EXPLORING AND MEASURING PROJECT COMPLEXITY IN METRORAIL PROJECTS

Abstract

Project complexity is one of the most common issues faced by metro rail projects due to its complex and interdependent characteristics. The challenging characteristics necessitate a study to identify and measure the impact of project complexity in metro rail projects. Hence, its strategic importance lies in enhancing project management and decision-making processes in the context of metro rail projects. This research provides valuable insights for the construction industry, effectively addressing and navigating complexity in metro rail projects. Initially this research identifies and analyzes the interdependence of Project complexity factors (PCFs) in metro rail projects using the Decision-Making Trial and Evaluation Laboratory (DEMATEL). The study provides qualitative and quantitative analysis of project complexities factors and their relationships. This study employs a case-based method for identifying PCFs and a DEMATEL method for analyzing the interdependence of complexity factors in metro rail projects. Initially, PCFs were identified through an extensive literature review. To validate and refine these factors, semi-structured interviews were conducted with thirty experienced professionals, each having 5-20 years of experience in roles such as project management, engineering, and planning. Further, elevated, and underground metro rail projects were purposefully selected as cases, for identifying the similarities and differences in PCFs. A questionnaire survey was conducted with various technical experts in metro rail projects. These experts rated the impact of PCFs on a five-point Likert scale, for the evaluation of the interdependence of PCFs. The DEMATEL technique was used to analyze the interdependencies of the PCFs. Later the study addresses the impact of project complexity on the performance parameters like time, cost, scope, quality, sustainability, and reliability. Machine learning (ML) was used as a powerful tool to predict the impact of project complexity. Despite the recognized challenges, there has been limited research utilizing ML models to assess the impact of project complexity in metro rail projects. An integrated predictive model was developed by combining three different ML algorithms: Support Vector Machine (SVM), Random Forest (RF), and Decision Tree (DT). This combined approach shows the unique strengths of each algorithm to create a more comprehensive and robust predictive model. This study aimed to analyze how project complexity influences key project performance parameters, including time, cost, scope, quality, sustainability, and reliability in metro rail projects. The integrated model showed improved performance compared to using each algorithm separately, indicating its potential for i delivering interactive results in predictive modeling. It accurately predicted time, scope, and cost, showcasing the model's robustness in predicting their impacts. However, challenges arose when predicting quality and sustainability, given their complex and multifactored influences. This model contributes to a better understanding and precise prediction of the impact of complexity in metro rail projects. BWM, a robust Multi-Criteria Decision-Making technique, was used to prioritize key complexities, and a PCI model was developed. Further, the developed PCI was validated through case studies and sensitivity analysis was performed to check the accuracy and applicability of the developed PCI model. The analysis revealed that the location complexity exerted the most substantial influence on project performance, followed by the environmental, organizational, technological, and contractual complexities. Sensitivity analysis revealed the varying impacts of complexity indices on the overall project complexity. The existing studies on project complexity identification and quantification were limited to megaprojects other than metro rail projects. Efforts to quantitatively study and analyze the impact of project complexity in metro rail projects are left unattended. The developed PCI model and its validation contribute to the field by providing a definite method to measure and manage complexity in metro rail projects.