Since they were first used in the First World War, when unmanned aircraft were guided by radio to attack zeppelins, military use of UAS has become increasingly widespread. While hobbyists operating drones in public areas or near private property for filming purposes has become a mere nuisance, drones of all sizes are also becoming easier to acquire and weaponise. Their improved speed and endurance capabilities pose a growing threat to defence systems and critical infrastructure. In the war against Ukraine, to give one example, Russia has used Iranian-made Shahed-136 drones to target critical infrastructure. Boasting a range of over 1,000km, these drones were able to bypass Ukrainian defences by flying low and striking from unexpected directions.
In response to these escalating UAS threats, counterunmanned aircraft systems (C-UAS) technologies are being designed and developed using AI and machine learning (ML) technologies. According to industry reports, the global AI in drone technology market size is expected to be worth around $206.9bn by 2033, up from $12.5bn in 2023. Once again, the rising importance of C-UAS is clear enough in the real world, with Ukraine and Israel just two of the countries to invest in the technology. Despite their increasing relevance in battlefields the world over, however, C-UAS technologies face the obvious challenge of keeping pace with developments in UAS – with failure meaning they’re unable to successfully counter the aerial threats they were designed to defeat.
“With an ever-evolving threat from state and non-state actors, drone warfare has clearly changed the character of warfare and with it has introduced significant challenges,” emphasises David Shank, former commandant at the US Army Air Defense Artillery School and Integrated Air & Missile Defence, and now a C-UAS consultant. “These challenges have been recognised in recent conflicts across the Middle East, Ukraine and the Nagorno-Karabakh area.”
Integrating AI
AI enables drones to be manoeuvred in complex environments with enhanced autonomy and intelligence, utilising data and increasing efficiency in their set tasks, but these increased capabilities also mirror their ability to be leveraged as offensive weapons. At any rate, these developments require defence authorities to counter the threat of AI-enabled drone attacks, particularly in cities, as an urgent priority. Certainly, there are good reasons for integrating AI into C-UAS systems – enabling as it does faster response times, increased precision, adaptability and the ability to monitor several drones simultaneously, all of which helps increase the protection of infrastructure and the safety of urban areas. Fundamentally, these developments can be understood technologically. AI-enabled C-UAS enhances the detection, identification, tracking and neutralisation of drones by using sensors like radar, radio frequency (RF) scanners and cameras. AI algorithms, for their part, help filter out noise and identify threats by analysing a drone’s characteristics, enabling the AI to monitor a drone’s movements and evaluate its threat levels to inform quick decision-making abilities.
$206.9bn
The predicted size of the global drone AI technology market by 2033.
Market.us
From there, AI can then suggest or autonomously deploy countermeasures, such as jamming signals that interfere with communication systems. AI may even be able to take over control of the drone or deploying projectiles or lasers to neutralise the threat.
“The introduction of AI has greatly decreased the necessity of service members being part of a full kill chain process,” says Shank. “Autonomous systems have taken those previously responsible for step-by-step actions ‘out of the loop’ allowing for machine-tomachine interfacing and AI actions to fill these steps up until a decision to engage is made.”
Upgrading traditional platforms
Not that AI-bolstered C-UAS requires a totally clean slate. Rather, traditional anti-aircraft platforms can be upgraded with AI-powered sensors to detect, track and identify UAS. Command and control (C2) systems, for their part, can also be improved with AI to enhance decision-making tools, even as it can be integrated into the targeting systems of traditional anti-aircraft weapons for automated or partly automated drone targeting. At the same time, AI-integrated C-UAS can be a component of a layered defence system, as the AI can efficiently coordinate different layers, prioritising larger threats with long-range systems and smaller drones with shorter-range defences. “Legacy systems will not go away overnight and the need to integrate these systems into and with modernised capabilities must take place,” emphasises Shank. “It is an engineering challenge, yet it lays a foundation for a layered 360-degree defence integrating legacy and modernised systems, all managed by a cyber-hardened C2 networked architecture.”
