UAVOS reports that it has successfully tested NAVAI, a vision-based navigation module for uncrewed aerial systems (UAS) that is designed to improve flight resilience when GNSS signals are degraded or temporarily unavailable.
NAVAI uses neural networks to match real-time camera imagery with pre-loaded terrain maps, helping UAS estimate their position by matching real-time camera imagery with pre-loaded terrain maps when satellite navigation cannot be relied upon. The module runs on embedded computing platforms and external mission computers and integrates with the APS Ground Control Station (GCS).
AI algorithms filter visual noise to recognise ground features in the event of cloud cover, haze, low light or seasonal changes in the landscape. A dedicated interface enables pilots on the ground to see a live overlay of what the drone ‘sees’, matched directly onto the mission map.
UAVOS has developed NAVAI for autonomous operations across commercial and industrial applications, as well as approved specialised programmes where applicable.
For more information