File image.
Russia’s Akvylon Research and Production Enterprise has begun pilot-series production of an artificial intelligence-based onboard computer designed for drones, the Center for Unmanned Systems and Technologies (CUST) told TASS on November 13.
“The product is designed for installation on autonomous unmanned aerial vehicles (UAVs), as well as ground-based and offshore robotic platforms. The system provides control, communications, navigation, and video stream processing. Skif’s neural network algorithms are responsible for detection, tracking, and segmentation. A single unit weighing no more than 200 grams combines the functions of a single-board computer, a flight controller, an AI board, and an interface board,” the CUST said.
The Skif integrates four components into a single unit, combining flight management, communications, navigation and onboard video processing.
The system’s neural-network algorithms are designed to detect, track and segment objects in video feeds, allowing the system to perform image-processing tasks directly onboard the platform.
The CUST said the developer identifies resistance to electronic warfare, autonomous operation without communications, redundant navigation and energy efficiency among the system’s main tactical advantages. These features are intended to allow unmanned platforms to maintain their functions in environments where communications are disrupted or navigation systems are degraded.
In addition, the development is a part of wider Russian efforts to replace foreign-made electronics used in unmanned systems.
The CUST noted that Akvylon is now playing an important role in producing domestically manufactured electronic components.
“The localization level of products manufactured by the CUST resident reaches 90%,” the company said.
The announcement comes amid growing efforts to introduce AI technology into Russian drone systems. In September, Russian military correspondent Andrey Filatov reported that Geran drones had begun using the Susanin autonomous optical navigation system, which allows them to orient themselves by the underlying terrain and continue their flight missions without GLONASS signals or communication with an operator.
ZALA, meanwhile, has been introducing its IRRA artificial intelligence-based target recognition and guidance system on Lancet loitering munitions. The onboard system automatically recognizes, tracks and locks onto targets using machine vision.
In an October report, the company said IRRA-equipped Lancets could maintain their lock on moving vehicles even when they changed direction, allowing the munitions to engage targets moving either away from or toward them.
_______________________________________________________________________________________________________________________
SouthFront: Analysis and Intelligence
NOW hosted at southfront.press
Previously, SouthFront: Analysis and Intelligence was at southfront.org.
The .org domain name had been blocked by the US (NATO) (https://southfront.press/southfront-org-blocked-by-u-s-controlled-global-internet-supervisor/) globally, outlawed and without any explanation
Back before that, from 2013 to 2015, SouthFront: Analysis and Intelligence was at southfront.com
