Automated model-based recognition of progress using daily construction photographs and IFC-based 4D models

Accurate and efficient tracking, analysis and visualization of the as-built status of projects are critical components of a successful project monitoring. Such information directly supports control decision-making and if automated, can significantly impact management of a project. In this paper, we describe an innovative automated approach for recognition of physical progress from unordered daily construction photologs and Building Information Models (BIMs). First, we describe an automated approach to photo-realistically reconstruct the as-built in 4D and label scene for occupancy. The IFC-based BIM is subsequently fused into the as-built scene by a registration-step and is traversed and labeled for visibility. Finally a machine learning scheme built upon a Bayesian probabilistic model is introduced that automatically detects physical progress in presence of occlusions and demonstrates that progress tracking at schedule activity-level could be fully automated. We present results on automated tracking, analysis and visualization of two ongoing building projects.