The current AI-based algorithms on detection and classification algorithms based on radar signatures (i.e. signals and images) shown a non-reliable solution for the detection and classification of small UAVs. A combined system with the EO/IR detection and classification based on AI-techniques could improve the required performances.

Ai4CUAV intends to improve the Threat Evaluation Subsystem of a counter UAV (C-UAV) through AI-based algorithms. Supposing the anti-drones composed by multiple heterogenous sensors, such as radars and EO/IR sensors, these algorithms “work” on radar signals and EO/IR images to enable the detection and classification of the killer-drones, as well as on drone trajectories to help to recognize a drone from another object.

This projected will build a shared database of RF and EO/IR signatures of different drones which can be used as training data and test set, which allows to compare different detection and classification techniques.

Ai4CUAV will investigate all the key SOTA of AI techniques from multiple sensor sources, including but not limited to, machine learning and deep learning. These techniques will be evaluated against the different use cases and scenarios, in order to assess the most adapted/promising ones. For the most promising techniques, algorithm prototyping and adaptation will be performed to assess preliminary performances through simulations.

This project is a follow on of the NATO SPS project n. G5633 “ANTI-DRONES – Innovative concept to detect, recognize and track killer-drones”, involving the core partners CNIT-RASS (NPD) and MTU (Co-Director), that will be concluded in September 2022 with good results, focused on the development of a new concept of anti-drone system, based on mini-radar technology and signal processing, able to detect, recognize and track the killer-drones – mini/micro UAS – in order to facilitate the neutralization of them minimizing the risk for people and assets. This AI-framework will be integrated into the ANTI-DRONES prototype, tested and evaluated by the end-user experts.

AI4CUAV project is a research project focused on a breakthrough innovation using AI. As such, it is expected to advance the accuracy of detection algorithms, and set forward AI-based solution of high complexity problems. From a prospect vision, Ai4CUAV would bring into market a novel and innovative AI application enabling intelligent ISR in complex situations.