Estonian manufacturer of ground robotic systems, Milrem Robotics, is developing an artificial intelligence (AI) algorithm for the automatic recognition of tactical UAVs and FPV drones. This prototype was recently tested at a proving ground during a hackathon near Tapa in northern Estonia.
Milrem Robotics hopes to integrate this module into the optoelectronic cameras of its THeMIS ground drones, which are being used in Ukraine. In the near future, they aim to evaluate the effectiveness of this new technology on the front lines, should peace talks between Moscow and Kyiv gain momentum.
At the end of January, Milrem Robotics began implementing a project to develop innovative counter-drone technology that challenges current practices on the Ukrainian front. The engineering team argues that the use of traditional electronic warfare systems is currently risky and often ineffective in Ukraine.
The project is driven by the need to counter wired drones controlled via fiber-optic cables, which the Russian army is using en masse. These devices have become a nightmare for Ukrainian drone defense units and specialized firms. According to experts interviewed, such drones are extremely difficult to detect and intercept, as they are controlled remotely through a spool of fiber-optic cable. Currently, the only options being explored by global research groups specializing in counter-drone technology are acoustic detection, which is still in its infancy, and AI-based cameras.
Thus, Milrem Robotics' engineering teams have developed technology that automatically detects visual anomalies in conventional optical and thermal video streams. This system, also known as an event camera, is already used in the public safety sector as stationary surveillance cameras. With AI algorithms, the event camera can be used to track suspicious individuals or vehicles. Milrem Robotics teams are now applying this technology to counter drones. The optoelectronic sensor mounted on the THeMIS drones will soon be equipped with a computer vision algorithm for automatic anomaly detection.
The developed model underwent rapid training to recognize the approach of reconnaissance drones like DJI and FPV drones from several hundred meters away. The system can assess the distance of the threat, its speed of approach, and send alerts to operators. The algorithm was trained over several weeks, and initial tests have been promising.
However, the developed prototype has faced some challenges when operating in forested and snowy terrain.
This extensive research effort involved annotating tens of thousands of videos showing drones flying at various altitudes and speeds. Most of the footage was provided by Ukrainian military partner units and during various tests conducted by the Estonian Defense League. This paramilitary organization is actively training in the use of FPV drones to prepare for potential combat against Russia.
Advanced technologies and artificial intelligence are undoubtedly impressive, but it is quite interesting to see what AI will offer the Estonians in countering Russian glide-guided two-ton air bombs, naval, artillery, and ballistic missiles...
It's a pity that the experience of Ukraine has taught the Baltics so little.
I wish the long-awaited peace to come as soon as possible!
Click 🔛
 | |
|---|---|
| Ukrainian War Crime [NSFW 18+ Only!] | The Simple Example of Modern Mass Deception |
I can if I want, but sometimes I'm lazy)