Creating Digital Twins of an indoor environment from an operator’s perspective

Abstract

Abstract In recent times, Virtual Commissioning (VC) has flourished in developing new manufacturing systems while ensuring high-quality, and flexible systems with reduced time and cost requirements. One application of VC is Virtual Reality, a technology that can enhance the process with more realistic visualization and validation. The highest level of VC can be achieved by utilizing a Digital Twin in the form of a 3D virtual map of an environment, however gathering a high-resolution, reliable, and accurate virtual map is crucial. Amongst various sensors, range sensors like LiDARs are advantageous in providing depth information of the surroundings, and further, the solid-state LiDARs produce high-frequency, high-resolution frames of depth information. With the help of the new Intel RealSense L515 LiDAR camera, data is collected from a point of view that corresponds to the human or operator, and a SLAM (Simultaneous Localization and Mapping) algorithm is implemented to generate 3D virtual maps of the environment. From experiments, it has been verified that the generated virtual maps are reliable and fairly accurate that they can be utilized as a Digital Twin in Virtual Commissioning for an indoor environment. For future work, a further investigation of the algorithm in terms of robustness toward various environments is suggested.