Ukraine rushes to develop AI-powered war drones

In Ukraine, a handful of startups are developing artificial intelligence (AI)-based systems to help fly a massive fleet of drones, taking warfare into uncharted territory as combatants vie for a technological edge in battle.

Ukraine hopes that the deployment of AI-powered drones along the front line will help overcome increasing Russian signal jamming and enable unmanned aerial vehicles (UAVs) to operate in larger groups.

AI drone development in Ukraine is essentially divided into visual systems that help identify targets and fly drones to them, terrain mapping for navigation, and more complex programs that allow UAVs to operate in interconnected “swarms.”

One company working on this topic is Swarmer. It develops software that connects drones in a network. Decisions can be implemented immediately across the entire group, and a human only needs to intervene to authorize automated attacks.

“If you try to scale this (with human pilots), it just doesn’t work,” Swarmer CEO Serhiy Kupriienko told Reuters at the company’s Kyiv office. “Controlling a swarm of 10 or 20 drones or robots is virtually impossible for humans.”

Swarmer is one of more than 200 technology companies that have sprung up since Russia’s large-scale invasion began in 2022. In it, civilians with IT backgrounds are developing drones and other devices to help Ukraine face a much larger enemy.

Other interesting topics

Ukraine intercepts all Russian drones targeting Kyiv overnight

“The latest airstrike alert in Kyiv was a drone threat. In the last two weeks, this is at least the fifth actual attempt by the enemy to attack the capital with drones,” says a Kyiv military official.

Kupriienko said human pilots struggle with missions involving more than five drones, while AI is capable of handling hundreds.

The system, called Styx, controls a network of reconnaissance and combat drones, both large and small, in the air and on the ground. Each drone can plan its own movements and predict the behavior of the others in the swarm, he said.

Kupriienko said automation would not only increase missions but also help protect drone pilots who operate close to the front line and are a priority target for enemy fire.

Swarmer’s technology is still under development and has only been experimentally tested on the battlefield, he added.

Samuel Bendett, adjunct senior fellow at the Center for a New American Security, said AI-based drone control systems would likely require human intervention to prevent the system from making mistakes in target selection.

There are major concerns about the ethics of weapons that exclude human judgment. A 2020 European Parliament research paper warned that such systems could violate international humanitarian law and lower the threshold for entering war.

AI is already being used in some of Ukraine’s long-range drone strikes, targeting military facilities and oil refineries hundreds of kilometers inside Russia.

A Ukrainian official who wished to remain anonymous told Reuters that the attacks sometimes involve a swarm of about 20 drones.

The main drones fly to the target, but it is the job of others to disable or distract air defenses along the way. To do this, they use a form of AI under human supervision to detect targets or threats and plan possible routes, the source added.

Signal interference

The need for AI-powered drones is becoming increasingly urgent as both sides introduce electronic warfare (EW) systems that jam signals between pilots and drones.

In particular, small, cheap FPV (First Person View) drones, which were the main way for both sides to attack enemy vehicles in 2023, are seeing their hit rates drop due to increasing jamming.

“We are already working with the concept that in the near future there will be no connection between pilot and UAV on the front lines,” said Max Makarchuk, the AI ​​lead at Brave1, a defense technology accelerator set up by the Ukrainian government.

According to Makarchuk, the percentage of FPVs that hit their target is constantly decreasing. Most FPV devices now achieve a hit rate of 30-50%, while for new pilots it can be as low as 10%.

He predicted that AI-controlled FPV drones could achieve hit rates of around 80%.

To counter the threat of electronic warfare, manufacturers like Swarmer have begun developing features that allow a drone to lock on to a target using its camera.

Electronic warfare (EW) systems create an invisible jamming dome over the equipment and soldiers they protect.

If contact between the pilot and the drone is lost, the pilot can no longer control it and the aircraft either falls to the ground or continues to fly straight ahead.

By automating the final part of a drone’s flight to its target, the pilot is no longer needed, thus eliminating the disruptive effect of electronic warfare.

AI-assisted drones have been in development for years, but have so far been considered expensive and experimental.

Bendett said Russia had developed AI-powered air and ground drones before the 2022 invasion and had some success.

In Ukraine, the main task of manufacturers is to develop a low-cost AI targeting system for drones that could be deployed en masse along the entire 1,000 km front line, where thousands of FPV drones are deployed every week.

Costs can be reduced by running AI programs on a Raspberry Pi, a small, inexpensive computer that has gained worldwide popularity even outside of the educational purposes for which it was designed.

Makarchuk said he estimates the cost of installing a simple targeting system that captures a shape visible to the drone’s camera to be only about $150 per drone.