he success of drone missions often relies on an accurate estimate of the drone’s position, attitude, and velocity. The classical navigation approach of fusing inertial sensor data with GNSS quickly fails if the GNSS signal is lost or unreliable. An alternative approach mitigat-ing this problem consists in fusing inertial and GNSS measurements with a model of the aerodynamic forces and moments exerted on the vehicle. We will first present an extension of this vehicle dynamic model (VDM) based method to use multiple redundant inertial sensors, including results with real flight data on a fixed-wing drone under simulated GNSS outage. We will also discuss the limits in reducing the noise from the multiple IMUs and improving the overall precision. We will then present the process of adapting the VDM navigation framework to a delta-wing drone in terms of model structure and the path to the working prototype with perspective applications.