Volumetric videos are three-dimensional video representations of objects or scenes. Unlike traditional 2D videos that only show a flat representation, volumetric videos allow for an immersive experience where the viewer can observe the object or scene from different perspectives. This is achieved by capturing data from multiple cameras, which is then converted into a 3D model. This 3D model is transmitted in real-time to a remote output device, allowing real-world environments to be experienced virtually in other locations. Automated segmentation of sub-models provides support in identifying relevant areas for the user. Additionally, distinguishing between dynamic and static sub-scenes can accurately depict a dynamic world.

In an industrial context, for example, the human operator can additionally receive direct support from a remotely connected expert in a virtual 1:1 on-site presence. The situation on the assembly line is recorded and transmitted to the remote expert using innovative volumetric reconstruction methods. Conversely, the expert can explain his or her recommendations to the worker in direct virtual presence and provide assistance either by means of direct actionable instructions or virtual annotations on the product itself. The system is scalable and supports assembly workers via mobile or head-mounted XR (extended reality) output units. Furthermore, this telepresence technology can be used for training purposes, teaching, or remote maintenance.