DroneShield will focus on upgrading its artificial intelligence (AI)-based detection and classification software in the coming years to develop an increasingly flexible approach that reduces reliance on data libraries, company officials told Janes in late March.
The Australian company– which specialises in countering potential threats like unmanned aircraft systems (UAS) and even in areas such as signals intelligence (SIGINT) and electronic warfare (EW) – rolled out its first fully machine learning/artificial intelligence (ML/AI)-based software to all existing customers in February as part of its quarterly software update programme.
DroneShield has launched several new products and updates to its C-UAS portfolio, such the DroneCannon MKII, and is looking to introduce regular updates to its AI-based detection and classification software. (DroneShield)
The new software – called DroneOptID and RfAI – enables near real-time detection and identification of signals of Interest, including unmanned robotic systems, the company said, along with other potential threats in the EW space. The software reduces false positives and dramatically increases the speed at which new threats are detected. It has been designed to work on all of the company’s counter-UAS products, such as the fixed site system DroneSentry, the vehicle-focused DroneSentry-X, and the body-worn RfPatrol.
DroneShield CEO Oleg Vornik told Janes the software improves on its predecessors using an algorithm that learns from all of the previous asymmetric threat signatures it has encountered. Rather than waiting for a new threat to be added to a database or library of systems, the AI-based solution is “able to make a determination on whether something is a threat or not”.
Looking to read the full article?
Gain unlimited access to Janes news and more...
