Error Analysis of Extensive Drone Flight Data for Satellite Navigation in Urban Environment

This paper describes a methodology to evaluate the challenges of Advanced Air Mobility in urban environment by adopting traditional navigation systems based on integrating GNSS and inertial measurements. A flight tests campaign has been developed by using a commercial drone in urban environments that are characterized by a different geometry of infrastructure. The selected application of drone mission profile is focused on building surfaces inspections for defects and damages detection and recognition. The image processing needed to obtain photogrammetry products such as point clouds and orthophotos is used to estimate the drone position error that is compared with the estimated value of Geometric Dilution Of Precision parameter of satellite-based navigation. Telemetry data recorded by the drone on-board systems were analyzed and the collected images were processed with commercial off the shelf software for aerial photogrammetry. A GNSS simulator has been set to read the satellite ephemeris data downloaded by open-source databases in order to reproduce the satellites constellation at the time and location of the executed flight missions. Thanks to the comparison between the drone position computed by the telemetry file and the one computed by the photogrammetry software during the images processing, a reference for the position error estimation has been defined. The computed value of the experimental drone position error has been compared with the estimated value of Geometric Dilution Of Precision obtained by the GNSS simulator. Synthetic charts can be developed for different Above Ground Level height values and different urban environment geometries to be consulted to assess the urban canyon impact at the mission time, supporting risk assessment when satellite-based navigation experiences challenging applications. This work examines the geometry of satellite receiver visibility and quantifies how many line of sight paths between the satellites and the drone are lost. The objective is to ensure that Unmanned Aircraft System navigation solutions are predicted correctly and simulated digitally under realistic operational conditions where line of sight restrictions cannot be neglected.