Modular Submersible Robot for Exploring Flooded Infrastructure in Post-Mining Areas – Development, Challenges and Initial Practical Application
Authors: Dr. Holger Witthaus, RAG Aktiengesellschaft, Mine Water Management Planning, Mine Water Concepts, Essen/Germany, Mareike Schubert, M.Sc., Univ.-Prof. Dr.-Ing. Oliver Langefeld, Clausthal University of Technology, Institute of Geotechnology and Mineral Resources, Department of Mining, Clausthal-Zellerfeld/Germany, Dipl.-Ing. Sebastian Matz, Fraunhofer IOSB, Advanced System Technology branch AST, Department of Underwater Robotics, Ilmenau/Germany
DOI: 10.66356/KTNO3991
The availability of novel, compact multi-sensor scanning technologies – particularly those integrating LiDAR and imaging sensors – enables time-efficient and often sufficiently precise acquisition of underground 3D geometries. Based on this data, colour-augmented point clouds and interconnected 3D models can be generated, opening up a wide range of applications: from geological and engineering-geological mapping to vivid 3D visualisations for exhibitions, such as those in show mines.
Modern methods for processing point clouds, such as so-called “Gaussian Splatting”, have led to significant improvements in rendering sharpness and model accuracy. Using selected examples from the field of historic mining, this contribution provides insights into the potential of current sensor and evaluation methods and illustrates how these technologies can be effectively employed in the documentation and presentation of underground structures.
