Russian Media: Russia Develops Autonomous Navigation System for Drones

Reference News Network reported on August 21 that the Russian newspaper "Izvestia" published a report on August 17 titled "The 'Eye of Mars' for Drones: Russia Develops Navigation System for Drones to Fly in Electronic Warfare Interference and Satellite Signal-Less Environments." The report states that Russian engineers have taken a significant step towards achieving full autonomy for drones—they have completed research work and begun equipping drones with a video stream-based autonomous navigation system. This technology is based on the visual odometry method, allowing drones to autonomously locate themselves without radio signals, satellite modules, or operators.

The onboard cameras of the drones continuously scan the ground, while a neural network module processes the relevant images in real-time and compares the visual data with pre-loaded digital terrain maps—satellite images or aerial photography. This is precisely how drones equipped with the system developed by engineers from Russia's "Kvant" company locate themselves in specific environments.

Meanwhile, Russian developers have successfully introduced an innovation that addresses the long-standing issue of cumulative error in inertial navigation systems. The new method compares the actual terrain below the aircraft with a reference (digital terrain map) to instantaneously correct cumulative deviations, ensuring precise targeting.

In professional circles, visual odometry is also known as "Mars technology." This term originates from NASA's Mars missions, during which the "Ingenuity" helicopter was used on Mars. Obviously, there is no GPS on Mars, and waiting for signals from Earth takes too long. Therefore, it locates itself based on the principle of an optical mouse, that is, "looking down the road," capturing micro-terrain with its camera and calculating displacement relative to the takeoff point. In 2024, after the mission concluded, the blueprints and code were open-sourced, and this method began to be applied in the field of drones.

Dmitry Kuzyakin, a Russian drone expert and chief designer at the Integrated Solutions Center for Unmanned Systems, told "Izvestia" that this navigation principle fundamentally changes the tactical application of drones.

"Just as the first-person view (FPV) drones in 2022 had a disruptive impact on the battlefield contact line, machine vision navigation is rewriting the rules of aerial combat," he said. "The equipment can operate in all weather conditions, move without antennas, and execute missions without direct control, GPS, or any external references."

Previously, among companies developing visual odometry-based technologies, the American company Skydio achieved the most significant results, including in combat drones. Russia has proven that it has also joined the ranks of countries possessing this technology. Kuzyakin stated, "The key is for engineers to initiate mass production and deliver the results to the combat units in the special military operation zone."

Military expert Yuri Lyamin noted that the reason for using the visual odometry solution as a foundation is primarily its resistance to interference. This characteristic is crucial in situations where satellite signals near military contact lines are systematically suppressed or deceived by electronic warfare means.

At the same time, the financial cost for drones to adopt the new technology is not high—it utilizes standard onboard cameras. Of course, like any technology, visual navigation has natural physical limitations. Thick clouds, fog, precipitation, and nighttime conditions can reduce operational quality.

Nevertheless, the introduction of visual odometry by Russian designers is a very important step, determining the survivability and combat effectiveness of the new generation of drone systems. (Translated by Tong Shiqun)