Towards Automated Progress Tracking of Erection of Concrete Structures

One of the main criticisms of the construction industry is that projects are too often completed behind schedule (and/or with cost overruns). Schedule delays may result from poor planning, but also from poor progress control, because, if progress deviation is identified too late, then actions can often not be taken to avoid the impact of these delays on the overall project schedule. Progress tracking of erection of concrete structures in particular is a very demanding task requiring intensive data collection. It is because erection of concrete structures involves many steps like erection of scaffolding, formwork and rebar assemblies, concrete placement, and removal of scaffolding and formwork. Current manual tracking methods, based on foremen daily reports, are typically time consuming and/or error prone. Three dimensional (3D) Laser Scanners (LADARs) are capable of capturing and recording the 3D status of construction sites with high accuracy in short periods of time and have thus the potential to effectively support progress tracking. An automated object recognition system has recently been developed to recognize project 3D CAD model objects from site laser scans. A novel system is proposed here which combines this 3D object recognition system with architect and engineer provided BIM and schedule information into a 4D object recognition system, with a focus on progress tracking. This new system improves the one originally proposed by Bosche et al. (2009). It is demonstrated with real life data acquired over the course of construction of the new Engineering V Building at the University of Waterloo.