NATO’s crystal ball? The program that claims to predict war
Advances in AI and data garnered from Russia’s invasion of Ukraine mean new systems are capable of predicting enemy attacks — but the process isn’t easy.
Anna Belova, a graduate of Kaliningrad state technical university in Russia’s militarised Baltic exclave, is a glamorous young woman with a blonde perm and five strings of pearls around her neck.
She is also a Russian spy, suspected of sniffing around the Swedish military.
With one click of a mouse, a dense web of Belova’s connections flashes up on the computer screen, ready for an intelligence analyst to trace the patterns in her biography, from her alma mater to a pronounced interest in a naval base just north of Stockholm.
“It’s like the mind palace of Sherlock Holmes,” said Rait Arro, a senior product developer at the Estonian military intelligence company SensusQ.
The fictional Belova is part of one of the less heralded revolutions in warfare and espionage. The modern battlefield is a torrent of data points, flooding in not only from radars, drone cameras and other sensors but also from more unorthodox sources such as Telegram channels and location signals from smartphones.
In theory this data is a tremendously powerful tool, rendering the situation “transparent” and making it all but impossible for an opponent to seize the element of surprise.
In practice, though, the task of weaving it together into strands of meaning and getting it to the people who need it most has often proved overwhelming, sucking up astronomical amounts of computing power to little effect.
After one British army exercise, the commanding officer memorably complained to Jack Watling, a land warfare expert at the Rusi think tank in London, that the software “ate like an elephant and shat like a mouse”.
Yet a new generation of technology, driven by advances in artificial intelligence and engorged with data from the Russian invasion of Ukraine, claims to be able not only to make sense of the present but to predict the future too.
SensusQ, which works with a number of western militaries and security services, claims that its “crystal ball” system can predict developments on a tactical level - such as where the next enemy attack on a particular sector will come from, and how hard it will hit, up to two days in advance. On a strategic level covering the grand sweep of military planning, the company estimates that it can forecast trends as far ahead as four to six months.
“You’re taking hundreds of thousands of indicators and trying to work out where the future is going,” Arro said. “We’re aiming to move from situational awareness to situational understanding: where is this manoeuvre coming from? Is it doctrinally motivated? How are we going to respond to it?”
Like most businesses with links to the military and intelligence worlds, SensusQ is guarded about the fancier end of its capabilities. It argues that even setting out a hypothetical case study would risk giving away sensitive insights to hostile powers.
Pressed for more detail, however, it pointed to an incident in 2018 when America was alarmed by the movement of a mechanised battalion of Syrian regime forces and Russian Wagner mercenaries near a platoon of US marines in the Syrian province of Deir Ezzor. The marines were outnumbered ten to one. Predicting the danger, the Americans immediately called in rocket artillery, combat aircraft, strike drones and helicopter gunships to strengthen their position. The enemy came under such intense bombardment that they had to back off after four hours with heavy losses. The decisive difference was foresight.
The underlying idea is not new. Systematic, a Danish software firm that is one of the biggest European players in the sector, has been around since the mid-1980s. Probably the best known outfit is Palantir, an American data analytics firm named after the magical seeing stones from the Lord of the Rings books and co-founded by the technology billionaire and Trump supporter Peter Thiel.
With close ties to the Pentagon and a starring role in its “Project Maven” algorithmic warfare project, Palantir is reported to be helping the American and British militaries feed the Ukrainians with intelligence strung together at a US facility in central Europe.
Marko Kaseleht, co-founder and chief executive of SensusQ, hit upon the idea of founding his own military intelligence firm while serving as an Estonian army squad commander in Afghanistan. He said “critical” intelligence-sharing failures had had fatal consequences for his unit when it was ambushed by the Taliban in 2009.
“One unit member was killed, and several others were seriously injured,” he said. “Afterwards we discovered that headquarters had intercepted Taliban communications setting up the ambush. However, outdated tools and analogue intelligence-sharing meant the information never reached us, lost in the chaos of multiple units and a vast area of operations.”
SensusQ, which was also founded by Villiko Nurmoja, another Estonian army veteran, has drawn a number of distinguished figures onto its board, including Riho Terras, a former head of the Estonian military, and Stephanie Bellistri, previously one of the top officers in the CIA. It is also advised by Toomas Hendrik Ilves, Estonia’s pugnacious former president.
It is tight-lipped about its list of customers, but alongside its headquarters in Tallinn it has outposts in London and Ukraine and supports the western coalition deployment in Iraq.
As well as guiding intelligence analysts through the morass of data, SensusQ and its Winning Mind software also have the potential to rapidly identify the most important threats or targets and offer advice on how to eliminate them with minimal resources.
One of the firm’s projects can map incoming fleets of enemy strike drones, evaluate their threat level and then suggest the best piece of weaponry for knocking them out of the sky, such as a jammer or a nearby air defence battery.
It has also developed a secure app called Verdandi, which is used by soldiers on the front line to file real-time reports on enemy activity.
The Times
