Digital twin technology is pivotal in the transition of manufacturing industries towards Industry 4.0, as it enables the creation of virtual representations of physical shop floors and production processes. This technology addresses manufacturing challenges by allowing the reuse and adjustment of production equipment in real-time, facilitating novel technologies, and supporting the adoption of flexibilities to add new product variants. The significance of resource-efficient and flexible production systems is highlighted by their ability to optimize resource utilization and enable reconfiguration through digital models. This study specifically investigates the differences between physical systems and their digital twins, focusing on the sustainable updating of virtual models of a flexible automation cell. Digital models of the flexible automation cell are acquired using 3D laser scanning techniques, capturing data as point clouds. The differences between new point cloud models and existing digital models are analyzed using CloudCompare software. Identified changes are extracted from the digital models as point clouds and converted into 3D mesh models through surface reconstruction techniques, thereby updating the digital twin. To address inaccuracies in the detailed extraction of digital models compared to physical models, an additional fusion step is implemented. This step integrates data from photogrammetry and 3D laser scanning, enhancing the point clouds and producing accurate 3D models of the automation cell. The main focus of this study is to determine the most effective approach for scanning an automation cell and identifying changes by comparing two digital models, thereby contributes to the field of digital twin technology with a novel methodology for sustainable virtual model updates.