181 / 2023-08-31 20:19:28

Quantitative analysis of different SLAM algorithms for Geomonitoring in an underground test field

Simultaneous Localization and Mapping,robot,underground geomonitoring,HDL Graph SLAM,RTAB-Map,accuracy and precision

Abstract Quantitative and reliable information provided by geomonitoring helps to identify hazards and to timely adopt appropriate measures. However, doing this job inherently exposes monitoring staff to a hazardous environment, especially in the field of underground geomonitoring. Since 2000, since robots have been widely applied in various fields, many studies have focused on the establishment of automated robotic systems as well as underground navigation and mapping. Only a few studies have conducted quantitative evaluations of the proposed or used methods, and almost none have provided systematic and comprehensive assessment of suitability of mapping robot for underground geomonitoring areas. In this study, the accuracy and precision of two selected Simultaneous Localization and Mapping methods (HDL Graph SLAM and RTAN-Map), implemented on a designed robot system, were systematically and quantitatively evaluated. In order to assess accuracy, an underground test field was developed, which includes a 20 m long tunnel for accurate analysis using control points and an 80 m nature ring area for accuracy analysis of points, line segments, and areas using placed surfaces. Experimental measurements were performed in this test field using a robot equipped with various sensors in conjunction with the selected SLAM. The resulting point cloud was compared with the reference point clouds measured by a total station and a terrestrial laser scanner. The accuracy and precision of the selected SLAM methods as well as the verifiability and reliability of the results were evaluated and discussed using metrics such as the coordinates of control points, cloud-to-cloud distance between the test and reference point cloud, normal vector, centre point coordinates and area of the placed surfaces.