Several start-ups in Ukraine are developing AI drones. The goal is to help manage a large fleet of drones, which will take the war into uncharted territory, reported "Reuters", quoted by BTA.
Ukraine hopes that deploying AI drones on the front line will help it overcome increasing signal jamming by the Russians, as well as allow unmanned aerial vehicles (UAVs) to operate in larger groups.
p>Development of UAVs with artificial intelligence in Ukraine is divided between visual systems that help identify targets and control drones, terrain mapping for navigation and more complex programs that allow UAVs to operate in interconnected " swarms".
Swarmer, one of the companies working on this, is developing software that connects drones into a network. Decisions can be executed instantly across the group, with a human stepping in only to give the go-ahead to automated strikes.
Swarmer CEO Sergey Kuprienko commented that when trying to scale up with human pilots, it just doesn't work, explaining that "for a swarm of 10 or 20 drones or a robot, it's virtually impossible for humans to they rule them."
Kuprienko said that while human pilots struggle to manage operations involving more than five drones, AI will be able to handle hundreds.
The system, called Styx, directs a network of reconnaissance and attack drones, large and small, in the air and on the ground. Each drone will be able to plan its own moves and predict the behavior of others in the swarm, he stressed.
In addition to expanding operations, Kuprienko explained that the automation will help protect drone pilots who operate near the front lines and are a priority target for enemy fire. Swarmer technology is still under development and has only been experimentally tested on the battlefield, he added.
Swarmer is one of more than 200 tech firms that have sprung up since Russia launched its full-scale invasion in 2022, with IT people developing drones and other devices to help Ukraine counter Russia.
Samuel Bendet, assistant senior fellow at the Center for a New American Security, said artificial intelligence drone control systems will likely need a human to prevent system errors in target selection.
There are widespread concerns about the ethics of weapons that preclude human judgment. A 2020 European Parliament research paper warned that such systems could commit violations of international humanitarian law and lower the threshold for starting war.
Artificial intelligence is already being used in some of the long-range drone strikes from Ukraine that target military installations and oil refineries deep inside Russian territory.
A Ukrainian official, speaking on condition of anonymity, said the attacks sometimes involve a swarm of about 20 drones and explained that the main drones fly to the target, but the job of the others is to eliminate or distract air defenses along the way. To do this, they use a form of AI with human supervision to help detect targets or threats and plan possible routes.
The need for AI drones is growing as both countries deploy electronic warfare (EW) systems that disrupt signals between pilots and drones, and in particular small, low-cost FPV (first person view) drones , which have become the primary way for both sides to hit enemy vehicles in 2023, see their hit rates drop as jamming increases.
According to Max Makarchuk, Head of AI for Brave1, the percentage of FPVs hitting their target is constantly dropping. Most FPV units are already seeing a hit rate of 30-50%, for new pilots it can be as little as 10% and predicts that AI-controlled FPV drones can score a hit rate of around 80%.
To counter the threat of electronic warfare, creators including Swarmer have begun developing features that allow a drone to lock onto a target via its camera.
Systems used in electronic warfare form an invisible jamming dome over the equipment and soldiers they protect. If a pilot loses contact with the drone, he can no longer control it and the aircraft either falls to the ground or continues to fly straight. But automating the last part of the drone's flight to its target means it no longer needs the pilot - thus nullifying the effect of EW jamming.
AI drones have been in development for years, but until now they were considered expensive and experimental.
Bendet pointed out that Russia had been developing aerial and ground-based artificial intelligence drones before the 2022 invasion and had claimed some successes.
In Ukraine, the main task of manufacturers is to develop an artificial intelligence guidance system for drones that is cheap. This would allow it to be deployed en masse across the entire 1,000 km frontline where thousands of FPV drones are in use every week.
Costs can be reduced by running AI programs on the Raspberry Pi, a small, inexpensive computer that has found global popularity beyond the educational purposes for which it was designed.
Max Makarchuk said he calculated the cost of putting a simple guidance system that would lock into a shape visible to the drone's camera at only about $150 per drone.