The ATTENTION project (Advanced Target Detection, Recognition and Identification based on Multispectral Inertial Odometry for Navigation) is a DEFRA research project aimed at creating the next generation of intelligent drone perception and navigation technologies. Modern defence operations increasingly rely on unmanned systems, yet drones must often function in environments where GPS signals are unreliable and where traditional cameras struggle with smoke, fog, darkness or camouflage. ATTENTION addresses these challenges by empowering drones with enhanced sensing, autonomous navigation, and robust target understanding—without relying on active sensors or constant communication links.
At the heart of the project is the development of multispectral vision, combining visible, short‑wave infrared (SWIR) and long‑wave infrared (LWIR) imaging. By fusing information across these complementary spectral bands using deep‑learning techniques, drones gain a far clearer and more resilient understanding of their surroundings. This allows them to detect and classify objects more accurately, even in conditions where normal cameras fail. The system also reduces false alarms and improves the ability to recognize camouflaged or concealed targets, contributing to safer and faster decision‑making for defence operators.
A second major innovation in ATTENTION is the ability for drones to navigate without GPS. Using advanced visual‑inertial odometry and AI‑driven image‑matching techniques, the drone can compare its onboard camera feed to stored satellite or aerial imagery, allowing it to estimate its position autonomously. This enables precise flight and localisation in GPS‑denied or jammed environments—crucial for operations in urban areas, dense forests, or remote locations. Because the drone can operate without active sensors or radio transmissions, it can also perform missions more discreetly when needed.
The project will integrate these sensing and navigation capabilities into a next‑generation UAS platform, jointly developed by the Vrije Universiteit Brussel (VUB), the Royal Military Academy (RMA), and industrial partner SABCA. The result will be a lightweight, highly capable drone demonstrator equipped with multispectral cameras, inertial sensors, and AI‑driven perception software, capable of advanced reconnaissance, persistent surveillance, and improved environmental understanding under challenging conditions.
Ultimately, ATTENTION strengthens Belgium’s defence capabilities by delivering technologies that improve situational awareness, enhance resilience in hostile environments, and reduce cognitive load on operators. Beyond its military relevance, the project also creates opportunities for dual‑use applications such as infrastructure inspection, emergency response, environmental monitoring, and operations in GPS‑challenged areas. By combining academic excellence with industrial expertise, ATTENTION lays the foundation for safer, smarter and more autonomous unmanned systems for the future.