Sentient signs contract extension with AeroVironment

Melbourne-based Sentient Vision Systems has signed a multi-year contract extension with US firm AeroVironment, a global leader in intelligent, multi-domain robotic systems, to provide enhanced airborne surveillance capabilities.

Sentient Vision Systems’ Kestrel Moving Target Indication (MTI) system has been part of every Group 1 Unmanned Aerial System (UAS) manufactured by AeroVironment since 2011. There are now thousands of Kestrel systems in service with AeroVironment’s customers worldwide.

The contract extension will see Sentient Vision Systems provide updates to its Kestrel MTI system. These will enhance AeroVironment’s solutions with operationally proven advanced computational and deep learning algorithms, arming the modern war-fighter with trusted, advanced Artificial Intelligence (AI) to bring enhanced accuracy and efficiency to complex and dangerous tasks.

“This contract extension re-affirms the commitment of both companies to providing world-leading, tactical airborne surveillance solutions,” said Dr Paul Boxer, Managing Director of Sentient Vision Systems. “We invest heavily in the R&D necessary to keep our product range ahead of the threat and it’s gratifying to get this endorsement of our ongoing work from such a respected global customer.”

AeroVironment’s Group 1 UAS product range includes the Wasp AE, Raven and Puma AE, all of which weigh less than 25lb (12kg). AeroVironment’s family of Group 1 UAS customers includes the U.S. Army, U.S. Marine Corps, Australian Army and 50 others worldwide.

Kestrel is a full-motion video processing software that automatically scans sensor imagery feed to detect and track very small, slow-moving targets in real time. Based on Sentient Vision Systems’ own AI system dubbed Sentient AI (a powerful combination of deep learning and advanced computer vision), Kestrel is a significant wide area search capability that provides up to 20 times increase in surveillance area by analysing EO/IR sensor imagery frame by frame and automatically detecting targets in real time. It indicates potential targets of interest that would otherwise be too small to be seen by an operator, or that are ambiguous and that are in a very ‘cluttered’ environment with lots of moving people and vehicles.