Continuous and Autonomous Digital Twinning of Large-Scale Dynamic Indoor Environments

Abstract

In the context of Industry 4.0, the interest in digitalizing manufacturing environments is steadily increasing. Hence, the need of a frequently updated digital twin of the facilities is also growing. In order to create a digital twin, specialized hardware is used to capture data with lidars and cameras. The data is then processed by a SLAM algorithm. However, the data acquisition process is typically done manually by multiple employees. This requires dedicated training and many working hours and hence is not feasible to do on a daily basis. In this paper, we present a solution to automate the data acquisition process, by combining an autonomous mobile robot and a scanning device. We show that the quality of the resulting point clouds matches the one of the manual scanning process and is hence ready to be deployed in real environments.