Ukrainian developers are working on technology based on artificial intelligence, which will make it possible to connect several dozen drones under the control of one pilot into a single network.
Points of attention
- The use of artificial intelligence will allow drones to operate in swarms and even automatically capture targets.
- Combat drones with AI will be able to increase the accuracy of strikes by up to 80%.
- The main directions of development include the creation of visual systems and programs that unite drones in a flock for joint performance of tasks.
AI will be able to control a swarm of dozens of combat drones
At once, several Ukrainian startups are developing artificial intelligence (AI) systems that enable unmanned aerial vehicles to work under the influence of electronic warfare (EW) and to unite in groups to perform tasks.
According to the agency, the work is carried out in two main directions. The first direction covers the creation of visual systems that would help identify targets and direct UAVs to them, as well as terrain mapping for navigation. The second direction is programs that enable drones to work in interconnected "flocks".
The Swarmer company, in particular, is working on similar technology. Engineers develop Styx software that connects drones in a swarm.
The system controls a reconnaissance network and attacks drones of various sizes in the air and on the ground. Each UAV can plan its route and predict the behaviour of others in the swarm. Human intervention is required only to allow the command to strike.
When you try to scale (with human pilots) it just doesn't work. Managing a swarm of 10 or 20 drones or robots is almost impossible for people, — noted Swarmer CEO Serhiy Kuprienko.
According to Serhiy Kuprienko, it is difficult for UAV operators to manage operations involving more than five devices, while AI can process hundreds of data simultaneously. In addition, automation will help protect operators working near the front lines.
Styx technology is currently in development and has only been tested on the battlefield as an experiment.
FPV drone hit percentage reduces due to the EW devices on the front line
In his turn, Max Makarchuk, the head of artificial intelligence in the Brave1 defence cluster, spoke about the work on the concept of drones with AI that will be able to work even if there is no communication on the front line in the future.
According to him, the front line is increasingly saturated with EW means, which means the percentage of FPV drones hitting the target is constantly falling. Therefore, there is a growing need for UAVs that can hit targets in obstacle conditions.
Max Makarchuk notes that the average percentage of FPV drones hitting the target is 30–50%. With AI, that number is expected to rise to around 80%.
To achieve this goal, Ukrainian companies, in particular Swarmer, equip UAVs with the function of capturing a target using a camera. Automated homing drones no longer require an operator to strike.
Currently, Ukrainian manufacturers must create a cheap AI guidance system for drones in order to use such drones en masse across the front line. According to Maks Makarchuk, installing a simple system that allows you to guide a UAV to a target automatically is almost $150 per device.