MOSCOW, May 12. /TASS/. The Kazan-based Arkodim company (a resident of the Center for Unmanned Systems and Technologies, CUST) has developed neural network analysis and tracking algorithms for a low-level air defense anti-drone system, the CUST reported.

The algorithms will reduce the cost of engaging a target 1,000 times compared to a conventional surface-to-air missile system (SAM), it added.

"Arkodim (Kazan) has implemented a suite of engineering solutions to automate the detection and targeting of enemy unmanned aerial vehicles (UAVs). Neural network analysis and tracking algorithms have been developed to date. <…> The developer has already completed R&D and laboratory testing of a dual-axis turret that can independently detect targets, including small UAVs. Neural network algorithms can also serve as the basis for the subsequent development of an anti-drone system for Russia's short-range air defense systems. According to the CUST resident calculations, the use of anti-drone systems with AI elements will reduce the cost of engaging a target 1,000 times compared to conventional air defense systems," the CUST said.

The company specified that the neural network analysis and tracking algorithms calculate ballistics and ensure reliable target tracking even during brief loss of visual contact. "In addition, the algorithms enable preemptive action. They accurately determine the target’s location after a certain period of time, significantly facilitating destruction with small arms," the CUST added.

The Center noted that neural network algorithms can be integrated into a turret equipped with a weapon--for example, a machine gun and laser. "Another application is an observation tower, which is designed to designate targets for external systems to neutralize unmanned aerial threats, including interceptor drones. <…> Arkodim presented its project at the Russian Venture Forum held in April. The developer estimates the investment required for the full implementation of the anti-drone solutions at 100-300 million rubles ($1.35-4.062 million)," the CUST stated.