Direct air combat between UAVs is here to stay, and the Ukraine war is the primary crucible of its development. We are already seeing reconnaissance and attack UAVs intercepted by kamikaze drones that collide with them, and others that deploy capture nets or communication/navigation signal jammers, for example, while AI-driven automatic target recognition and autopilots capable of agile combat manoeuvring are developing rapidly. However, air defence is one of the most demanding of military disciplines, and there is a death spiral of cost and complexity awaiting anyone who tries to pack too many capabilities into a platform bounded by the constraints in size, weight, power and cost that typically face drone developers, writes Peter Donaldson.

Air defence is a problem best tackled with a system-of-systems approach, which Britain had implemented by 1940. This historical solution provides a useful analogue for counter uncrewed aerial system (C-UAS) development today. UK air defences in the Battle of Britain linked the Chain Home radar system via the telephone network with sector stations – that provided command and control – and airfields supporting manned fighters equipped with two-way voice radios. That infrastructure provided fighter direction capabilities that allowed the aircraft designers to focus on key attributes such as speed, rate of climb, manoeuvrability and firepower in a machine simple enough to be operated effectively by one person with the technology of the day.

UAV interceptors, likewise, need to be as simple as is compatible with effectiveness, and their integration into a layered and networked C-UAS architecture is the way to go. The building blocks of such architectures consist of networked ground-based and even airborne radars, electro-optical/infrared (EO/IR) sensors and radio frequency detectors, command and control software to fuse detection data and allocate targets to ‘shooters’ and a secure, low-latency data link. Integrating the interceptor at the sharp end of this kill chain – alongside other effectors such as jammers, lasers, guns and missiles – mitigates several of the engineering challenges facing interceptor developers.

It greatly reduces the sensing and processing burden because the UAV doesn’t need a high-performance radar, EO/IR system or AI computer, and can therefore be smaller, lighter and cheaper. Fed most of the target data it needs over the link, it can manage with a simple sensor for the final approach and terminal guidance.

Furthermore, the interceptor can be launched on demand instead of wasting energy in a standing patrol, meaning that its fuel tank/battery capacity can be smaller, solving the loiter-time/performance trade-off.

The battle management software can make intelligent decisions such as allocating a net capture drone to a target over a populated area, sending a kamikaze interceptor to a single drone over an open field, or tackling a large swarm with a combination of ground-based jammers and lasers, and sending interceptors after ‘leakers’. If enough interceptors are available, they can be used as a swarm to counter a swarm, with orchestration and deconfliction provided by the central command system.

Consequently, interceptors can be both relatively dumb and cheap, and specialised tools, each used only when it is the most appropriate and cost-effective solution. This way, the entire system can be more effective and resilient, and that cost-and-risk death spiral can be avoided.