<

Blog

>

Counter-UAS: From Detection to Decision

How multi-sensor data fusion and a digital targeting web connect what sensors detect to how an organisation responds.

Why connecting sensors isn't enough

In the first part of this series, I argued that counter-UAS is a systems integration problem, and that the hard part is getting existing sensors, effectors and decision-makers to work as one. Connecting sensors to a common operational picture is the first step. On its own, though, it doesn't make an effective multi-sensor system. Putting several independent feeds on the same screen gives an operator more information to process, and that doesn't always leave them with a better understanding of what's happening around them.

The bigger gain comes from combining measurements from different sensors into information that's more useful, reliable or precise than any one of them could provide alone.

How multi-sensor data fusion works

Multi-sensor data fusion is a core technical capability at StirlingX, and it draws on decades of experience in autonomous systems, tracking and information fusion. We combine measurements from different types of sensor, track multiple targets across multiple sources, and work out which observations are likely to belong to the same physical object.

Take an acoustic sensor that picks up the first sign of an approaching aircraft. That cue can point an electro-optical camera at the right area of sky. The camera's measurements are then associated with any other observations available and folded into a single estimate of the target's position and movement. Where sensors provide complementary or partly independent information, the fused estimate can be more accurate, more continuous and more resilient than any individual measurement.

In our architecture, those two jobs sit with our two platforms. LEWES, our connected autonomy platform, does the cueing, tasking the right sensor to the right place at the right time. MAYNE, our data analytics platform, carries out the fusion, bringing every source into a single track picture.

Fusion also brings in semantic information alongside conventional tracking measurements. Radar micro-Doppler signatures can reveal how a target moves, including whether it has rotating parts. Computer vision models can classify objects from electro-optical imagery. Passive RF observations can indicate what type of system is operating and how it communicates.

Bringing these together can improve track association, help identify targets and reduce false alarms, depending on the quality and reliability of the measurements available. It also gives the system a far more informative picture of the operating environment than a list of separate detections.

Adding more sensors doesn't automatically give you better information. In information-theoretic terms, the aim is to exploit complementary observations while properly accounting for uncertainty, dependence between sensors and the limits of each one. The value comes from combining measurements in a way that's mathematically consistent and makes sense operationally. That's how scattered observations become information an operator can act on.

From a shared picture to a digital targeting web

A clear operational picture helps identify and track potential threats, but counter-UAS also has to connect that picture to the decisions and responses available to whoever is being protected. That's the role of a digital targeting web, a concept the UK's 2025 Strategic Defence Review placed at the centre of its plans, with more than £1 billion committed and delivery targeted for 2027.

A digital targeting web avoids fixed, proprietary links between individual sensors and effectors. It provides the infrastructure that lets multiple sensing systems, decision-makers and response options share information. A detection from one sensor can feed a fused track, prompt another sensor to take a closer look, and inform the choice of response. Together, LEWES and MAYNE provide that layer between sensor and decision, so any sensor can reach any decision-maker and any effector.

In practice, a low-cost RF detector and a high-performance air defence radar can both feed the same operational picture, even though they differ widely in capability, cost and purpose. That same picture can support responses ranging from electronic warfare to kinetic interceptors, depending on the threat, the operating environment and the rules of engagement in force.

Choosing a proportionate response

Having that choice matters because uncrewed aircraft vary so widely in cost, sophistication and the way they're used. The cheapest available response isn't always the proportionate one. A proportionate response balances effectiveness, available resources, operational risk and the value of the asset being protected.

Built for interoperability

Our architecture is sensor-agnostic and effector-agnostic. Integrating a new device still takes engineering work, because different systems have different interfaces, data models and control requirements. What the architecture provides is support for interoperability through appropriate standards, APIs and integration mechanisms, so customers aren't dependent on any single sensor or effector manufacturer.

Customers can then choose the combination of systems that suits their operational requirements today, and change it as threats and technology move on.

Share this post

StirlingX brand mark

Ready to see what’s possible?

Experience secure autonomy in action. Partner with StirlingX and transform the way you monitor, manage and protect your assets.