How Ukraine uses cheap AI-guided drones to deadly effect against Russia

Ukraine’s drone war is evolving rapidly. Once a cheap answer to Russia’s artillery dominance, Ukrainian small and inexpensive Read more of our recent coverage of the Ukraine war

They are carrying more explosives and flying farther per dollar, says Andrey Liscovich of the Ukraine Defence Fund, which crowdsources non-lethal aid. When FPVs were introduced at scale at the beginning of 2023 they could fly 10km or so, notes Yaroslav Filimonov, the CEO of Kvertus, a Ukrainian firm which makes anti-drone gear. Now 30km flights are routine, thanks to more powerful antennae and signal-repeater drones that let them communicate with base stations from greater distances. They are becoming more diverse, too. Large “bomber" drones scatter 10kg landmines on Russian supply roads. FPV “interceptor" drones have now taken out more than 850 Russian surveillance drones in the air, according to Tochnyi, a research group, easing the burden on the country’s air-defence missile stocks and undermining Russia’s ability to co-ordinate strikes.

The biggest change of all is that electronic warfare—essentially jamming—has consumed the battlefield. That started with the jamming of GPS signals from satellites, which caused American precision-guided weapons to fail at alarming rates. It has extended to the communication link between pilot and drone. By the beginning of this year, says Gundbert Scherf, the co-founder and CEO of Helsing, a German military AI firm, jamming was no longer “spotty" but “pervasive". Dronemakers are having to change their radios and antennae at ever shorter intervals. They face a dilemma: if they rely on common frequencies, the parts are easy to source but jamming is worse; rare frequencies are jammed less, but require more obscure hardware. The solution is to avoid jamming altogether by relying on AI to guide the drone to its target in the final stages of flight.

Rudimentary object-recognition software has been in use, on both sides, for over a year. But it is getting better. Lorenz Meier of Auterion, a firm based in America, says that between spring and summer his firm’s software, known as Skynode, managed to double the range at which a drone could engage a target, from 500 metres to 1km or so. He says that improvements in the resolution of images captured by drones have since increased that further. The Economist understands that AI systems are sometimes locking onto targets at perhaps double that distance, far beyond the range at which basic jammers could take out the drone.

Data from the battlefield suggest that the hit rate for these AI-guided drones is currently above 80%. That is higher than the rate of manually piloted drones. As important, the training burden declines dramatically. Mr Liscovich notes that, although there are now more highly experienced drone pilots, some with thousands of hours of flying time, the average quality of Ukrainian personnel has fallen over time as less motivated people are conscripted. “We can train an operator within 30 minutes and the quality of the engagement doesn’t depend on their piloting skills," says Mr Meier.

The result is that Ukraine has become the furnace of a new kind of software-defined warfare which combines precision with mass. Helsing is selling Ukraine 4,000 of its HF-1 strike drones, which it says will have the same payload as the Russian Lancet, around 5kg or so, and perhaps triple the range (up to 100km) but at a lower price (the Lancet costs around $30,000). Auterion, which had its first combat engagement in the spring, plans to field tens of thousands of drones powered by its software by early next year, with each unit (a chip pre-loaded with software) about the cost of an Android smartphone.

In both cases the drones themselves are made in Ukraine, by Ukrainians. One advantage of that is scale. Auterion’s largest partner in Ukraine, one of many, churns out 300,000 drones per year. Although recent Chinese sanctions have threatened to disrupt Ukraine’s drone supply chain, Mr Meier says that alternatives from Taiwan are now available. “We’re sitting on top of a consumer and automotive electronic supply chain—tens of thousands are nothing for these industries." Many Ukrainian manufacturers are pre-emptively switching to Ukrainian and European components too.

Mr Meier reckons that fewer than a tenth of drones are AI-guided at present. But that number is rising. Ukraine’s innovation cycles are relentless, with feedback loops in some cases down to a few days. The simplicity of FPV drones, which Ukraine builds from off-the-shelf components, can also act as a hindrance: it makes them easy to replicate. One manufacturer says Russian reverse engineering can be as quick as three weeks, but encryption means that the software can be protected from copying.

Yaroslav Azhnyuk, the founder of The Fourth Law, says that his own “autonomy module" is around $50 to $100 per unit for Ukrainian customers buying thousands of units. “There are Western firms which charge one to two orders of magnitude more for systems with similar or weaker capabilities," he claims. After pioneering much of the technology, Ukraine is still ahead of Russia—just ahead in airframes, but perhaps by three months in software for autonomy, which is encrypted and harder to reverse engineer.

Finance remains a huge problem. The Ukrainian state is buying more than it once was, but voluntary foundations still provide at least a third of all drones used by the army. Western companies are yet to fully commit to Ukraine’s burgeoning industry. Mr Azhnyuk complains that his other tech business—Petcube, a remote-controlled system for pet owners to take care of pets using lasers and mechanical treat-givers—received more in foreign investment than the whole of the Ukrainian drone sector in 2023.

The tech entrepreneur rejects talk of military automation as some kind of dystopian future. “Using AI to accurately target is far more ethical than lobbing missiles and artillery," he says. Ultimately, a human still has to make the final call on any engagement, says Mr Scherf. But Western and Ukrainian companies are busy working on deep-strike drones whose AI systems will be able to hunt for a wide range of potential targets far from the human operator. Mr Azhnyuk of The Fourth Law sees current technology as just the start. He hopes to have a prototype of a fully automated system, from launch to strike, built by early next year.

To stay on top of the both the biggest European stories and our defence and international security coverage, sign up to Café Europa and The War Room, our weekly subscriber-only newsletters.

© 2024, The Economist Newspaper Limited. All rights reserved. From The Economist, published under licence. The original content can be found on www.economist.com