Threat Analysis – Russian AI-enabled explosive-laden FPV-drones
Disclaimer: With the courtesy of the Counter-IED Centre of Excellence, we are providing you with this threat analysis document for your further use. Please do not share this information openly on the internet, as the original document is classified as OFFICIAL SENSITIVE /// DISTRIBUTION LIMITED TO MILITARY & LAW ENFORCEMENT.
REFERENCES:
A) C-IED CoE S325/2024 – Threat Analysis – 2024 Russian handbook on explosive-laden first-person view drones, 16 October 2024
Bottom Line Up Front: Russia has implemented artificial intelligence (AI) technology into explosive-laden first-person view (FPV) drones: the associated capability of autonomous guidance, target acquisition and engagement is allowing attack drones to overpass adversary electromagnetic warfare countermeasures.
- AIM
The purpose of this report is to provide a better understanding of the Russian drone tactics, techniques and procedures using the Овод-C (OVOD-S “GADFLY-S”) and similar explosive-laden drones fitted with AI, along with the potential consequences from that.
- BACKGROUND
The OVOD-S/ Овод-C drone is a first-person view quadcopter drone. It was developed by Andrey IVANOV and produced by the company «РЛВТ» (РУССКАЯ ЛАБОРАТОРИЯ ВОЗДУШНОГО ТРАНСПОРТА, “Russian laboratory for air transport”) established in TULA, Russian Federation, 180 km south of Moscow.
The cost of one Овод-C drone is not higher than 69,000 Rubles (about 710 US $). The drone has a payload of maximum 3.5 kg (1.5 kg less than the basic OVOD version without AI capability, due to the additional payload posed by the AI-related computer).
Those drones can be equipped with several kinds of payloads like specific drone munitions or improvised explosive devices (IEDs) including modified conventional munitions. The flight duration is estimated between 8 and 10 minutes, and the maximum range is up to 8 km.
In the final operating version of Овод-C drone, an AI-enabled homing system utilizes a combination of algorithmic tracking through technical vision and a neural network-based system. It analyses video feeds from onboard cameras, which enables the drone to interpret its surroundings, pinpoint potential targets, and autonomously investigate suspicious entities – such as an isolated bush in an open field that might conceal a camouflaged enemy vehicle or outpost. Upon closer inspection, utilizing both standard and thermal imaging, the drone can confirm the nature of the target with a high degree of accuracy.
The homing system, powered by advanced technical vision and neural network technologies, facilitates object recognition and precise targeting. Operators can designate targets with ease, simply by pressing a button, and the drone will pursue them even under the interfering effects of electronic warfare. Furthermore, the OVOD-S features a remote detonation capability, a self-destruction mechanism for sensitive mission scenarios, and a multi-layered protection system for its warhead.
Mr. Martin ROSENKRANZ wrote an article on 6 July 2024 (militaeraktuell.at) describing how similar the performance of the Russian площадь (PLOSHCHAD, “square”) onboard AI-powered homing system (allegedly built in the OVOD-S) is in comparison with the Skynode S from the US-Swiss company Auterion, which is under operational use by Ukrainian Forces.
Figure 1 – Video capture showing a Овод-C drone with the homing system emplaced on top during tests (Image Source: Youtube)
3. OTHER TYPES OF RUSSIAN AI-SUPPORTED EXPLOSIVE-LADEN FPV DRONES
During 2024 and early 2025, Russian Armed Forces have been equipped with several other models of attack FPV drones taking benefit of AI, examples as follows:
- Скворец-Z (“Starling-Z”) by Технодрон (“Technodrone”) company.
- Скворец PRO (“Starling PRO”) by Технодрон (“Technodrone”) company.
- Гранат-3 (“Granat-3”) by ОКБ Астрон (“Astrohn”) company.
- Mикроб (“Microbe”) by КБ Микроб (“Microbe”) company (it also uses Ploshchad system).
- Русак-С (“Rusak-S”) by АО Стратим (“Stratin”) company.
- Ветер-8, Ветер-10 (“Veter-8”, “Veter-10”) by AO Котлин-Новатор (“Cotlin Novator”) company
On the other hand, it has been also reported the use of AI module for target recognition and autonomous engagement inside last Russian versions of Iranian origin “Shahed” loitering munitions.
Figure 2 – Video capture showing AI-based recognition of targets by a Rusak-S drone (Image Source: www.focus.ua)
4. ANALYSIS OF TACTICS, TECHNIQUES AND PROCEDURES USING OVOD-S DRONES
Analysing several videos from open sources, the Russian Armed Forces are mainly engaging main battle tanks and armoured vehicles with the OVOD-S drone, obviously in a one-way mode. The troops are guiding the drones to the area of engagement via radio frequency, then, and with machine vision capability, the drones can detect targets on their own. In the moment an OVOD-S tracks a target, the AI-enabled guidance system likely takes over control. This makes the drones highly resilient against electronic countermeasures in the end phase of an attack.
The fact that these types of drones have only a short playtime indicates that they cannot be used for reconnaissance or surveillance. Therefore, someone or something must conduct reconnaissance or surveillance, and submit the information about opponent forces to alert the drone operators to engage.
Ukrainian sources are reporting OVOD-S drones to decrease speed once the operator marks the target and the drone has to act autonomously: it poses a relative advantage for the target forces to try to engage the drone with their weapon systems.
In fact, the Russian publication “RIA Novosti” claims that the OVOD-S drone can destroy targets “… without human intervention…”.
Nonetheless, the analysis of the referenced YouTube video (in which the designer is showing the performance tests of OVOD-S drone and describing its functioning and capabilities) indicates that the potential target is captured by the operator: “… the effectiveness of this solution is due to the fact that when we go beyond the limits of radio visibility, that is, behind a forest, behind a hill, behind a building, our connection ends there. Having captured an object, its target, in advance – during a loss of communication, the device continues to track the target and destroys it regardless of whether it is a static or dynamic target…”, says the drone developer.
5. ASSESSMENT
Although it is still in the way of improvement in effectiveness, the addition of AI capabilities for the autonomous or semi-autonomous use of explosive-laden FPV drones poses a realistic challenge to adversary forces, which would see how their electromagnetic warfare counter measures against a small-unmanned aircraft system (s-UAS) are not functional against explosive-laden FPV drones.
Considering the quick development and progressively decreasing costs of commercial-off-the-shelf (COTS) AI-powered electronics, it is estimated the accessibility of COTS AI devices able to recognize and select targets in a short term: it would allow violent extremist organizations (VEOs) to expand their scope in the potential use of explosive-laden drones.
