.jpg)
AIRaD – AI Radar for Drone Detection
I designed and built a complete portable X-band radar system, from the RF front-end and antenna array to the power supply and acquisition chain and integrated it with advanced signal processing to generate real-time radar images. On top of this, I developed dedicated deep learning models for the autonomous classification of drones/UAVs and their payload. The prototype was validated in real-world field trials with different types of drones, demonstrating reliable radar-based drone detection using AI.
The prototype can be deployed as a portable drone/UAV surveillance node to protect critical areas such as airports, prisons, industrial sites, borders, private property or temporary event locations. Because it is battery-powered and lightweight, it can be installed quickly and used in almost any scenario, including remote or temporary sites without fixed infrastructure and on the go. Being radar-based, it works day and night and in adverse weather, while the integrated AI automatically detects and classifies drones and their payload in real time, reducing the need for continuous human monitoring. It can also serve as a flexible research and test platform for developing and validating new radar and AI algorithms for counter-UAS applications.
Tiny AI: Recognition and Classification of sounds with embedded low-cost devices
See the unseen / SKY100