Dual-RTK and IMU-Based System for Accurate Drone Payload Pose Estimation With PPS Time Synchronization

In this work, the authors propose a novel electronic architecture named VESTA (Versatile External Synchronization and Telemetry Architecture), capable of independently estimating position and attitude with centimeter-level accuracy, specifically designed to be mounted as a payload for multicopter drones. VESTA integrates GPS-RTK and inertial-sensing technologies to generate a high-precision telemetry data stream, which is transmitted via a dedicated UART (Universal Asynchronous Receiver-Transmitter) interface. The key innovation of VESTA lies in its robust temporal synchronization mechanism: by exploiting GPS time signals, including UTC timestamps and PPS (Pulse Per Second) signals, each telemetry sample is uniquely timestamped with millisecond resolution. To the best of our knowledge, current commercial and custom drone navigation systems lack a reliable method for achieving this level of absolute time synchronization between inertial and GNSS data. This limitation hinders precise data fusion and complicates integration with other systems requiring temporally aligned information. VESTA addresses this gap, enabling accurate and unambiguous time correlation across subsystems. The positioning accuracy and synchronization performance of VESTA were validated through comparative testing with the DJI Matrice 300 RTK, one of the most advanced commercial UAS platforms, which notably does not support GPS-based timestamping. A Leica total station, providing millimeter-level ground truth, was employed to track the drone’s trajectory and quantitatively assess the estimation errors of both systems.