Denial of global navigation satellite systems is a relatively inexpensive, but very effective weapon in the adversary’s arsenal, removing the capability to accurately localise and therefore navigate. Current alternatives to the global positioning system, such as visual simultaneous localisation and mapping (vSLAM), inertial navigation systems (INS) or dead reckoning all have their advantages and inherent limitations. Truly effective INS systems are large and heavy, and therefore unsuitable for smaller drones. Flare Bright have developed an approach to significantly boost INS performance in drones, and have demonstrated a means of achieving stealthy, GPS-free, unjammable flight through sophisticated machine learning techniques. We have taken a traditional micro-electromechanical system (MEMS) INS and integrated a platform digital twin flight model, capping the small error rate at a linear rate, which makes it both measurable and controllable.