This focus on C2 makes sense. With the addition of such a platform, encompassing a joint common operational network, information can be shared more rapidly across different C-UAS units. This approach also increases a commander’s decision space, and eases the burden of users as the AI provides more information. From there, AI also allows for distributed operations, enabling multiple traditional anti-aircraft units to work together while covering larger areas by managing the network and assigning targets based on various factors. When human reaction time is a limiting factor, autonomous operations can also be conducted, particularly in high-threat environments such as missile defence. For Shank, leveraging a high-tech solution – comprising advanced electro-optical (EO) and infrared (IR) sensors, integrated with AI software and a C2 platform – enables users to autonomously scan, track and classify drones. This increases the levels of operational effectiveness while maintaining a low probability of intercept (LPI) and low probability of detection (LPD). With these strengths in mind, at any rate, it’s unsurprising that AI-powered C-UAS is enjoying attention from the world’s greatest military power. Earlier this year, for instance, the US Department of Defense’s (DoD) Chief Digital and Artificial Intelligence Office (CDAO) reported that it was seeking demonstrations for sensors enabling increased and enhanced detection of UAS.
“CDAO’s interest is to ensure C-UAS end systems (sensors and effectors) and associated C2 systems are open and accessible for both the data and software,” says Commander Jessica Anderson, a spokeswoman at CDAO. “Data availability is a core building block of developing and utilising advanced algorithms. Another area is improving the decision systems to assist the human operating in selecting the appropriate weapon of choice to adjudicate targets. We maintain humans in the loop but understand AI assistance can improve decision cycles.”
Certainly, by upgrading traditional anti-aircraft systems with AI-enhanced C-UAS capabilities, militaries can significantly improve their ability to defend against the growing threat of drones, while equally maintaining the flexibility and robustness of existing platforms. “A key step in implementing advanced algorithms is opening the data,” is how Anderson puts it. “This aligns with CDAO’s approach with OpenDAGIR by ensuring each contract we enter into has the appropriate language to expose the right data for government/ mission use.”
Not that any of this is straightforward. After all, the benefits of integrating AI into traditional platforms are clouded by the increased cybersecurity risks inherent in the technology. Troops also need to ensure that new AI platforms can communicate effectively with legacy systems. In practice, that means training personnel to operate novel systems ultimately, informing autonomous or semi-autonomous decision-making.
Continuing research and development
Beyond these technical challenges, there are ethical questions to be answered here too. As AI-integrated C-UAS systems become more widely used, after all, regulations will need to be updated to ensure their safety and ethical use. Military lawyers and technicians are also under pressure to define the boundaries of AI autonomy in threat detection and engagement – unsurprisingly an especially important concern in civilian areas. In practice, legislation will be needed to govern the deployment of C-UAS and ensure accountability in decision-making processes, particularly with the use of autonomous systems. Furthermore, establishing international standards can help to manage the cross-border implications of AI-driven C-UAS and prevent misuse – while also promoting collaboration across shared or contested airspaces. As Shank notes, that’s shadowed by continued innovation in the technical space. “Research, development, testing and experimentation (RDT&E) continues in an effort to maintain pace with adversaries, though there are numerous C-UAS capabilities which already exist at a technology readiness level (TRL) of 9,” says Shank, referring to the maturity level of technology, whereby level 9 means the system has been proved through successful mission operations. “Continuity of system integration with joint forces will be key to disseminating this technology and information across multi-domain operations.”
“Beyond RDT&E are unit field training and major exercises at the ‘tier 1’ level among individual services and joint and combined forces,” the expert adds, referencing the highest level of operations or units that are involved in the most critical, complex and demanding missions. “These opportunities will help introduce emerging drone and counter-drone technologies, demonstrating increased efforts in support of combined operations and providing service members hands-on training.”
These efforts are surely worthwhile. For if the killing fields of Ukraine have already shown the impact of drones, and indeed C-UAS, AI may yet prove pivotal in future aerial defences. If, after all, China finally tries to take Taiwan, drones could play crucial roles in surveillance and electronic warfare – and help overwhelm Taipei’s air defences. The same could plausibly be said of the Middle East, where any future conflict between Israel and Iran will undoubtedly be shaped by UASs and the technology to counter them. Given these geopolitical tensions, at any rate, the investments of the DoD and others surely make sense.
