This research addresses challenges in construction project progress monitoring by introducing an innovative methodology. The Activity Level Progress Monitoring System (ALPMS) utilizes construction site images and a 4D Building Information Model (BIM) to create a Digital Twin information system. It employs deep learning-based semantic segmentation on BIM-registered orthographic images to estimate activity-wise completion percentages, enhancing progress reporting beyond binary forms. 

Automating project schedule updates involves linking reality models with construction schedules. Reality models are aligned using 3D BIM or control points, followed by segmentation to identify building components. Activity progress is calculated and stored with LOT (location-object-task) data. Advanced NLP models, such as LLMs, extract LOT details from schedules, enabling automated updates. Integrating BIM and NLP with innovations like few-shot learning and prompt engineering offers significant potential for streamlining construction progress tracking and schedule management.

Real-time construction monitoring integrates SLAM-based navigation, AI-powered progress detection, and AR visualization. A robotic system captures site data, aligning it with BIM for accurate tracking. AI-driven segmentation measures progress, while AR devices display updates. Multiple team members access the model for real-time collaboration, enabling timely decisions and corrective actions.