As part of the digitization of the working world, manufacturing that will become more and more flexible or even autonomous, it is essential to be able to simulate real, continuously changing processes in the virtual world.
Our R&D activities join the real world and the virtual world to form “digital twins” of physical environments by capturing static and dynamic scenes with cameras that continuously transmit the current status of the real world to the virtual world. This could even be used to capture a production system. The generated images and point clouds are segmented to extract relevant objects in 2D or 3D. Object detection is based on machine learning processes, and our object detection processes can help collaborating robots interpret their environment and decide how they should respond. The digitized version of a real scene can be enriched with animated, virtual models, allowing for uses such as implementing plans based directly on data from dynamically recorded reality.