions are presented in Table 4. It is observed that each project is allocated once, and the selected suppliers are assigned to a maximum of two construction projects each. In Table 5, the achieved goals are compared against the planned values. The solution in this altered scenario is less favorable compared with the normal situation, particularly concerning the first three most critical factors (G1, G2, G3). This outcome is expected due to the exclusion of three suppliers initially chosen in the normal situation from Case 2.

Table 4 Supplier selection and allocation in case 2
Full size table
Table 5 Goals planned vs achieved in case 2
Full size table

4.4 Implications for theory and practice

4.4.1 Theoretical implications

This research enriches the theoretical landscape by blending diverse methodologies, notably integrating Fuzzy Goal Programming, GIS-based risk assessment, and the F-AHP. This multifaceted approach not only advances the understanding of MCDM but also tailors it specifically to the nuanced requirements of post-conflict reconstruction scenarios. This result is supported by Govindan et al. (2020), who emphasized the importance of using multiple methodologies to enhance decision-making processes. Such a synthesis is pivotal in offering a comprehensive framework that not only addresses but also adapts to the evolving complexities inherent in construction supplier evaluation within such zones. Additionally, the incorporation of real-world data and case studies enhances the theoretical relevance of the model, grounding abstract concepts in tangible scenarios that reflect the current challenges faced in supply chain management within conflict-impacted environments.

4.4.2 Managerial implications

From a practical standpoint, this study provides robust tools for improving decision-making in crisis situations. By systematizing the evaluation and allocation of suppliers to construction projects, the model underscores the critical importance of integrating risk and humanitarian considerations into procurement strategies. Such insights are invaluable for NGOs, donors, and local authorities engaged in reconstruction efforts, offering them a methodologically sound approach to enhance their operations. The application of this model not only promises enhanced efficiency and effectiveness in project implementations but also fosters greater transparency and accountability in environments that traditionally suffer from high uncertainty and risk.

Furthermore, the strategic recommendations outlined in this study serve as a guiding framework for entities involved in reconstruction efforts. By adopting the proposed methodologies, these organizations can better navigate the complexities of supplier selection and project allocation in ways that align with both immediate project goals and long-term developmental objectives. This dual focus on operational efficiency and strategic foresight exemplifies the practical applications of the research, providing a scalable and adaptable solution that can be customized for various conflict-affected regions worldwide.

5 Conclusion

This study has developed an innovative hybrid methodology aimed at optimizing supplier selection and assignment for construction projects critical to the reconstruction efforts in Syria post-crisis and following the February 2023 earthquake. By integrating Geographic Information Systems (GIS) and the F-AHP, the approach effectively identifies the risk values associated with each potential supplier, facilitating the selection of the most suitable entities. The subsequent application of fuzzy goal programming ensures that the selection and assignment processes are guided by a robust multi-objective optimization framework, which incorporates a comprehensive set of criteria covering technical, financial, humanitarian, and risk aspects.

The findings of this research significantly contribute to both theoretical understanding and practical applications in the field of crisis management and reconstruction. Theoretically, it bridges the gap in multi-criteria decision-making models by incorporating complex and dynamic environments such as conflict zones. Practically, it offers a transparent, systematic framework that enhances decision-making in supplier selection and project assignments, which is critical for effective reconstruction efforts. The real-case application using data collected from northern Syria not only demonstrates the methodology’s applicability and effectiveness but also its adaptability to other conflict-affected regions globally.

Addressing the research questions posed at the outset, the study highlights that a diverse array of supplier selection criteria, previously under-considered in conflict zones, are indeed crucial for the construction industry. The integration of GIS-based risk analysis and multi-criteria decision-making tools like F-AHP has proven effective in enhancing the efficiency and outcomes of construction projects in such challenging environments. These methodologies allow for a nuanced consideration of various risk factors and ensure that the most capable suppliers are selected and assigned to projects that they are best suited for.

Despite its innovative approach and contributions, the study acknowledges certain limitations, such as its focus on a relatively small geographic area and the omission of some potential risk factors. Future research could address these limitations by expanding the geographical scope of the study and incorporating more dynamic models that account for additional risk factors and real-time data. This would not only enhance the robustness of the model but also its applicability to a wider array of scenarios in conflict-affected regions.

In conclusion, this study not only enhances our understanding of supplier evaluation in post-conflict reconstruction but also contributes to more effective resource utilization, transparency in decision-making, and ultimately, the speedy recovery of affected communities. By continuing to refine and adapt this approach, it holds significant promise for aiding reconstruction efforts not just in Syria, but in any region emerging from crisis.