An earned-value-analysis (EVA)-based project control framework in large-scale scaffolding projects using linear regression modeling

Abstract

In large-scale industrial construction projects, scaffolding activities account for a large amount of the construction budget, and overlooking the scaffolding management can lead to budget overruns and schedule delays. The scaffolding activities can be categorized by classifications and types based on the nature of the scaffold builds. To ensure the project progress on track, it is critical to measure project performance based on project progress data. However, given the nature of scaffolding activities, it has been challenging to track and utilize the scaffolding data for analytical purposes. Therefore, this paper proposes a project control framework based on Earned-value analysis (EVA), in which linear regression models are used for productivity prediction. Three scenarios of productivity based on historical data (i.e., low, medium, and high productivity) are introduced. The proposed framework is implemented in a real construction project for validation. The results have shown that the proposed framework can efficiently evaluate the project progresses integrated with the EVA. The construction companies, such as general contractors and scaffolding sub-contractors, can use this method for site progress tracking. For future work, the EVA can be integrated with other non-linear predictive models (e.g., neural network) for productivity prediction. The EVA results can be integrated with data visualization to create situational awareness for construction practitioners.