One of the most important means for digital transformation is the Digital Twin which is the most exact representation of the manufacturing process in a virtual model. The Digital Twin provides many advantages for the optimization of production processes, but it is currently still used very rarely. This is primarily due to the fact that companies are reluctant to incur generation costs and do not have the required IT know-how. One approach to overcome these obstacles is to simplify the generation of the Digital Twin by using object recognition. This can be used to automatically scan a production system and transfer it to a model. However, suitable methods of object recognition are needed to achieve added value. Suitable acquisition methods are also necessary to compensate the impact of darkness, dirt and occlusion. This paper provides a comprehensive overview of the recent advances in 3D object recognition of indoor objects using Convolutional Neural Networks (CNN). The comparison of main recognition methods based on methods of geometric shape descriptor and supervised learning and their strenghts and weakness is also included. The focus lies on the specific requirements and constrains in an industrial environment like tight assembly, light, dirt, occlusion, incomplete data sets.