The proposed schedule-update system consists of two parts. First, the actual construction progress of each activity is measured. For this purpose, collected as-built data for the building under construction are compared with the 4D as-planned model. The comparison consists of 3D registration and object matching. The 3D registration process used in this study improves upon that proposed by Kim et al. (2013a). Their automated 3D registration method can be applied efficiently to small point clouds, whereas this study’s proposed method can align a large set of as-built data with an asplanned model. To determine whether or not a given structural component has already been constructed—based on a comparison of the aligned as-built data with the as-planned model—the proposed method applies the object-matching method devised by Kim et al. (2013a), which is an improvement over the methods proposed by Bosch e et al. (2009) and Golparvar-Fard et al. (2009), even in the presence of occlusions. In the study by Kim et al. (2013a), a revision process was proposed—based on the sequence of activity execution and the topological connectivity between structural components— to modify the inaccurate as-built status of a component if that inaccuracy was due to an incomplete data set. For more details regarding the measurement process, please see Kim et al. (2013a).