نوع مقاله : مقاله پژوهشی
عنوان مقاله English
نویسندگان English
Rework in large concrete dam projects (LCDP) is a costly and delay-inducing phenomenon, arising from errors, design changes, or execution defects, which severely threatens the quality, productivity, and safety of the structure. This research employs a novel hybrid multi-criteria decision-making (MCDM) and machine learning (ML) approach to identify, weight, and prioritize the factors influencing rework. Data were collected from 26 expert specialists using a two-level questionnaire (10 main criteria and 100 sub-criteria) based on a Likert scale. The MCDM weighting was performed using four methods (Simple Multi-Attribute Rating Technique - SMART, Shannon's Entropy, Criteria Importance Through Inter Criteria Correlation, and Rank Sum) and the geometric mean. This process identified the top three criteria as Design and Technology (weight ≈0.31), Geological and Environmental (0.18), and Legal and Contractual (0.12). The top ten sub-criteria included groundwater infiltration, high cost of digital technologies, cybersecurity issues, defects in concrete compaction, and political pressures (with a low weight gap of ≈0.0008). Among methods, Partial Least Squares Regression (PLSR) demonstrated the best performance with a test mean squared error of 0.0028 and a test R² of 0.9984. Regularized Bayesian regression and Ridge regression models, with R² > 0.997, ranked next. The predicted importance from the top ML models confirmed that geotechnical factors (e.g. groundwater infiltration: ≈3.9), concrete quality (compaction: ≈5.6), and digital technologies (high cost: ≈3.3) were the most critical, showing complete alignment with the MCDM weights. The integrated hybrid approach increased prediction accuracy to 99.8%, providing a dynamic tool for proactive project management.
کلیدواژهها English
Rework is widely recognized as one of the most critical challenges in large-scale construction projects, particularly in complex infrastructure systems such as concrete dams. Rework refers to the need to redo or correct completed construction activities due to errors, omissions, design changes, quality deficiencies, or unexpected environmental conditions. In dam construction projects, rework can lead to significant financial losses, schedule delays, inefficient resource utilization, and even structural safety risks. Given the enormous investments associated with dam projects and their long service life, minimizing rework is essential for ensuring construction efficiency, structural reliability, and long-term sustainability. Concrete dams represent highly complex engineering systems that involve multidisciplinary design, extensive geotechnical investigations, large-scale material placement, and strict quality control processes. The presence of uncertainties related to geological conditions, groundwater flow, material performance, and construction methods further complicates project execution. As a result, even minor deviations in design, planning, or execution may trigger extensive rework activities that affect project performance. Previous studies have examined various causes of rework in construction projects, including design deficiencies, inadequate communication among project stakeholders, poor construction practices, contractual ambiguities, and environmental uncertainties. However, most existing approaches rely heavily on subjective assessments or single analytical techniques, which may not fully capture the complex relationships among the multiple factors influencing rework. Consequently, there is a need for integrated analytical frameworks that combine expert knowledge with advanced computational techniques in order to provide more accurate and reliable decision support. In response to this need, the present study proposes a hybrid analytical framework that integrates multi-criteria decision-making (MCDM) techniques with machine learning (ML) models to systematically identify and prioritize the key factors contributing to rework in concrete dam projects. The main objective of this research is to determine the relative importance of various technical, managerial, environmental, and contractual factors affecting rework and to develop a predictive framework capable of validating and reinforcing the prioritization results. More specifically, the study aims to: identify the critical criteria and subcriteria associated with rework in concrete dam construction; evaluate and weight these factors using multiple MCDM techniques; analyze the relationships among the identified variables through machine learning models; and ultimately provide a comprehensive and data-driven decision-support tool for project managers, engineers, and policymakers involved in dam construction projects. By integrating expert judgment with advanced data-driven modeling, the research seeks to enhance the accuracy of rework risk assessment and to support more effective preventive management strategies in large-scale hydraulic infrastructure projects.
The research methodology employed in this study is based on a hybrid analytical approach combining multi-criteria decision-making techniques and machine learning algorithms. The study relied on expert knowledge obtained from professionals with substantial experience in dam construction and related engineering fields. A total of 26 experts participated in the study, representing various professional backgrounds including structural engineering, geotechnical engineering, project management, quality control, and construction supervision. Their expertise ensured that the collected data reflected diverse perspectives on the causes and implications of rework in concrete dam projects. To systematically capture expert opinions, a structured two-level questionnaire was developed. At the first level, ten main criteria were identified based on a comprehensive review of the literature and consultations with domain experts. These criteria represented the primary categories of factors that may contribute to rework in dam construction projects, including design and technological factors, geological and environmental conditions, legal and contractual aspects, construction quality issues, managerial and organizational factors, and other related domains influencing project performance. At the second level, a total of one hundred detailed subcriteria were defined under the ten main criteria. These subcriteria represented specific factors that may lead to rework during different stages of dam construction. Examples of such factors include groundwater infiltration into the dam foundation, insufficient geological investigations, inadequate drainage design, poor concrete compaction, segregation or voids in concrete placement, design inconsistencies, high costs associated with digital construction technologies, cybersecurity risks in digital systems, political pressures affecting project decisions, and contractual conflicts between stakeholders. Experts evaluated the importance of each subcriterion using a Likert-scale rating system. The collected responses were then converted into quantitative values for further analysis. In order to determine the relative importance of the criteria and subcriteria, four different multi-criteria decision-making methods were employed. These methods included the Simple Multi-Attribute Rating Technique, Shannon entropy weighting, a correlation-based importance assessment method, and the rank centroid technique. Each of these methods provides a different perspective on the weighting process, thereby reducing methodological bias. To obtain a unified set of weights, the results of the four methods were aggregated using the geometric mean. This integration process produced a robust and balanced set of weights representing the relative importance of the criteria and subcriteria influencing rework in concrete dam projects. In the second stage of the analysis, machine learning models were developed to evaluate the predictive capability of the identified factors and to verify the results obtained from the MCDM analysis. A dataset was constructed using the expert evaluation scores as input variables. Twelve different machine learning regression models were trained and evaluated in order to identify the most accurate predictive approach. Among these models, particular attention was given to Partial Least Squares Regression (PLSR), Ridge Regression, and Regularized Bayesian Ridge Regression due to their effectiveness in handling high-dimensional datasets and correlated predictors. The models were trained using appropriate training datasets and subsequently validated using testing datasets to assess their predictive accuracy. Model performance was evaluated using statistical indicators such as mean squared error (MSE) and the coefficient of determination (R²). In addition to prediction accuracy, the relative importance of variables identified by the machine learning models was analyzed and compared with the weights obtained from the MCDM analysis in order to evaluate the consistency of the results.
