Assessment of a robust MEMS-based RTK/INS system for UAV applications

Recent years has seen a significant increase in UAVs being used in large-scale surveying projects. One reason for this success is that MEMS-based GNSS/INS systems have become available, fitting the strict weight and power budgets of UAV missions as well as driving down the costs. These GNSS/INS systems need to be robust and need to provide reliable and accurate results, in real-time or post-processing, to limit manual corrections and avoid having to re-do surveys. In this paper we present Septentrio's new RTK/INS solution, specifically developed for UAV survey and inspection applications. In collaboration with a survey UAV manufacturer the performance of the RTK/INS system is assessed onboard a UAV in an environment typical for surveying applications. The performance of the RTK/INS system is compared with other commercially available MEMS-based GNSS/INS systems. The systems are mounted together on a large X8 octocopter UAV (28 kg). All systems use dual-antenna GNSS receivers and share the same two antennas, separated by a two-meter baseline, for GNSS attitude. The GNSS attitude adds information to the MEMS-based INS filter, which for a multi-copter can improve attitude accuracy during low dynamics (e.g. hovering, straight and level flight) and improves reliability at startup. A high-grade LiDAR mounted under the UAV provides an application-oriented indirect reference. The LiDAR data collected onboard the UAV is geo-referenced with the position and attitude solutions of each of the GNSS/INS systems under test to generate LiDAR point clouds. In addition, data is collected with the same LiDAR mounted on the ground. This provides a millimeter-accurate geo-referenced model of known structures. The performance of each GNSS/INS solution can be assessed by comparing the accuracy of the UAV-based point clouds with respect to the ground-based point cloud on those features. Since the same LiDAR data is used for each of the GNSS/INS systems, this serves as an independent reference to determine the final geo-referencing accuracy of each of the systems.