Error bound analysis of marker-based visual navigation in urban air mobility: a parameter study and experimental verification for vertical operations

Abstract

This paper presents an experimental validation and system analysis of a two-step error bounding method for marker-based visual navigation in vertical flight operations. Building on prior simulation studies, we verify the proposed Lie group–based parameterization approach using real flight data collected at Magdeburg-Cochstedt Airport. The method effectively bounds positioning errors within 4-sigma confidence intervals (99.9%) for vertical takeoff and landing below 43m altitude under real-world conditions. We further analyze challenges at higher altitudes, where error bounds degrade due to poor measurement geometry. To address this, we incorporate additional visual features available at regulated Vertiport sites, such as V-pattern and inner square markers, that improve error bounds by up to 70%. Our results demonstrate the potential of leveraging existing visual markers to enhance navigation reliability and efficiency in urban air mobility and U-space operations.