The results of the multi-criteria decision-making analysis revealed clear patterns in the relative importance of the factors contributing to rework in concrete dam projects. Among the ten main criteria evaluated in the study, design and technological factors were identified as the most influential category, with an approximate weight of 0.31. This finding highlights the critical role of accurate design documentation, advanced engineering technologies, and effective coordination among design disciplines in preventing construction errors and minimizing rework. The second most significant category was geological and environmental factors, which obtained a weight of approximately 0.18. This result reflects the importance of accurate geological investigations and appropriate management of environmental conditions in dam construction projects. Geological uncertainties, groundwater conditions, and environmental constraints may significantly influence the stability and constructability of dam structures, thereby increasing the likelihood of rework if not properly addressed during the planning and design stages. Legal and contractual factors ranked third among the main criteria, with a weight of approximately 0.12. This finding emphasizes the influence of contract management, regulatory requirements, and legal frameworks on project execution. Issues such as contractual ambiguities, change orders, and disputes between project stakeholders can lead to modifications in construction activities, which often result in rework and additional project costs. At the subcriteria level, several factors emerged as particularly critical contributors to rework. Groundwater infiltration into the dam foundation or structural components was identified as one of the most influential technical factors. Similarly, deficiencies in concrete compaction and placement practices were found to significantly increase the likelihood of rework due to their impact on structural integrity and durability. In addition, technological challenges such as the high costs of implementing digital construction technologies and concerns related to cybersecurity in digital project management systems were identified as emerging factors affecting modern construction environments. The machine learning analysis further reinforced the findings of the MCDM approach. Among the twelve evaluated models, Partial Least Squares Regression demonstrated the highest predictive accuracy, achieving a test mean squared error of approximately 0.0028 and a coefficient of determination of approximately 0.9984. These results indicate an extremely strong predictive relationship between the identified factors and the occurrence of rework. Ridge Regression and Regularized Bayesian Ridge Regression also demonstrated high levels of predictive performance, with coefficients of determination close to 0.997. The analysis of feature importance within these models revealed strong agreement with the priorities obtained from the MCDM analysis. In particular, geotechnical factors such as groundwater infiltration, construction quality factors such as inadequate concrete compaction, and technological challenges such as high implementation costs of digital systems were consistently identified as the most influential predictors of rework.
The consistency between the results of the MCDM analysis and the machine learning models confirms the reliability of the proposed hybrid framework. By combining expert-based decision analysis with data-driven predictive modeling, the study provides a comprehensive understanding of the complex relationships among the factors influencing rework in concrete dam construction projects.
The findings of this study demonstrate that the integration of multi-criteria decision-making techniques and machine learning models provides a powerful analytical framework for identifying and prioritizing the factors affecting rework in concrete dam projects. The results indicate that design and technological issues, geological and environmental conditions, and legal and contractual factors are the most influential categories contributing to rework in such projects. At a more detailed level, specific factors such as groundwater infiltration, inadequate concrete compaction, technological implementation challenges, and contractual complexities play significant roles in increasing the likelihood of rework during dam construction. These findings highlight the importance of comprehensive geological investigations, rigorous quality control in concrete placement, improved coordination among design and construction teams, and clear contractual frameworks in order to minimize rework risks. The machine learning analysis confirmed the robustness of the prioritization results and demonstrated the high predictive capability of the proposed framework. The superior performance of the Partial Least Squares Regression model indicates that linear regression models with appropriate regularization can effectively capture the relationships among complex engineering variables even when the available dataset is relatively small. Overall, the hybrid MCDM–ML approach developed in this research offers a practical decision-support tool for engineers, project managers, and policymakers involved in dam construction projects. By providing a systematic and data-driven method for evaluating rework risk factors, the framework can help improve project planning, reduce unnecessary construction modifications, and enhance the overall efficiency and sustainability of dam infrastructure development. Future research may expand this framework by incorporating larger datasets from real construction projects, integrating additional machine learning techniques, and exploring the applicability of the proposed approach to other types of large-scale infrastructure projects such as tunnels, bridges, and hydropower facilities. Such developments could further strengthen the ability of decision-makers to manage construction risks and improve project performance in complex engineering environments.
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
The study was funded by Urmia University, Country Iran.
All authors contributed equally to the conceptualization of the article and writing of the original and subsequent drafts. All authors have read and agreed to the published version of the manuscript.
Statement: there is nothing to disclose.
Data available on request from the authors.
The authors would like to thank all participants in the present study.
The authors would like to thank anonymous reviewers for their valuable suggestions in manuscript revision.
The authors avoided data fabrication, falsification, and plagiarism, and any form of misconduct.
The authors declare no conflict of interest.
