Evaluating the performance of GNSS RTK and UAV (drone) technologies under vrs mode in challenging environments

Abstract

Reliable geospatial data acquisition in obstructed environments is important for engineering surveying, but buildings, vegetation, multipath, and operational restrictions can reduce positioning reliability. This field study compared Global Navigation Satellite System Real-Time Kinematic (GNSS RTK), an RTK-enabled unmanned aerial vehicle (UAV) using Virtual Reference Station (VRS) corrections, and handheld Simultaneous Localisation and Mapping (SLAM) across urban, forested, mixed-use, and semi-urban sites in Sri Lanka. Static GNSS and closed Total Station traverses established the reference coordinates. GNSS RTK was tested at all four sites, UAV RTK only in the urban corridor, and SLAM in the forested and mixeduse sites. Horizontal performance was assessed using root mean square error (RMSE), median, 95th percentile, maximum error, and availability. GNSS RTK RMSE values ranged from 0.064 to 0.121 m, UAV RTK recorded 0.106 m in the urban corridor, and SLAM recorded 0.160 m in the forest and 0.299 m in the mixed-use site. GNSS RTK gave the lowest errors where a stable FIX was available; UAV RTK provided rapid coverage, while SLAM maintained continuous mapping with higher drift-related error. Technology selection should therefore reflect required accuracy, coverage, access, and operating constraints.