Publications / 2025 Proceedings of the 42nd ISARC, Montreal, Canada
This paper presents an automated workflow for weekly construction progress reporting, streamlining data integration and analysis. The proposed approach focuses on three key areas: planned activities, weekly performance, and projected progress. Inputs include an updated baseline schedule and weekly inspection data, such as images. Outputs, generated autonomously, provide project status and performance metrics, including Earned Value and Planned Value. Leveraging multimodal large language models (MLLMs), the system processes text and images, enabling seamless data integration. Key contributions include a simplified, reliable reporting process that reflects actual construction execution and planning while reducing time and resource demands. The paper also addresses implementation challenges, AI-driven solutions, and scalability for broader construction reporting applications.