‘We are reaching a new level of automated warfare technologies,’ Russian military experts concur as they comment on the early use of AI-enabled technology in battlefield scenarios. What exactly are these technologies, what missions are they deployed in and how are the Russian AI-assisted vehicles faring against the similar technology employed by Ukraine?
Back in February, Russia’s defence ministry (MoD) announced the upcoming rollout of AI-enabled UAVs. In October 2023, Denis Manturov, the-then minister of industry and trade noted, ‘The changing battlefield realities are pushing for the introduction of AI technology. The indications are, the decision time is often limited to mere seconds. Such speed is simply unfeasibly without a heavier reliance on the AI.’
President Putin has also repeatedly stressed the importance of using AI for military purposes. ‘Over the past few years, the defence contractors have multiplied their output and supplies of the most in-demand armaments <…> The next big step is the development and mass production of AI-based technology with an eye to the extensive military use thereof,’ the Russian leader was quoted as saying.
According to the latest MoD’s announcement, the use of AI is already gaining traction in the ongoing special operation. Defence minister Andrei Belousov has been briefed by drone operators who are tasked with receiving the UAV-provided data and identifying the opponent’s targets in real time, while processing the video feeds is aided by the AI technology.
Experts seem to concur that AI is fast becoming an integral part of the current warfare. ‘The use of AI is going to speed up and improve the accuracy of the missions while boosting the operators’ productivity,’ says Alexander Kaniovsky, the deputy CEO and co-founder of Fly Drone.
‘The AI technology will significantly reduce the time it takes to hit enemy targets,’ military expert Boris Rozhin agrees. ‘Granted, so far the AI cannot fully replace human inputs, but the operators capable of using the technology will see an uptick in productivity and speed,’ adds Alexei Rogozin who heads the Centre for the Development Transportation Technology.
A video feed is an image taken by a drone-mounted camera and relayed to the control centre. It can be an enemy location captured by a reconnaissance UAV from a high altitude. But even experienced recon officers often have a hard time identifying the objects shown in a picture. Worse, those can be camouflaged. Identifying such objects – a tank, an artillery piece, a trench, a pillbox etc. – amid the surrounding terrain is exactly what the AI technology is supposed to help with.
‘The AI involves a suite of processing and computing algorithms that hone their efficiency as they learn from a mounting body of data available. The larger the amount of the data input, the more sophisticated the algorithm,’ Rogozin explains.
The algorithm is trained to identify the attributes of specific objects, including in settings that render those inaccessible to the human eye.
According to Alexei Rogozin, today’s battlefield is overflowing with data, but ‘only a tiny portion of it gets processed to a point where it is usable in the AI algorithms’. ‘It packs a humongous potential for military use. That is what we are beginning to tap into now,’ the expert adds.
Amid the shrinking decision times, the AI can both facilitate the operators’ job and bump up the hitting accuracy.
‘The use of AI in processing the video feed is often referred to as machine vision,’ Alexei Rogozin continues.
‘I reckon there are three avenues where the use of this technology will be most productive. One is detecting hidden objects and unobtrusive movement. Two is enhancing the quality of the images that are often compressed or blurred due to the atmospheric phenomena. Three is drone navigation, which means the GPS- and GLONASS-type satellite systems can be replaced with visible portions of the terrain being the primary navigator.’
The expert believes all of the three major usage scenarios are being explored and tested, including in real combat.
The AI and neural networks can help further automate drone operations. ‘Some of the drones in current battlefield use feature neural net processors underlying the AI guidance systems. These UAVs can distinguish infantry personnel from their equipment. Although the resulting quality may be spotty, this technology should be used more extensively. The future is already here,’ says Boris Rozhin.
The AI has the capacity to increase the drones’ survival rate.
According to Alexander Kaniovsky, ‘machine vision and other AI features allow modern drones to travel a certain portion of target distance disconnected from their operations as well as to avoid electronic warfare (EW) jamming’. It helps UAVs to be protected from enemy detection.
‘The use of AI will further reduce the EW efficiency,’ Rozhin predicts.
Finally, Alexander Kaniovsky argues that the AI can help plan large-scale drone operations and analyse a bulk of reconnaissance data obtained by the drones.
However, the introduction of the AI technology is not something specific to Russia. According to various media outlets, Ukrainians have been using similar technology in active combat scenarios. In early April, CNN reported on the attacks of the Russian oil refineries perpetrated by the Ukrainian drones that ‘have even begun to integrate a basic form of artificial intelligence’. Ukraine’s deputy defence minister Kateryna Chernohorenko indicated that they were developing their own AI-enabled vehicles.
But according to Dmitry Kuzyakin, the director of the Centre for Integrated Unmanned Solutions, Russia is outstripping Ukraine in the extent of the AI-assisted drone usage. ‘Importantly, Ukraine does not produce these drones. Those are being supplied by the West. It leads us to believe that the Russian army will give its Ukrainian counterpart the run for its money,’ Boris Rozhin explains.
We are currently at an early stage of the extensive AI use in battlefield scenarios. If Rozhin is to be believed, we are about to see the upgrade and mass production of a range of AI-enabled vehicles. ‘Further testbed experiments will certainly uncover quite a few inevitable flaws. But we are already reaching a new level of automated warfare technologies,’ he says.
‘The AI’s role will be expanding. I think the widespread integration of the AI into multiple weapons systems will almost happen in the blink of an eye and begin an integral part of the decision cycle,’ Alexei Rogozin concludes.