Data-driven model for delay estimation in construction projects

The global construction industry continues to face persistent delays, a complex challenge that varies across projects and organizations. This study aims to develop a data-driven model for predicting delays in construction projects. A regression-based model is proposed, and its performance is compared to that of the K-Nearest Neighbors (KNN) algorithm. Several key factors contributing to construction delays are identified through a comprehensive review of existing literature and consultations with industry experts. The correlations between these factors and project delays are analyzed to understand how they interact and collectively impact project timelines. The results reveal that clearly defined goals and objectives, the availability of consultant resources, the number of internal stakeholders, and the project’s location are among the most influential factors. The proposed models are validated using real-world data, and the findings show that the multiple linear regression model outperforms the KNN algorithm, achieving an R² value of 95%. This high level of accuracy demonstrates the model’s effectiveness in explaining variance in project delays. The study presents a practical tool that enables project managers and stakeholders to better anticipate and manage potential delays. By facilitating more informed decision-making, the model supports more effective planning, resource allocation, and overall project delivery.