In recent years, 3D reality capture has become the preferred method of capturing the as-built condition of any physical asset. Millions of bits of data are converted into 3D point clouds, creating an accurate virtual representation of what was scanned. This data can then be used to create highly accurate models that can be used by designers, construction managers, and engineers. However, even a modest project can generate hundreds or thousands of scans; each one requiring point cloud registration. Not only must scans be taken and registered the first time around; throughout an asset’s lifespan, it is in exactly these situations that targetless or cloud-to-cloud registration can run into challenges.