Machine Learning-Based Construction Planning and Forecasting Model

Abstract

Construction planning and scheduling are crucial aspects of project management that require a lot of time and resources to manage effectively. Machine learning and artificial intelligence techniques have shown great potential in improving construction planning and scheduling by providing more accurate insights into project progress and forecasting. This paper proposed a machine learning model that utilizes regularly updated site data to generate predictions of quantity variances from the plan and enable a more accurate forecast of future progress based on historical data on concrete activities. Also, the outputs of this model can be used when creating a schedule for a new project. New schedules created with the help of this model will be more consistent and reliable due to its vast data pool and ability to generate realistic forecasts from this data. The model utilizes data from completed and other ongoing projects to generate insights and provide a more accurate and efficient construction planning and scheduling solution. Within the scope of this study, different attributes of concrete pouring activities of different projects and locations were used as input data for a machine learning process, and then, using this model on test data, the forecast concrete quantities were obtained. This model provides a more advanced solution than traditional project management tools by incorporating machine learning techniques while significantly improving construction planning, scheduling accuracy, and efficiency, leading to more successful projects and increased profitability for construction companies