A sense of things to come

Shortly before his death, looking back on a career that swept him from the battlefield to Downing Street to a seat in the House of Lords, the Duke of Wellington presented a final maxim for his successors to follow. “All the business of war, and indeed all the business of life,” he said, “is to endeavour to find out what you don’t know from what you do – that’s what I called ‘guessing what was at the other side of the hill’.” When the Duke made this statement in 1852 and decades after his famous victory over Napoleon at Waterloo, the tools available to generals were clearly rather limited. Hussars could scout for enemy positions, their horses sometimes shadowed by observation balloons. Spies could snatch documents, and sentries could be captured and interrogated. But often, intelligence gathering was really about guesswork, about putting a plan into action and hoping it worked.

Now, of course, everything is different. Buoyed by satellites and drones, today’s warfighters are increasingly able to understand the precise layout of the battlefield in an instant, spotting friends and foes, whatever the terrain. As Professor Paul Thomas puts it: “Modern operations, while raising the tempo of this demand to the extreme, are enabled by modern sensing technology, which can detect stealthy and manoeuvring targets at impressive ranges.” However, as the fellow at UK’s Defence Science and Technology Laboratory (Dstl) continues, contemporary situational awareness has issues all its own – especially in the complex field of sensors. Ensuring devices work quickly and efficiently, alerting soldiers to any incoming threat, is vital. Yet, older platforms are often inappropriate, putting undue burdens on personnel, even as the systems themselves struggle with the rigours of contemporary warfare.

Now, however, Thomas and his colleagues are working to drag this technology into the 21st century. Known as SAPIENT (Sensing for Asset Protection with Integrated Electronic Networked Technology), they’ve developed a new architecture aimed at making situational awareness robust enough to deal with any situation. Exploring the latest developments in modularity and AI, it removes the cognitive stress from soldiers themselves, instead giving them personalised information for whatever fire they’re fighting. Nor is SAPIENT merely being employed in the service of His Majesty. On the contrary, the system has already proved useful in Nato exercises, while the developers are already planning on honing it yet further.

Sensing a problem

It’s hard to overstate the power of situational awareness over today’s militaries. Encompassing a field globally worth around $28bn – a figure set to more than double by 2030 – it’s fundamental to how modern militaries go about their business. This instinctively makes sense: understand where and how the enemy is moving, and you stand a much better chance of keeping your own troops secure. This importance is equally obvious, moreover, if you consider the many practical dangers armies now face. In Ukraine, for instance, both sides have had to deal with drone attacks, with a swarm of 30 unmanned aircraft recently attacking Zelensky’s forces near Kyiv. And if humans alone could never hope to catch such threats – let alone do anything to stop them – sensors can often prove more successful. Ukraine itself has instituted just such a network, deploying a suite of counter-unmanned aerial systems (C-UASs) to defeat Russian drones.

Yet, as Thomas explains, traditional situational awareness platforms suffer from a number of problems. That begins, he suggests, with a lack of bandwidth. Scarce enough on the battlefield as it is, the situation is made even worse by the fact that sensors have traditionally collected information locally, before sending it to be processed and understood in some central office elsewhere. The result, says Thomas, is “bandwidth-breaking data streams” being moved over “creaking” communications systems – hardly ideal when lives and materiel are at stake. In a similar vein, older platforms have largely relied on individuals to process the raw information sensors gather. But as Thomas warns, the “cognitive burden” involved can often be intolerable, especially when we know that attention spans invariably slump over time. Combined with the fierce sophistication of modern operations – tanks, infantry and logistics all need to be protected together – and it’s unsurprising that existing networks can be expensive to run for long periods of time.

Certainly, this seems clear from the numbers. According to recent work by the Molfar NGO, for instance, the period from 13 September to 17 October 2022 saw Ukraine spend $28m to defeat Russian drones. It goes without saying, meanwhile, that beyond the financial impact, there are inherent risks for armies without sufficient situational awareness. According to GCHQ, Russia has reportedly lost a number of fighter jets because pilots flew over enemy territory by mistake. They aren’t the only ones to suffer in this way either. In 2014, for example, a US air crew in Afghanistan launched two laser-guided bombs against friendly troops on the ground, killing six. “The key members executing the close air support mission collectively failed to effectively execute the fundamentals,” noted the subsequent report into the accident, “which resulted in poor situational awareness and improper target identification”.

Unmanned, unbound

With all this in mind, it should come as no surprise that, in 2013, Dstl announced it was working to improve situational awareness functionality. Jointly funded by the MoD and Innovate UK, from the beginning, SAPIENT has sought to use technology to upgrade both military and civilian operations. Whatever the target audience, at any rate, it’s clear that Thomas is extremely excited about what the new set-up can provide. Born out of a desire to exploit the “dual revolutions of AI and autonomy”, he argues SAPIENT, in its latest iteration, has “a wide variety of applications”.

How, then, does SAPIENT work in practice? Crucial is the way it abandons the centralisation inherent to older situational awareness systems. Rather than indiscriminately sending reams of data back to headquarters, then expecting humans to sift through it, SAPIENT instead uses AI to decide what really matters – only providing troops with truly actionable information. That’s echoed by a related automated function, which uses AI to make decisions on the spot.

By way of example, it could mean that a camera or sensor chooses to zoom in on a particular enemy position, providing intelligence officers with more robust guidance automatically. That’s twinned with other gadgets, which help minimise the risk of false alarms, and which correlate what a SAPIENT-linked sensor is saying. At its best, in short, Thomas says that Dstl’s work means operators have “key information about the unfolding scenario” – allowing them to make decisions faster and with more accuracy.

If that weren’t impressive enough, moreover, SAPIENT also offers fixes in terms of interoperability. Boasting a so-called open architecture, it’s easy to integrate across a number of different sensors – even while the system is in action. Considering the complexity of modern joint arms militaries, and the fact that organisations like the US Army hire literally thousands of external contractors, this is probably wise. At the same time, Thomas notes that flexibility means systems can be sharpened at speed, allowing operators to combat “rapidly-advancing threats and enabling adoption of improved technology”. Even better, it doesn’t lock the MoD into working with particular vendors, hardly immaterial given how frequently such partnerships go wrong.

Simplicity and reliability

The real-world practicality of SAPIENT’s open architecture was obvious in 2021, when it was used at Nato’s C-UAS Technical Interoperability Exercise (TIE) in the Netherlands. Apart from demonstrating the simplicity of SAPIENT to suppliers – Thomas reports that implementing the platform into their equipment was “straightforward” – it has also helped developers hone their product even further. Among other things, that includes adopting Google’s ‘protobuf’ serialisation protocol, a data format that offers lower bandwidth usage and backwards compatibility. 2022’s TIE meetup, again in the Netherlands, proved the wisdom of these changes: 40 different systems were integrated using SAPIENT, while Nato has adopted the platform as one of its C-UAS standards.

Not that Thomas and his fellow developers are happy to sit back just yet. As part of SAPIENT’s original aspiration to straddle military and civilian life, Dstl has been working with the National Police Chiefs’ Council, as well as the Home Office, to integrate new technology into their own capabilities.

Similar relationships with equipment suppliers have been formed as well, meaning that police forces across the UK will soon have access to what Thomas calls a “multi-modality sensor, effector and decision-making suite” compatible with SAPIENT. That, he adds, “creates a supplier marketplace that benefits the MoD’s procurement activities”. Together with other schemes, notably the opening up SAPIENT to other areas of the public and private sectors, it’s hard not to feel that Dstl’s invention is here to stay. If only the Iron Duke himself was around to see it happen.